The nearest perspectives of applications of artificial neural networks for diagnostics, modeling, and control in science and industry (methods, results,

The nearest perspectives of applications of The nearest perspectives of applications of artificial neural networks for diagnostics, artificial neural networks for diagnostics,

modeling, and control in science and industry modeling, and control in science and industry (methods, results, demonstrations & hands-(methods, results, demonstrations & hands-

on)on)

Victor S. Abrukov(2008)

e-mail: [email protected]

Chuvash State University, RussiaChuvash State University, Russia

Department of Thermo PhysicsDepartment of Thermo Physics

OutlineOutline

1. IntroductionEx-1: Deflagration to detonation transitionEx-2: Automatic control system of steam generating

unitEx-3: Wave propagation on a free surface of a fluidEx-4: Burning Rate vs Temperature profile 2. Perspectives of others (laser scattering technique

for remote sensing, , smart sensors for monitoring of industrial equipment (smart watch, soft sensors), etc.

3. Conclusions

Introduction - 1

Commercial applications of Data Mining (ANN, etc) in areas such as e-commerce, market-basket analysis, text-mining, and web-mining have taken on a central focus in last years.

There is a significant amount of innovative Data Mining work taking place in the context of scientific and engineering applications but that is not well-represented in the mainstream of Data Mining.

Introduction - 2

Scientific Data Mining techniques are Scientific Data Mining techniques are being applied to diverse fields such as being applied to diverse fields such as physics, chemistry, astronomy, physics, chemistry, astronomy, engineering, etc. engineering, etc.

Data Mining frequently enhances existing Data Mining frequently enhances existing analysis methods based on statistics and analysis methods based on statistics and exploratory data analysis. exploratory data analysis.

The goalThe goal

The goal of the seminar is: The goal of the seminar is: Share experiences Share experiences Learn how ANN can be applied to Learn how ANN can be applied to

their data their data Practice in it (You will the first user Practice in it (You will the first user

of Data Fusor in all of the world!)of Data Fusor in all of the world!)

Introduction (main!)

No Science without Data Base

and

Data Analysis

Data Base

There are three ways for creation Data Base: Real experiment Numeral experiment Analytical Solution

We have obtained DB. What do we have to do further?

Data Analysis

We have to create a model of DB that help We have to create a model of DB that help us to use DB (for understanding, us to use DB (for understanding, diagnostics, testing, control, … pleasures).diagnostics, testing, control, … pleasures).

What is the best way?What is the best way?

I think – Artificial Intelligence.I think – Artificial Intelligence.

Data Fusor is a part of AI toolsData Fusor is a part of AI tools

ANN is a part of Data FusorANN is a part of Data Fusor

EX-1: Modeling of Deflagration-to-EX-1: Modeling of Deflagration-to-Detonation Transition for Pulse Detonation Transition for Pulse

Detonation Engine.Detonation Engine.

Ex-1

A deflagration-to-detonation transition A deflagration-to-detonation transition under various experiment conditions was under various experiment conditions was studied by Santora R.J. at al (USA). studied by Santora R.J. at al (USA).

The data that were presented by it had The data that were presented by it had many blanks (about 60%). A task of filling many blanks (about 60%). A task of filling them by means of ANN was set. them by means of ANN was set.

The results obtained are presented in The results obtained are presented in Table. Table.

Ex-1

Ex-2: ANN ANN Model of Model of AAutomatic utomatic CControl ontrol SSystem of ystem of BBoiler oiler UUnitnit

The work has been aimed on a development of work has been aimed on a development of principles of creation of a new model of principles of creation of a new model of automatic control system of a boiler unit in automatic control system of a boiler unit in transient (non-linear) regimes (change of load, transient (non-linear) regimes (change of load, launch and stop of the boiler aggregate, etc) on a launch and stop of the boiler aggregate, etc) on a basis of ANN. basis of ANN.

The following problems were set The following problems were set and solvedand solved -1 -1

- - A complete set of non-linear differential A complete set of non-linear differential equations for the two-phase flow in cylindrical equations for the two-phase flow in cylindrical coordinate system circumscribing processes in a coordinate system circumscribing processes in a superheater was formulated. It described also a superheater was formulated. It described also a process of injection of water into a superheater process of injection of water into a superheater during control procedure of vapor temperature in during control procedure of vapor temperature in a superheater. a superheater.

- The obtained set of equations was converted into - The obtained set of equations was converted into a system of finite-difference equations. a system of finite-difference equations.

