Doing research in Power Electronics

by Bimal K. Bose

My View

6 IEEE INDUSTRIAL ELECTRONICS MAGAZINE MARCH 2015

Do you do research in power electronics? Are you a profes-sor or graduate student in a uni-versity or an engineer in an industrial research laboratory? Power electron-ics research is not different from any other area of engineering or scientific research. For doing research, we dis-cipline and dedicate our mind to make inventions or generate new knowledge that helps to solve our problems and contribute to the advancement of our civilization in a broad perspective. The sincerity, purity, and tranquility of our mind, possibly blended with some spir-ituality, help us concentrate our mind for doing research. Research accom-plishment gives supreme satisfaction of mind. Note that doing research and learning always go together. Learning is essentially a lifelong process. Albert Einstein said that we cease to learn only when we die.

How does our brain function for doing research? The human brain, the thinking machine with a biological neu-ral network that gives us natural intelli-gence, is the most complex machine on earth. Neurobiologists have attempted to understand the structure of the brain and its functioning over a pro-longed period of time, but these remain extremely inadequate even today. The neural network [1] in our brain consists of the interconnection of billions of neu-rons or nerve cells, where the synaptic junction of each input dendrite is filled with neurotransmitter fluid. The im-pedance of this junction contributes to the intelligence or associative-memory

property of each cell. The intelligence of the brain is thus distributed in the cells of the whole neural network. The supervised learning from our educa-tion conditions the junction impedanc-es to acquire knowledge in a specific domain, such as power electronics. It is interesting to note that our brain does not have a computerlike central memory. The neural network has the ability to interpolate or extrapolate this knowledge to create new knowledge. The brain has the additional cognitive capability to invent, which we do not really understand.

The creative capability of the hu-man brain is tremendous. We use hardly more than 5% of our creative capability for doing research, and the remaining is mostly wasted in the triviality of our daily thoughts. The research can be de-fined as fundamental or basic type and applied or application oriented. Thomas Edison, the wizard of applied experimen-tal research in electrical engineering, de-fined genius as 1% inspiration and 99% perspiration. Edison had 1,093 U.S. pat-ents even though he did not complete his high school education. However, according to Charles Steinmetz, the wiz-ard of basic research, genius is defined as 99% inspiration and 1% perspiration. These are, of course, extreme examples in the early days of engineering re-search. A modern research project typi-cally requires idea formulationsystem analysisdesignsimulation studyand validation by experiment. Finally, suffice it to say that a good researcher should also be a good communicator in both writing and speaking.

Now, let me fall back to power elec-tronics. What is special about power

electronics? It is a complex and inter-disciplinary technology that basically deals with the conversion and control of electrical power using switching-mode power semiconductor devices. The ap-plications of power electronics include regulated dc and ac power supplies, electrochemical processes, heating and lighting control, electronic welding, power line volt-ampere reactive (VAR) and harmonic compensation, high- voltage dc (HVDC), flexible ac trans-mission systems (FACTS), photovoltaic (PV) systems, fuel cell power conver-sions, high-frequency (HF) heating, and motor drives. After several decades of technology evolution, power electron-ics applications have recently become extremely important for energy saving, electric/hybrid vehicles, the smart grid, renewable energy systems, and bulk energy storage, besides the usual ap-plications in industrial automation and high-efficiency energy systems.

In general, doing research in power electronics requires expertise in power semiconductor and peripheral devices, converter circuits, control theories, electrical machines, digital signal pro-cessors (DSPs), field programmable gate arrays, power systems, and com-puter-aided design and simulation tech-niques. Recently, artificial intelligence (AI) techniques, such as fuzzy logic and artificial neural networks (ANNs), are advancing the frontier of power electronics. Each of these component disciplines is advancing rapidly, thus presenting greater challenges to power electronics researchers. A thorough knowledge of the application environ-ment is essential for doing research in a power electronics project.

Doing Research in Power Electronics

Digital Object Identifier 10.1109/MIE.2014.2364471

Date of publication: 19 March 2015

MARCH 2015 IEEE INDUSTRIAL ELECTRONICS MAGAZINE 7

In this column, discussions will mainly be based on my own experi-ence. The formulation of research ideas, project planning, proposal preparation, steps in doctoral research projects, and writing papers for publication will be covered. This column is aimed at young researchers at universities based on ex-amples from my own research career, reviewing my experience in power elec-tronics research, and illustrating my key contributions. I have spent more than 30 years in universities and more than 11 years in a premier industrial research laboratory, which practically spans the whole era of the modern power elec-tronics evolution. Hopefully, my knowl-edge and experience will be useful to the readers.

Research Areas in Power ElectronicsAs I mentioned, power electronics is a complex and interdisciplinary technol-ogy, and doing research in this area requires a comprehensive background in electrical engineering and beyond. Figure 1 shows some key inventions related to power electronics [2]. The research in power electronics can be broadly classified into research on devices, converter systems, motor drives, and general energy systems, which are summarized, respectively, in Figures 25. The motor drive area is al-ways included in power electronics be-cause complexity in this area is mainly due to power electronics.

The research on devices, particu-larly power semiconductor devices, is extremely important because evolution in this area has essentially brought on the modern power electronics revolu-tion. The present trend of research and development (R&D) on silicon and large-bandgap power semiconductor devices will continue until the power device characteristics and ratings are signifi-cantly improved, approaching an ideal switch. However, note that the basic research on power semiconductor and peripheral devices (including machines) does not strictly fall in the mainstream of power electronics, except for the eval-uation of their performances. Generally, every research project in power elec-tronics has the usual implementation

stages of design, analysis, modeling, computer simulation study, and experi-mental evaluation.

The converter systems and motor drives technologies have significantly matured in recent years, although there are ample opportunities for doing inno-vative and incremental research in these areas. The motor drives area is generally more complex and requires expertise in more interdisciplinary areas. Recently, power electronics applications have ex-panded into complex energy systems because of their integration with the utility systems, and research in this area

is expanding. Figure 5 gives some ex-amples of general energy systems. The modern complex smart or intelligent grid, where the conventional high-power fossil fuel, nuclear, and hydroelectric generators are integrated with distrib-uted renewable energy systems (such as wind and PV) along with bulk energy storage devices [such as batteries, fly-wheel, pumped storage, ultracapacitors, superconducting magnet energy storage (SMES), and hydrogen], need extensive research efforts for system stabilities, bus voltage, frequency control, power quality, optimum resource utilization

D-,7

8 IEEE INDUSTRIAL ELECTRONICS MAGAZINE MARCH 2015

with supply- demand interactive energy management, economical electricity to consumers, higher energy efficiency, higher system reliability and security, fault detection and protection, fault-tol-erant control, etc. The system complex-ity increases if it incorporates power electronics-based HVDC transmissions, FACTS, static synchronous compensa-tors, uninterruptable power systems, etc., to achieve the above- mentioned objectives. The AI applications in power

electronics are ushering in a new tech-nology frontier in this perspective and require further exploration. Additional information on research areas can be obtained from the power electronics literature. Power electronics is often de-fined as a maturing enabling technology. However, both ends of the technology spectrum, i.e., power semiconductor de-vices and complex energy systems with power electronics, require extensive R&D efforts.