EquationsEquations

EquationsEquations

The following problems were The following problems were set and solvedset and solved - 2 - 2

Adjustable and controlling parameters were Adjustable and controlling parameters were selected by means of experimental data and selected by means of experimental data and experts experience. experts experience.

The rate of a change of vapor temperature on an The rate of a change of vapor temperature on an exit of the boiler aggregate was selected as a exit of the boiler aggregate was selected as a controlled parameter. The water mass flow via controlled parameter. The water mass flow via injector was selected as controlling parameter. injector was selected as controlling parameter.

The analytical connection between them was The analytical connection between them was determined by means of the mathematical determined by means of the mathematical apparatus of Lee derivatives. apparatus of Lee derivatives.

The analytical connectionsThe analytical connections

The following problems were set The following problems were set and solvedand solved - 3 - 3

The database for training of ANN was The database for training of ANN was created by means of the obtained created by means of the obtained analytical connection and an optimal analytical connection and an optimal architecture of ANN was determined. architecture of ANN was determined.

The training of ANN was executed and a The training of ANN was executed and a model of a control system as well as a model of a control system as well as a skeleton diagram of a position of a gate skeleton diagram of a position of a gate valve governing a water discharge on valve governing a water discharge on injection were created.injection were created.

Data Base for trainingData Base for training

The following problems were set The following problems were set and solvedand solved - 4 - 4

The prototype of an automatic control system of temperature vapor in a superheater of a boiler unit TGME-464 was designed.

It consists of the program emulator of ANN established on the personal computer, industrial microcontroller Philips P89LPC935, cable lines and switch gears.  

The obtained prototype can work both in a mode of advice, and in a mode of automatic control. It can be simply enough integrated in present control systems.

The features of ANN technologies of control (property of adaptability to new conditions and ability to self-training) provide simplicity of modernization and escalating of capabilities.

kk

kk TT

1

1

kk

kk TT

1

1

Scheme of adaptive controlScheme of adaptive control

Ex – 3: Wave propagation on Ex – 3: Wave propagation on a free surface of a fluida free surface of a fluidThe tasks of hard shock about a layer of fluid and The tasks of hard shock about a layer of fluid and

of wave propagation on a free surface of a of wave propagation on a free surface of a fluid were the problems (afluid were the problems (at an impulse t an impulse formulation, the deformation of the bottom of formulation, the deformation of the bottom of an earthquake occurs for a rather short time, an earthquake occurs for a rather short time, i.e., there is a bottom shock about a fluidi.e., there is a bottom shock about a fluid))

With the help of an available analytical solution With the help of an available analytical solution the data base consisting of values of the data base consisting of values of dimensionless coordinates, times and the dimensionless coordinates, times and the velocity on a free surface of a fluid were velocity on a free surface of a fluid were obtained.obtained.

Ex – 3: A part of Data BaseEx – 3: A part of Data Base

The values of The values of coordinates and coordinates and times were times were considered the input considered the input characteristics, and characteristics, and the values of the values of velocities the output.velocities the output.

t x U

3,5 20,00005 0,00011

3,5 19,50006 0,00203

3,5 19,00008 0,00609

3,5 18,50012 0,01119

3,5 18,00017 0,01872

3,5 17,50027 0,03013

3,5 17,00041 0,04731

3,5 16,50065 0,07301

3,5 16,00101 0,1107

3,5 15,50157 0,16531

3,5 15,0024 0,24282

3,5 14,50364 0,3506

3,5 14,00546 0,49698

3,5 13,50806 0,69042

3,5 13,01172 0,93814

3,5 12,51675 1,24358

3,5 12,02349 1,60347

3,5 11,53227 2,00405

3,5 11,04333 2,4185

3,5 10,55672 2,80624

3,5 10,07223 3,11672

3,5 9,58932 3,29733

3,5 9,10715 3,30458

3,5 8,62465 3,11544

3,5 8,14067 2,7343

3,5 7,6542 2,19415

3,5 7,16448 1,55162

3,5 6,67113 0,87784

3,5 6,17419 0,24754

3,5 5,67409 0,27129

3,5 5,1716 0,6273

3,5 4,66773 0,79289

3,5 4,16364 0,77083

3,5 3,66036 0,60398

3,5 3,15855 0,359

3,5 2,65865 0,08995

3,5 2,16068 0,11851

3,5 1,66273 0,18944

3,5 1,1622 0,20892

3,5 0,65864 0,18826

3,5 0,1529 0,0484

4 20,00025 0,00293

4 19,50028 0,01218

4 19,00038 0,03098

4 18,50056 0,05268

4 18,00083 0,08221

Ex – 3: Scheme of ANNEx – 3: Scheme of ANN

Ex – 3: resultsEx – 3: results

The results of ANN calculations of the velocity values as related to different coordinates and times together with analytical results are presented in Figures.