Doing Research in UniversitiesNow, let us fall back to research in univer-sities [3]. Have you just completed your Ph.D. degree in a university and are look-ing for a job to start your career? Gener-ally, you have two choices: join the faculty of a university or be an engineer in an in-dustrial research laboratory. Of course, you also have the option to be a post-doctoral fellow in a university to work with a senior professor or start your own business as an independent researcher.

1) Power Semiconductor Devices:A) Materials (Silicon, Silicon Carbide, Gallium Nitride, Diamond, Etc.)B) Present Devices [Diode, Thyristor, GTO, Triac, Power MetalOxideSemiconductor Field-EffectTransistor (MOSFET), IGBT, IGCT, Intelligent Power Module (IPM), or Power Integrated Circuit (PIC)]with Incremental ChangeorDevelopment of a New Device

[Voltage and Current Ratings, Leakage Current, Safe Operating Area (SOA), Conduction Drop, JunctionTemperature, Turn-On and Turn-Off Times, Minority Carrier Storage, Rate of Current Change (di/dt),Rate of Voltage Change (dv/dt), Switching Frequency, Series and Parallel Operation, Power Loss,Thermal Impedance, Cooling, Snubber or Snubberless, Gate Drive, Fault Diagnosis and Protection, and Applications]Design, Fabrication, Packaging, Analysis, Modeling, Simulation Studies, Performance Prediction,and Experimental Evaluation

2) Peripheral Devices:Present Devices (Capacitor, Inductor, Resistor, Transformer, Battery, Fuel Cell, PV Cell,Light-Emitting Diode, SMES, DSP,Application-Specific Integrated Circuit, Field-Programmable Gate Array, Etc.) with Incremental ChangeorDevelopment of a New DeviceDesign, Fabrication, Packaging, Analysis, Modeling, Simulation Studies, Performance Prediction,and Experimental Evaluation

Present Converters (Voltage-Fed, Current-Fed, Hybrid, HF Link, Etc.) with Incremental ChangeorDevelopment of a New Topology [Control Strategy, Response, Loss and Efficiency, Line and Load Harmonics, Power Quality,Filtering, Soft Switching, Dead-Time Compensation, Power (P), Reactive Power (Q), Total Harmonic Distortion (THD),Power Factor (PF), Displacement Power Factor (DPF), Electromagnetic Interference (EMI),Line/Converter/Load Faults, Fault Diagnostics and Fault-Tolerant Control, Hardware/Software Implementation, Etc.]

Present PWM Techniques [Sinusoidal Pulsewidth Modulation (SPWM), Selected Harmonic Elimination (SHE),Static Volt Ampere Reactive (Var) Compensator Hysteresis Band (SVC.HB)]with Incremental ChangeorDevelopment of a New TechniqueDesign, Analysis, Modeling, Simulation Studies, Performance Prediction, and Experimental Evaluation

Design, Analysis, Modeling, Simulation Studies, Performance Prediction, and Experimental Evaluation

1) Converters:

2) PWM Techniques:

FIGURE 2 Research topics on devices.

FIGURE 3 Research topics on converter systems.

MARCH 2015 IEEE INDUSTRIAL ELECTRONICS MAGAZINE 9

Whatever it is, you have essentially cho-sen research as the main activity in your career path. It is very likely that you will select a university to start a faculty ca-reer. So, welcome as a tenure-track assis-tant professor in electrical engineering.

What are the merits and challenges in a university career? First, among the merits, you have a lot of freedom. Essen-tially, you are your own boss. You wont have to report for work during the 8 a.m.5 p.m. time frame. In a university, if the department is big, you seldom meet your department head, and he will never ask you how your work is go-ing. By adding Prof. with your name, you maintain a prestige and ego in the career. A university professor is an em-blem of intellectual thinking and is high-ly respected in society. You can travel to conferences as you want, provided you control the travel funds. A university job in the United States is typically for nine months out of the year. This means that in the summer you can work for

extra income. In addition, you can also do industrial consulting, typically one day/week, to boost your income. A pro-fessors nine-month salary may not be

necessarily less than that of an engineer for 12 months in industry. A professor, of course, has to be a good speaker as well as a good writer. For success in

2) Control StrategyPresent Control Strategies [Vector Control, DTC Control, Model Referencing Adaptive Control (MRAC),Self-Tuning Regulator (STR),Variable Structure Control (SMC), Model Predictive Control, Fuzzy and Neural Controls,Genetic Algorithm (GA) Control, Sensorless Control, DisturbanceCompensation, Fault-Tolerant Control and Other Scalar, Optimal, and Adaptive Controls,Hardware/Software Implementation] with Incremental Change

1) MachinesPresent Machines [Induction Motor (IM), Permanent Magnet Synchronous Motor (PMSM),Wound-Field Synchronous Motor (WFSM), Switched Reluctance Motor (SRM)with Radial, Axial, or Linear Geometry] with Incremental Change

Design, Fabrications, Analysis, Modeling, Simulation, Performance Prediction, and Experimental Evaluation

or

orDevelopment of New Machine(Volume, Weight, Power/Torque Density, Parameters, Losses and Efficiency, Cooling, Pulsating Torque,Acoustic Noise, Faults, Etc.)

Development of a New StrategyAnalysis, Design, Modeling, Simulation Studies, Performance Prediction, and Experimental Evaluation

3) Estimation and MeasurementsPresent Techniques [Torque, Slip, Flux, Speed, Position, Acceleration, Disturbance (Line and Load),Response, Accuracy, Harmonic Effects, Machine Parameters (Stator and Rotor Resistances, and Stator andRotor Inductances), dv/dt Effect on Insulation, Bearing Current, Acoustic Noise, Machine VoltageBoost for Long Cable, Fault Diagnosis (Online and Offline), Hardware/SoftwareImplementation, Etc.] with Incremental Changeor

Development of a New Strategy

Design, Analysis, Modeling, Simulation Studies, Performance Prediction, and Experimental Evaluation

FIGURE 4 Research topics on machines and motor drives.

Examples:3.%%%'#'-++.)-#%Po0+2,-&0#-".)+)#-(+ (+--+2-(+!)+-#'!0#-"#''+-#('2,-&-(.))%2 '-+--'-(,#''.-('(&(.,2,-&3#+(+#0#-"#''#,-+#.-'+!2-(+!e3!'-#%%2/#--"#%+',)(+--#('2,-&0#-" $-#%#-2.))%#,3+',,,#('2,-&'-!+-0#-"3' -"(+#''+-#('2,-&,'-!+-0#-"-#%#-2+#"+(.!" (%-!-(++',,,#('3&+-+#2,-&'-!+-0#-"(,,#%.%+''ew%'+!2 2,-&,'.%$'+!2-(+!

.,vo%-!' +*.'2p,*"+&('#,)(0+*.%#-2+%##%#-2,-#%#-2,-#&-#('&,.+&'-('-+(% ##'2,.+#-2 .%-#!'(,#s, .%---(%+'-('-+(%()-#&.&%( %(w,-(')-.%,-.#,,#!''%2,#s,&(%#'!,#&.%-#(')+ (+&')+#-#(''1)+#&'-%ev%.-#('

FIGURE 5 Research topics on general energy systems.

10 IEEE INDUSTRIAL ELECTRONICS MAGAZINE MARCH 2015

your career, these communication skills are extremely important.

The biggest challenge in a university is that a professor has to bring in research funds in a sustained way throughout his career. Although most universities have the status of nonprofit tax-exempt insti-tution, they tend to operate like profit-making corporations. For this reason, most of the faculty, particularly in the engineering departments, are always under tremendous pressure to bring in more and more research funds. Since most of the graduate students (M.S. and Ph.D. students) come from abroad and are usually supported by research funds, if you do not have research funds, you will hardly have any graduate students, i.e., you will not have any research proj-ects and, therefore, no publications. As a tenured faculty member, you are left with a full-time teaching load of four to five courses. Essentially, you are a dead professor. Again, it is a one-way street. Once you have left the research career, it is difficult to come back to it.

When you join the university as a new nontenured faculty member, the univer-sity will most likely give you a seed or start-up fund to start your research program, buy some lab equipment, and hire one or two graduate students as research assistants. You will be lucky to join as a team member of an ongoing research program with a funding stream. The research can also be supported by teaching assistantships (TAs) in the department. After a few years, a tenure committee in the department will evalu-ate your teaching, publications, research funding, and other activities and then decide your eligibility for tenure. As a tenured faculty, you will be promoted to associate professor. If not tenured, you can apply to another university or switch to an industry job. If everything moves smoothly after getting tenure, you will be a full professor in the course of time and remain in that position until retirement. Some professors switch to administra-tive positions for more money and pow-er, but with less intellectual attainment. Some universities create distinguished professors or prestigious chaired profes-sors, with endowment funds that pro-vide a high salary and initial tenure, to attract distinguished professionals. The

chairs are normally expected to have some industrial experience and high connections that can attract more fund-ing and enhance the universitys reputa-tion. There is no mandatory retirement age for a university professor. This is another advantage of a university career.

Awards and HonorsAs a researcher in power electronics, whether in a university or in industry, you normally become a Member of the IEEE. The IEEE is the largest internation-al professional organization in the world. Eventually, through your professional contributions, you can become eligible to be an IEEE Fellow [4]. The IEEE Fel-lowship is very prestigious, particularly in a university career. With this award, you become eligible for promotions and higher responsibilities in life. However, this award is highly competitive glob-ally, and the Fellow award is limited to only 0.1% of the total IEEE membership. Along with the Fellowship award, there is also opportunities for for IEEE main professional Society awards, technical field awards, and medals [5], which can be summarized as follows:

IEEE Power Electronics Society tRichard M. Bass Outstanding Young Power Electronics Engineer Award tR. David Middlebrook Achievement Award

IEEE Industrial Electronics Society t Dr.-Ing. Eugene Mittelmann Achieve-ment Award t Dr. Bimal Bose Energy Systems Award t Rudolf Chope R&D Award t J. David Irwin Early Career Award

IEEE Industry Applications Society tOutstanding Achievement Award tGerald Kliman Innovator Award t Andrew W. Smith Outstanding Young Member Award

IEEE Technical Field Awards t William E. Newell Power Electron-ics Award tRichard Harold Kaufman Award tNikola Tesla Award

IEEE Medals t Power Engineering Medal (re-placed the original Lamme Medal) tMedal of Honor.

The Medal of Honor is the highest award in the IEEE and is often defined

as the Nobel Prize in electrical engi-neering (as there is no Nobel Prize in engineering). In 2014, the Medal of Hon-or was awarded to power electronics scientist B.J. Baliga for the invention of the insulated-gate bipolar transis-tor (IGBT). Note that an Early Career or Young Member Award may be given before earning IEEE Fellowship.

Research Idea, Project Plan, and Proposal PreparationA research project in a university may be nonfunded or funded. In nonfunded research, a graduate student can get support from a TA, a professor may have his or her own research grant to sup-port the student, or a visiting research scholar from abroad may get support from his or her own government. Visit-ing research scholars are normally bril-liant because they are selected by their governments on a competitive basis to do research under reputed professors abroad. Experienced visiting professors and postdoctoral research fellows nor-mally give the best performance with lit-tle supervision, which helps to enhance the professors reputation.

In most cases, the research is funded and a professor has to write proposals to bring in funds. A proposal may be unso-licited or based on a request for propos-al (RFP), and the funding agencies may be government [such as the National Science Foundation (NSF) or the Depart-ment of Energy (DOE)] or private indus-tries. The research may be fundamental or application oriented, as mentioned before. The NSF, for example, promotes fundamental research, whereas the DOE research is generally applied. The industrial projects are normally small and solicited and require solving prob-lems related to products. Unfortunately, the success rate for government-funded proposals is very small, which causes a large waste of effort and a tremendous amount of frustration.

How do you generate ideas for re-search projects? A mature knowledge with a broad perspective of the technol-ogy helps with the generation of research ideas. Often, there is cross-fertilization of ideas, i.e., ideas gained in one technology area can be applied to others. A profes-sor can maintain an idea or knowledge

MARCH 2015 IEEE INDUSTRIAL ELECTRONICS MAGAZINE 11

book for this. After attending a confer-ence presentation or reading a paper, the question should arise in his or her mind, Have I learned anything new? The ideas can be jotted down in the notebook. Sometimes, research ideas flash in our thoughtful mind when we are relaxing in a chair, walking alone, etc. Solitude is the breeding ground for new ideas. The ideas should be innovative, advanced, and timely in the current technology trend. Once the idea crystalizes in the mind, it should be verified by an intensive litera-ture review to be sure that it is sound and has not been solved before. Figure 6 gives guidelines for planning a research project and proposal preparation. An appro-priate project title and objective of the research should be written clearly and convincingly. These and the technical background of the proposal with appro-priate references will demonstrate the im-portance of the project and the authors knowledge base to handle it. The other parts of the proposal, i.e., formulation and scheduling of the tasks, background of the project investigators (particularly the principal investigator), lab facilities, and budget, are also important.