Ex – 3: results (t=2 sec)Ex – 3: results (t=2 sec)(coordinate y is U – velocity)(coordinate y is U – velocity)

-2

-1

0

1

2

3

4

5

0 5 10 15 20 25

coordinate x

co

ord

inate

y

Neuro

Num

Ex – 3: results (t=3,5 sec)Ex – 3: results (t=3,5 sec)

-1

-0,5

0

0,5

1

1,5

2

2,5

3

3,5

4

0 5 10 15 20

coordinate x

coo

rdin

ate

y

.

NumNeuro

Ex – 3: results (t=4 sec)Ex – 3: results (t=4 sec)

-1

-0,5

0

0,5

1

1,5

2

2,5

3

3,5

4

0 5 10 15 20

coordinate x

coor

din

ate

y

Neuro

Num

Ex – 4: Burning Rate vs Temperature Profile

The “black box” computational model of a flame temperature profile prediction is created.

It allows to predict the temperature profiles by means of data about heat of combustion, burning rate, and pressure

WITHOUT NEW MEASUREMENT EVERY TIMEEVERY TIME

EX - 4EX - 4

Ex - 4Ex - 4

The scheme of construction of an ANN PCW The scheme of construction of an ANN PCW model was as follows. model was as follows.

We have taken experimental data (from We have taken experimental data (from scientific and applied literature) about scientific and applied literature) about

- pressure of burning, P, pressure of burning, P, - burning rate, U, burning rate, U, - heat of propellant burning, Q, heat of propellant burning, Q, - as well as temperature profile (values of as well as temperature profile (values of

temperature, T and coordinates, x) for some temperature, T and coordinates, x) for some propellantspropellants

Ex - 4Ex - 4

Ex – 4Ex – 4

The different sets of values of pressure, The different sets of values of pressure, burning rate, heat of combustion and burning rate, heat of combustion and coordinates were installed on the input coordinates were installed on the input layer of ANN. The corresponding values layer of ANN. The corresponding values of temperature were installed on the of temperature were installed on the output layer of ANN. output layer of ANN.

Ex - 4Ex - 4

By means of a training tool named the method of “back By means of a training tool named the method of “back propagation of errors”1, we have created the ANN PCW propagation of errors”1, we have created the ANN PCW model. model.

This model is a “black box” type which consists of latent This model is a “black box” type which consists of latent connections between variables. connections between variables.

The “black box” obtained we have used in an invented The “black box” obtained we have used in an invented experimental investigation for the prediction of a temperature experimental investigation for the prediction of a temperature profile as follows. One of the experimental data that have not profile as follows. One of the experimental data that have not been used for training of the ANN we have used for checking been used for training of the ANN we have used for checking of the “black box” obtained. The values of pressure, burning of the “black box” obtained. The values of pressure, burning rate, heat of burning and coordinates were installed on the rate, heat of burning and coordinates were installed on the input layer of the “black box”. Then, the correspondence input layer of the “black box”. Then, the correspondence values of temperature T’, were obtained on the exit of the values of temperature T’, were obtained on the exit of the output layer of the “black box”. output layer of the “black box”.

The results we have obtained are presented in Table 2.The results we have obtained are presented in Table 2.

Ex - 4Ex - 4

Ex - 4Ex - 4 An analysis of Table depicts that the “black box” An analysis of Table depicts that the “black box”

enough correctly “determines” the temperature enough correctly “determines” the temperature profile by means of data about burning rate, pressure profile by means of data about burning rate, pressure and heat of burning (or maximum temperature of and heat of burning (or maximum temperature of flame).flame).

In our opinion, this way of determination the In our opinion, this way of determination the temperature profiles could be considered as temperature profiles could be considered as perspective unique tool for cases when we have no perspective unique tool for cases when we have no any experimental results for concrete propellants and any experimental results for concrete propellants and would like to have them. would like to have them.

In order to make it all we need to do is to collect a In order to make it all we need to do is to collect a lot of experimental data published in scientific and lot of experimental data published in scientific and technical literature concerning temperature profiles technical literature concerning temperature profiles for various propellants upon various condition. for various propellants upon various condition.

Then we could create the common “black box” Then we could create the common “black box” model: Burning Rate vs Temperature Profile.model: Burning Rate vs Temperature Profile.

PerspectivesPerspectives

ConclusionsConclusions

Practice is the best of all Practice is the best of all instructors instructors