Doctoral ResearchA full-time doctoral student in a univer-sity typically takes four years to com-plete his or her courses, qualifying and comprehensive exams, and the research work after the M.S. degree, whereas an M.S. degree may take two years for com-pletion. A doctoral project should be an original contribution in technology and should include the steps of idea formu-lation, a literature study, a system analy-sis, a computer simulation, and finally an experimental investigation to validate the results, as indicated in Figure 7. A university professor in a good university normally demands one or two transac-tions publications from the doctoral dis-sertation. The scope of an M.S. thesis is much narrower and may involve either a simulation or experimental study, finally resulting in a simple conference paper.

In the beginning, the advising profes-sor should define the research project for the student. It will be unfortunate if the professor is too busy and does not have a clear idea for the research and, therefore, depends on the student to

define his own project. The outline of the project idea with some description and scope of the work should be sum-marized in a one-page write-up and dis-cussed with the student extensively. If the student has any novel idea, it should be fully honored. Then, an extensive lit-erature study should follow to be sure that the idea is original and that no-body has done it before. The selection of a proper topic is the most important step in doctoral research. Following the start, the professor should have a week-ly meeting with the student for an ex-tensive discussion with back-and-forth learning and pursuing progress. Any gap in the students knowledge should be filled up by self-study or by the pro-fessor. Needless to say, a professor be-comes famous by the work of his or her graduate students. Therefore, his inti-mate association with the project until its successful completion is important.

In implementing the project, col-lect pieces of relevant knowledge from different sources; document them in a notebook; and connect, interpolate, and extrapolate them to generate the new knowledge. It is a good idea to formu-late a questionnaire related to the dif-ferent aspects of the project. Solving a project essentially consists of answering a questionnaire. The problems should be analyzed thoroughly from all angles to assure success. Suffice it to say that the student should be a true friend and companion of the professor throughout the project. The student should write a monthly progress report to the professor, which can be used as a periodic progress report for the funded project. Finally, the thesis is composed and approved by the doctoral committee after the defense presentation. A professors workload tends to be heavy with more graduate projects in addition his teaching load and

Formulate the project idea.

Search the literature and be sure the idea is novel and timely.

Analyze the idea and plan proposal preparation.

Generate the proposal title.

Write an abstract or objective of the project.

Write a technical background or prior art.

Formulate the tasks with a brief description.

Prepare a time schedule of tasks.

Identify the project investigators and write a brief curriculum vitae.

Review the laboratory facilities.

Formulate a budget and complete the proposal.

Submit the proposal.

FIGURE 6 Planning a project and proposal preparation.

12 IEEE INDUSTRIAL ELECTRONICS MAGAZINE MARCH 2015

other department activities. With a large number of students, he or she tends to be more of a project manager rather than a research adviser, and the quality of re-search can deteriorate, thus deteriorat-ing the publication standard.

Writing a Paper for PublicationHave you ever published any paper in IEEE publications, particularly in one of the transactions [7]? The transactions papers with your biography and photo at the end are prestigious and bring you status in the professional community. Published IEEE papers are included in IEEE Xplore (ieeexplore.org), and their importance can be found in citations and the h-index given in Google Scholar (scholar.google.com). The quality of the publication is extremely important. In-ferior-quality papers will give you a bad name. Typically, one transactions paper can be considered equivalent to four or five conference papers. Again, a high-quality paper with an original contribu-tion can be equivalent to ten mediocre papers, and a paper with an invention

(see Figure 1) may be worth 20 high-qual-ity papers. If you are a young, untenured professor, transactions publications will help you get tenure and promotions. If you are a tenured professor, more trans-actions publications will promote your fame, help you become an IEEE Fellow, and may subsequently bring other IEEE and non-IEEE awards and honors. Gradu-ally, the door of your career may open for new avenues of success. Needless to say, publications are extremely important in the academic community for survival following the axiom publish or perish. If you are an industrial researcher, pub-lication may not be that important, but it brings fame to your career. However, if you are an engineer in industry and try-ing to transition to a university career, you must build up a publication base. Above all, publications bring the tremen-dous satisfaction of career accomplish-ments. A scientist without publications is forgotten quickly.

If you have done research and the results are of archival value, these are publishable as papers. The material

may be of current interest or may have a potential interest in the future. Of course, state-of-the-art technology sur-vey papers by experienced authors are also worthy of transactions publication. Proceedings of the IEEE often publishes prestigious survey papers, which of-ten get best paper awards (such as the Donald G. Fink Award). The research results in archival literature may be im-portant for immediate applications or future applications after a prolonged period of time. For a new and emerg-ing technology, often analytical results with validation by a simulation study may suffice for a transactions paper. Otherwise, experimental results are de-manded to substantiate analytical and simulation results. The reason is that a simulation is only as good as the model, which means that if the modeling is not accurate, the simulation cannot give trustworthy results. Anyway, when a contribution has been made, you need to judge carefully if it is worthy of be-ing a transactions or conference paper. Once you have decided that the material is transactions worthy, the first step is to organize the material very carefully. In addition to having good technical content, writing a good paper is an art and tests your knowledge of English and your writing skills. It is no wonder that a majority of submitted papers, particu-larly those written by foreign nationals, are rejected. A typical flowchart for writ-ing a paper is shown in Figure 8 [7].

Writing a good paper is like telling a story to somebody, which should be clear, concise, and well organized with a logical flow of expressions. It is always a good idea to read some good papers written by reputed authors. The title of the paper should clearly reflect your contribution. For every paper, the lead student normally becomes the first au-thor even though the advisor has made the primary contribution. This helps the student successfully establish his or her career. Then, coauthors should be added in the order of magnitude of their contribution. It is unethical to add a coauthor who has not made any contribution to the paper. In the same way, it is not ethical to add the name of a department head, project manager, or financial supporter as a coauthor

What is your background?What project experience do you have?

Do you have any topic preference?

Select a project topic.

Formulate the project topic andwrite the implementation steps in one page.

Review the literature and iterate the topic.

Analyze the system and design.

Evaluate the system performance.

Develop the control strategy and analyze the system.

Simulate the system and evaluate the performance.

Integrate the system and perform laboratory tests.

Write the thesis and papers.

FIGURE 7 Steps in a doctoral research project.

MARCH 2015 IEEE INDUSTRIAL ELECTRONICS MAGAZINE 13

unless he or she has made contribu-tions. Note that plagiarism is an offense, and submitting multiple publications of the same material to different jour-nals with slight alterations is highly un-ethical. The next step is to collect the references in the proper format. The references are important for writing the introduction of the paper. The reviewer of the manuscript will become angry if his or her contribution is not cited in the paper. Planning figures with the appropriate labels and titles is a very crucial step in the preparation of the paper. The figures and captions should be fully explanatory and clearly convey their contribution to the paper. A figure is worth a thousand words. The figures should be finalized after the preparation of the full draft paper. Then, plan the dif-ferent sections and subsections with the appropriate title and assign the fig-ures to the sections. Organize the main equations with the appropriate symbols and definition of the symbols locally. The equations are the ornaments of a research paper, and if possible, there should be a few equations. Use the symbols that are commonly used in textbooks. The derivation of the equa-tions, if necessary, should be included briefly in the appendix. Organize all the points in detail and in proper sequence in each section and subsection before starting the draft paper preparation. Correct English composition, grammar, and spelling are extremely important in paper writing. Needless to say, despite being an excellent contribution, a ma-jority of papers are rejected because of poor English. If a professor depends solely on his graduate student for writ-ing, it is almost certain that the paper will be rejected. Again, even if the paper is written by an experienced professor himself, rejection is not uncommon.

The most difficult part of the paper is writing the introduction. Here, in the beginning, you should clearly highlight the importance of your contribution in a convincing way. Then, past contribu-tions in this area should be reviewed with proper references, emphasizing why your contribution is novel and su-perior to others. The remaining parts of the paper consist of a simple and clear description of the content in logical

sequence. Finally, summarize your con-tribution and its significance in the conclusion. After writing the full draft paper, revise it several times to improve and polish the English. After the peer re-view of your paper, typically by three re-viewers, it may be rejected or accepted with recommendations for revision.

If you are a senior and established professor with a lot of experience, you should possibly consider writing a book. A good book can give you a lot of visibility in the professional communi-ty. However, writing a book may not be easy. Many professors start the proj-ect, but very few complete it. A book project requires a lot of extra studies and continuous progress without any interruption. A young professor in the midst of career building should never undertake a book project.

Doing Research in IndustryThis article would be incomplete without some discussion of the pros and cons of industrial jobs. Some discussion was included in the Doing Research in Uni-versities section. A discussion of gov-ernment research labs is also included in this section. Some students, after a long period in the university environment, find it tiring and monotonous and would like to experience the outside world by taking a job in industry. There is definite-ly satisfaction in doing real- world, prod-uct-oriented practical projects with your own hands and interacting with a lot of people. As I mentioned before, some industrial experience is an asset if you plan to switch to a university career later as a prestigious chaired professor. In that case, building a publication base is essential. The salary in an industrial job may be higher, but not necessarily. There is also the possibility of moving to a higher management position (such as vice president) later in your career with high compensation and power that are unheard of in a university. However, a managerial job is generally less secure than that of a contributor. As a manager, you may have a reputation only within the perimeter of your company. Besides, unlike a renowned university professor, you are forgotten quickly as soon as you quit the management position. As a contributor, you will have to report the

progress of your project to your man-ager, who may be less mature and less educated than you. He or she may also have an arrogant personality that may be difficult for you to bear. If you meet him/her in the corridor, he or she may ask you, How is the project going? Your travels are restricted mostly to business travels. There is no tenure system in industry, and you may be laid off with short notice if the companys financial condition is not good or if it changes its research direction.

A research lab in a large corpora-tion may undertake large government-funded projects, or problem-solving tasks for their product departments, in addition to its own assessed-fund tech-nology development projects. Much of the discussion on academic research given in the previous section is also applicable in industrial projects. Since industries are profit oriented, the R&D activity is highly organized to econo-mize cost. The publication of patents is of much higher priority than the pub-lication of papers. In fact, paper publi-cation is often denied or delayed until

Is the contribution publishable?

Generate an appropriate paper title.

List the author/coauthors.

Collect the relevant references.

Plan the figures and tables with titles.

Collect the points and organizethe sections and subsections.

Write a draft paper.

Finalize all the figures and tables.

Prepare the final paper(revisereviserevise).

FIGURE 8 Steps for writing a research paper.

14 IEEE INDUSTRIAL ELECTRONICS MAGAZINE MARCH 2015

the application or issue of the patent. Efficient team work (with compatibility of temperament) for a large project un-der a project manager is important. A monthly progress report and periodic design review (with action items) on the progress of the project by an expert review team are essential. This is dis-cussed further in the My Experience in Research section.

My Experience in ResearchI started my power electronics research in 1958 in the University of Wisconsin, Madison, where I completed my M.S. degree in 1960 under the support of the United States Agency for International Development (USAID). My project was concerned with investigating three-phase diode-bridge rectifier harmonics on a long transmission line. In those days, thyristors were not available com-mercially (they were introduced in 1958), but silicon power diodes were available in high power for industrial applications.

The field of power electronics was then known as industrial electronics. The term power electronics was introduced in the beginning of the 1970s. My Wisconsin project consisted of some theoretical study and experimental investigation with a model distributed parameter (LC) transmission line in the lab. With the help of a harmonic wave analyzer, I could demonstrate the distribution of harmonics along the line and the reso-nance voltage boosting effect with a par-ticular harmonic.

I returned to India in 1960 and joined the Bengal Engineering College [now the Indian Institute of Engineering Sci-ence and Technology (IIEST)], Shibpur, to start my career and begin my doc-toral research in Ramey magnetic am-plifiers (MAs). My research project was somewhat hybrid among MAs, power transistors, and thyristors. The proj-ects were developmental, analytical, and experimental studies of a magnetic servoamplifier for a position-servo with

a two-phase induction motor and mul-tichannel telemetry encoding systems [8], [9]. The U.S. National Aeronau-tics and Space Administration (NASA) Langley Research Center, United States, considered the encoding system for satellite-to-earth data communication. I received my doctorate degree in 1966. From 1966 to 1971, I supervised several MA-based research projects on a four-quadrant analog multiplier, a dc-to-dc converter, a magnetic servoamplifier, and a digital voltmeter.

The year 1971 was historic in my ca-reer, when I emigrated to United States with an ambitious goal in life. I joined Rensselaer Polytechnic Institute (RPI), Troy, New York, as a faculty member to organize its power electronics teaching and research program with the help of GE Corporate R&D (GE-CRD) (now GE Global Research Center), Schenectady (Figure 9). Power electronics was an emerging technology at that time with intense R&D activities by large corporations, such as GE, Westinghouse, and Siemens. Coming from India with an MA background, the work was quite challenging to me. The only other university in the United States with a power electronics program at that time was the University of Missouri, Columbia. I found that the graduate stu-dents were extremely brilliant in RPI and were excited to do research projects in the emerging power electronics area. I completed a number of projects, which consisted of developing a transistor ac switch and its application in a matrix con-verter [19], Triac speed control of a three-phase induction motor, a three-phase ac power control with transistors, a thyris-tor self-oscillating inverter, etc. During this period, I came in close contact with GE-CRD researchers in power electron-ics. They offered me a part-time (one day/week) consulting research project on a thyristor HF resonant-link cyclocon-verter for a motor drive [24], which was quite challenging for me. The project was successful, and I could for the first time demonstrate that the cycloconverter could operate at a programmable (lead-ing-lagging) line displacement power fac-tor (DPF) [as a static VAR compensator (SVC) in extreme cases]. It was also the M.S. project of a GE-CRD engineer who helped me complete the project.

FIGURE 9 Doing research in the GE-CRD laboratory on DSP control of EV (ETXII) project. (From left) Paul Szczesny, Bimal Bose, and Hunt Sutherland [11].

MARCH 2015 IEEE INDUSTRIAL ELECTRONICS MAGAZINE 15

I decided to transition to GE-CRD [11] in 1976 after a 16-year university career. My main motivation was to learn power electronics with hands-on ex-perience and work on some real-world large industrial projects. In those days, GE-CRD was the worlds top research center (called the ivory tower) in power electronics. It was like Bell Laboratory where the transistor was invented. Pow-er electronics scientists from all over the world used to visit us in Schenectady. I could see so many world-class scientists across the hall in Building 37, where my office was located. My first project was an autosequential current-fed inverter analysis and simulation with William McMurray (see William McMurray: The Guru of Power Electronics). Bill was the founding father and guru of power electronics, and his papers on force-commutated thyristor inverters were classic contributions that set the stage for the modern power electron-ics evolution. I used to worship him like God. From him, I learned that research ideas do not necessarily come within the 8 a.m.5 p.m. work day in office. The thoughts linger most of the time beyond the office hours, and often new ideas come when I am taking bath, walk-ing alone in the evening, or even in the midnight when I suddenly wake up with the flash of a new idea. There is no dif-ference between scientific research and transcendental meditation [11]. A substantial part of my time in GE (until 1987) was involved with electric vehicle (EV) and hybrid electric vehicle (HEV) projects. The EV project was the first major initiative by the U.S. government after the Arab oil embargo in the 1970s. I was the principal engineer for micropro-cessor/DSP control development, which was quite a difficult subject for a power electronics engineer in those days. Our first EV project (ETV1) was a splendid success [21], and it was demonstrated before Queen Elizabeth II of England. My last GE project was an EV drive (ETXII) with an interior permanent magnet (IPM) synchronous motor drive [12] using distributed TMS320 DSP-based control. Gradually, the IPM synchronous motor for EV/HEV drives was accepted all over the world. My other important projects in GE were control development

of a linear inductor machine for railroad propulsion, a microcomputer-based hy-brid (SPWM-SHE) PWM controller [22] of an inverter, scalar decoupled control of induction motors, control of SRM drives, sliding mode control of induction motors, maximum power point tracking (MPPT) control of residential PV sys-tems [26], etc. I noticed that my manager strongly discouraged simulation stud-ies, which he thought were a waste of time. I do not trust simulation results was his comment. He would only trust experimental waveforms on a scope or a multichannel recorder when the invert-er is working and the machine is run-ning. The company mandate was that if I needed to study fundamentals related to a project, I must do it in my home on my own time. Company time was only for problem solving. Although I was very publication minded, publications were considered a waste of time in the com-pany environment. Instead, writing pat-ent applications was highly encouraged. Often, paper writing was not permitted at all or was permitted with a long de-lay after patent application. During my GE career of 11 years, I lost practically 50% of potential research publications because of this company policy. In this period, I continued as an adjunct fac-ulty member of RPI, where I advised a large number of graduate students and taught a graduate course on ac drives in the evening. I published my first edited book, Adjustable Speed AC Drive Systems (1982) [13], and my first authored book, Power Electronics and AC Drives (1986) [14], in my GE days with great hurdles.

I had to work hard on the weekends for these books with the door shut against my family members. The authored book was translated into several languages immediately after publication.

I decided to return to a university career in 1987 after spending 11 years in industry. The university is my origi-nal home, and I love this career. I joined the University of Tennessee, Knoxville, as the Condra Chair of Excellence (En-dowed Chair Professor) in Power Elec-tronics with initial tenure granted to me. Concurrently, I started working as a chief/distinguished scientist in the new-ly established Electric Power Research InstitutePower Electronics Applica-tions Center. Part of my responsibility as a chief scientist was to promote pow-er electronics education and research in the United States. In addition to my reg-ular graduate students, I was fortunate to get a large number of visiting profes-sors and research scholars from abroad to come and work in my laboratory with financial support from their respective governments. All of them were brilliant scholars. Unfortunately, I was not very lucky to get mega-funded projects from the U.S. government agencies. In fact, I hardly tried for it. I love the university career for its freedom and prestige but hate to be a super-salesman seeking re-search funds. In my opinion, the govern-ment should maintain a roster of expert researchers in the country and solicit their contributions instead of research-ers searching for government funds. Some of my research contributions in the University of Tennessee included a

WILLIAM MCMURRAY: THE GURU OF POWER ELECTRONICS

William McMurray (Figure S1) was a power electronics scientist with GE-CRD for 35 years (19531986). He received the honorary doctor of law degree from Concordia University, Canada, in 1986. He be-came an IEEE Fellow in 1980 and a Life Fellow in 1994. He received the IEEE Newell Award (1978), the IEEE Lamme Medal (1984), and the IEEE Millennium Medal (2000) for his research contributions. He authored the book The Theory and Design of Cycloconverters (MIT Press, 1972) and was a contributing author in the historic book Principles of Inverter Circuits by Bedford and Hoft (New York, Wiley, 1964). Bill was a chain smoker. In the later part of life, he suffered from emphysema, which was the cause of his death in 2006.

FIGURE S1 William McMurray (19262006) [10].

16 IEEE INDUSTRIAL ELECTRONICS MAGAZINE MARCH 2015

soft-switched inverter for motor drives, HF nonresonant link power conversion for EV drives using a MOS-controlled thyristor [25], fuzzy control of dc and induction motor drives, fuzzy control of wind generation systems, neural net-work-based drive feedback signal esti-mation for vector drives, neural control of space vector modulations (SVMs) of two-level and multilevel converters [28], [29], converter faults investiga-tion, automated IM drive control design by expert systems, sensorless vector control of IMs, and high-temperature superconductivitysynchronous mo-tor (SM) ship propulsion with multilevel converters. A number of these projects were government funded. We did a lot of pioneering work in the application of AI techniques in power electronics [1], [27]. Unfortunately, however, the area did not pick up the desired momentum possibly because of its general unfamil-iarity in the power electronics communi-ty and very few industrial applications. During my days in the University of Ten-nessee (19872014), I traveled abroad extensively to give tutorials, IEEE Distin-guished Lectures, invited seminars, and keynote addresses. During this time, I also published two authored books [1], [15] and three edited books [16][18].

Some of my key contributions can be summarized as follows [33]:

I invented the transistor ac switch [19] and demonstrated it for acac direct power conversion in 1973 (published in 1976). This has been recognized as a key milestone contri-bution in matrix converter research [20]. The IGBT-based ac switch is now

universally used in matrix converters. The matrix converter for acac con-version was formally introduced later by Venturini in 1980.

I pioneered microprocessor con-trol of power electronics systems [14], [16]. The first microproces-sor-controlled fully functional commercial motor drive system paper for EV applications was published in 1979 [21] after Intel 8080 was introduced in 1970. I also introduced microprocessor (Intel 8086)-based SPWM (hybrid with SHE) of fully functional voltage-fed inverter VFI for IM drives [22]. Mi-croprocessors and DSPs are now universally used in the control of power electronic systems.

I invented an adaptive hysteresis-band PWM current control method of voltage-fed inverters used for IPM synchronous motor drives in 1989 [23]. This method is now widely used for commercial direct torque control (DTC) drives and other applications.

I demonstrated the first thyristor cy-cloconverter-based ac-HFac-ac reso-nant link power conversion system for motor drives that could operate at a programmable ( leading-lagging) power factor at the line side in 1975 [24]. In extreme cases, it could oper-ate as an SVC.

I proposed an HF active filter in the dc link to eliminate electrolytic ca-pacitors in a voltage-fed converter system [32].

I introduced the HF nonresonant link soft-switched power conver-sion for ac drives [25].

I introduced the MPPT control al-gorithm in PV power systems in 1984 [26] (IEEE prize paper), which is routinely used today.

I pioneered AI (expert system, fuzzy logic, and neural network) applica-tions in the control and estimation of power electronic systems, which is now an emerging technology [1], [15], [27]. These include ANN-based space vector PWM for two- [1] and multilevel converters [28], [29].

I published the first textbook on modern power electronics and ac drives in the English language in 1986 [14], [33].

I built the power electronics pro-gram in the University of Tennes-see from zero-ground to the center stage of the world during 19872002. This provided the favorable base for building the present national center for the smart grid project [known as the Center for Ultra-Wide-Area Re-silient Electric Energy Transmission Networks (CURENT)] by the NSF and the DOE (http://curent.utk.edu).

I promoted power electronics glob-ally through extensive seminars, tutorials, books, invited presenta-tions, and keynote addresses [30].

ConclusionThe article has provided a comprehen-sive and personalized review on doing research in power electronics, which in-cludes my experience and contributions during my career that spans more than four decades, covering the entire period of the modern power electronics evolu-tion. Although my experience has been

&vee,##"".&ve--(v!" !# &ve #" &ve # "# .&

!## e.& "# # %"'( &!# ("""ev# .$eNev !&!#ve! # "#% '#

FIGURE 10 My ten-point instructions to young scientists.

MARCH 2015 IEEE INDUSTRIAL ELECTRONICS MAGAZINE 17

highlighted in the article, it should be em-phasized that the technology has been enriched by the relentless contributions of many researchers over a long period of time. The article is generally addressed to young researchers who are about to embark on a new career. Research in the university and industry is generally a vast topic and is covered here briefly, empha-sizing the salient points. The advantages and demerits of both the career paths are outlined. Doctoral research that includes the formulation of project ideas, proposal preparation, execution of the project, and writing papers for publication has been highlighted in the discussion. Hopefully, my research experience, including my own contributions, will be of interest and benefit to the readers. Finally, I would like to conclude the article with some advice [31] for young researchers (Figure 10). I have tried to follow these instructions in my own career.

AcknowledgmentsI am deeply indebted to my graduate students and visiting scholars who contributed so much for the develop-ment of my career. Most of the names are cited with the references at the end of this article. I am also grateful to Prof. T.H. Liu of the National Taiwan Univer-sity of Science and Technology for re-viewing the whole article and making some suggestions. The presentation of a similar topic was given in his univer-sity with his invitation [3].

BiographyBimal K. Bose ([email protected]) was a faculty member at the Bengal Engineer-ing College, India, [currently the Indian Institute of Engineering Science and Technology (IIEST)] from 1960 to 1971. From 1971 to 1976, he was an associate professor of electrical engineering at the Rensselaer Polytechnic Institute, Troy, New York. From 1976 to 1987, he was a re-search engineer in General Electric Cor-porate Research and Development (GE-CRD) (now GE Global Research Center), Schenectady, New York. He has been the Condra Chaired Professor (Chair of Excellence) at the University of Tennes-see, Knoxville, since 1987. He special-izes in power electronics, motor drives, and artificial intelligence applications.

He has authored more than 250 papers, holds 21 U.S. patents, and authored/edited seven books in power electron-ics. He is a recipient of a number of awards, including the IEEE Power Elec-tronics Society Newell Award (2005), the IEEE Millennium Medal (2000), the IEEE Meritorious Achievement Award in Continuing Education (1997), the IEEE Lamme Medal (1996), the IEEE In-dustrial Electronics Society (IES) Mittel-mann Award (for lifetime achievement in power electronics and motor drives) (1994), the IEEE Region 3 Outstanding Engineer Award (1994), the IEEE Indus-try Applications Society Outstanding Achievement Award (1993), the Calcutta University Mouat Gold Medal (1970), the IIEST Distinguished Alumnus Award (2006), the D.Sc. degree (honoris causa) from IIEST (2013), and a number of prize paper awards. IEEE Industrial Electron-ics Magazine published a special issue (June 2009) Honoring Dr. Bimal Bose and Celebrating His Contributions in Power Electronics with a cover photo. The IES Dr. Bimal Bose Energy Systems Award was established in 2014, which is funded by the IEEE Foundation and the IES. He is a Life Fellow of the IEEE.

References[1] B. K. Bose, Modern Power Electronics and AC

Drives. Upper Saddle River, NJ: Prentice Hall, 2001.[2] B. K. Bose, Power electronics and mo-

tor drivesRecent progress and perspec-tive, IEEE Trans. Ind. Electron., vol. 56, no. 7, pp. 581588, Feb. 2009.

[3] B. K. Bose, How to do research in power elec-tronics and motor drives, presented at the Seminar in National Taiwan University of Sci-ence and Technology, Apr. 16, 2008.

[4] B. K. Bose, How to be an IEEE Fellow, IEEE IE Society Newslett., vol. 52, pp. 67, Dec. 2005.

[5] B. K. Bose, IEEE medalsThe most covetable awards, IEEE PELS Newslett., vol. 20, 1st. Qr., pp. 1113, 2008.

[6] M. Liserre, L. G. Franquelo, I. Nagy, and C. Wen, Honoring Dr. Bimal Bose and celebrating his contributions in power electronics, IEEE Ind. Electron. Mag. (IEEE IES Special Issue), vol. 3, no. 1, pp. 25, 1214, June 2009.

[7] B. K. Bose, How to get a paper accepted in transactions, IEEE IE Soc. Newslett., vol. 53, pp. 67, Mar. 2006. (Also in IEEE PELS Newslett., Second Qr., vol. 18, no. 2, pp. 46, 2006).

[8] B. K. Bose, A magnetic amplifier telemetry en-coder circuit, Proc. IEEE, vol. 52, no. 9, p. 1076, 1964.

[9] B. K. Bose, An improved telemetry encoding circuit by square-loop cores and SCRs, Proc. IEEE, vol. 54, no. 3, pp. 440442, 1966.

[10] B. K. Bose, Power electronicsHistorical perspective and my experience, IEEE Ind. Ap-plicat. Mag., vol. 20, pp. 29, Mar./Apr. 2014.

[11] B. K. Bose, Eleven years in corporate environ-mentMy experience, IEEE IE Soc. Newslett., vol. 54, pp. 68, June 2006.

[12] B. K. Bose, A high performance inverter-fed drive system of an interior permanent magnet synchronous machine, IEEE Trans. Ind. Appli-cat., vol. 24, pp. 989997, Nov./Dec. 1988.

[13] B. K. Bose, Ed., Adjustable Speed AC Drive Sys-tems. Piscataway, NJ: IEEE Press, 1982.

[14] B. K. Bose, Power Electronics and AC Drives. Upper Saddle River, NJ: Prentice Hall, 1986.

[15] B. K. Bose, Power Electronics and Motor DrivesAdvances and Trends. Burlington, MA: Academic, 2006.

[16] B. K. Bose, Ed., Microcomputer Control of Power Electronics and Drives. Piscataway, NJ: IEEE Press, 1987.

[17] B. K. Bose, Ed., Modern Power Electronics. Pis-cataway, NJ: IEEE Press, 1992.

[18] B. K. Bose, Ed., Power Electronics and Variable Frequency Drives. Piscataway, NJ: IEEE Press, 1997.

[19] V. Jones and B. K. Bose, A frequency step-up cycloconverter using power transistors I inverse-series mode, Int. J. Electron., vol. 41, no. 6, pp. 573587, 1976.

[20] T. Friedli and J. W. Kolar, Milestones in matrix converter research, IEEE J. Ind. Applicat., vol. 1, no. 1, pp. 214, 2012.

[21] B. K. Bose, A microprocessor based control system for a near-term electric vehicle, IEEE Trans. Ind. Applicat., vol. 17, no. 6, pp. 626631, Nov./Dec. 1981 (Also in IEEE IAS Annu. Meeting Conf. Rec., 1979, pp. 743748).

[22] B. K. Bose and H. A. Sutherland, A high perfor-mance pulse-width modulator for an inverter-fed drive system using a microcomputer, IEEE Trans. Ind. Applicat., vol. 19, pp. 235243. Mar./Apr. 1983.

[23] B. K. Bose, An adaptive hysteresis band cur-rent control technique of a voltage-fed PWM inverter for machine drive system, IEEE Trans. Ind. Electron., vol. 37, pp. 402408, Oct. 1990.

[24] B. K. Bose and P. Espalage, High frequency link power conversion, in IEEE Trans. Ind. Ap-plicat., vol. 13, pp. 387393, Sept./Oct. 1977.

[25] L. Hui, B. Ozpineci, and B. K. Bose, A soft-switched high frequency non-resonant link integral pulse modulated dc-ac converter for ac motor drive, in IEEE IECON Conf. Rec., 1998, pp. 726732.

[26] B. K. Bose and P. Szczesny, A microcomputer based control of residential photovoltaic pow-er conditioning system, IEEE Trans. Ind. Appli-cat., vol. 21, pp. 11821191, Sept./Oct. 1985 (also presented at the IAS Conf., 1984).

[27] B. K. Bose, Neural network applications in power electronics and motor drivesAn introduction and perspective, IEEE Trans. Ind. Electron., vol. 54, no. 1, pp. 1433, Feb. 2007.

[28] S. K. Mondal, B. K. Bose, V. Oleschuk, and J. O. P. Pinto, A neural network based space vector PWM controller for a three-level volt-age-fed inverter induction motor drive, IEEE Trans. Ind. Applicat., vol. 38, pp. 660669, May/June 2002.

[29] N. P. Filho, J. O. P. Pinto, B. K. Bose, and L. E. B. da Silva, A neural network based space vec-tor PWM of a five-level voltage-fed inverter, in IEEE Industry Applications Society Conf. Rec., pp. 21812187, 2004.

[30] M. Liserre, Dr. Bimal K. Bose: A reference for generations, IEEE Ind. Electron. Mag., vol. 3, no. 2, pp. 25, 2009.

[31] B. K. Bose, Fulfilling my lifelong dream, IEEE Ind. Applicat. Mag., vol. 19, p. 88, Sept./Oct. 2013.

[32] B. K. Bose and D. Kastha, Electrolytic capacitor elimination in power electronic system by high frequency active filters, in IEEE IAS Annu. Meet-ing Conf. Rec., 1991, pp. 869878.

[33] Wikipedia. (2015.). Wikipedia online encyclo-pedia. [Online]. http://en.wikipedia.org/wiki/Bimal_ Kumar_Bose