Autonomous Mobile Platform

8/3/2019 Autonomous Mobile Platform

1/5

Autonomous Mobile Platform II

Krzysztof Jaskot, Artur Babiarz

Institute of Automatic Control

Silesian University of Technology

Gliwice, Polande-mail: [email protected], [email protected]

Abstract The paper presents design of autonomous mobile

platform based on the all terrain 1/8th scale four wheel drive

radio control model. In this paper was considered problem of

automatic control of mobile platform using information from

GPS system, electronic compass and encoder. The mobile

platform is equipped in two-stroke glow engine, heavy-duty

drive train and wide-track suspension and controller based on

ARM7 microcontroller and using MaxStream XBee Pro

2.4GHz radio modem communication module. The base station

equipment is also described. Communication protocol betweenmobile platform and base station is presented. The paper

presents also an application of electronic compass to measure

azimuth of mobile platform. Problem of speed and distance

control is described. Autonomous mobile platform is a machine

that can operate in a human-made environment. By control in

this case we will understand to be able to avoid collisions with

obstacles (other mobile platforms and walls) during drive.

Results of real application are also shown. Results of work on

autonomous mobile platform that can operate in human

environment are presented. The obtained properties of the

system have been effected that it can be used for future

research and autopilot design project.

Keywords: mobile platform, sensors, communication,

microcontroller

I. INTRODUCTION

The aim of the project was to create an autonomous

mobile platform, which could operate in open terrain.

Autonomous mobile platform is a machine that can operate

in a human-made environment [4,5,6,7]. The key to

autonomy is a control system built on the basis of

information concerning the position and goal.Use GPS NMEA (National Marine Electronic

Association) protocol allows you to obtain information intext form about the current location of the object. Afteradding information about the intermediate target points(Waypoint), we can receive information about the currentdirection [8]. After the experiments conducted and described

in the article [2] been amended accordingly to increase theaccuracy of determining the direction of movement andspeed control.

The work was considered problem use informationderived from the GPS and IMU (Inertial MeasurementSystem) [3] as a source of control signal. In addition, weuses also information from the speed sensor. IMU currently

provides only information from the electronic compass. Thisinformation is needed to determine the azimuth of mobileplatform.

II. MOBILE PLATFORM

As described in the article [2] mobile platform has been

built using a remote-controlled car on a scale 1/8th (length

55cm, width 43cm) and it was delivered by the HPI Racing.

Selected terrain model with an independent suspension and

four-wheel drive (4WD), because we wanted to create an

autonomous platform that can operate in open terrain. The

appearance of old version of mobile platform with installed

controller and the GPS system is shown in figure 1.

Figure 1. Old version of mobile platform.

As the propulsion system used in this model, two-strokeinternal combustion engine with a capacity of 3.5cm3 andpower 2HP. This allows the dispersal model to speed about60km/h. In addition to the chassis and drive train in thecomposition of the platform includes two servos, which areresponsible for controlling the throttle/brake and course.

These two servos give us the ability to control the platformtraction. After the experiments described in the article [2] wechange the appearance of mobile platform. We addaluminum frame with installed GPS and 2.4GHz antenna.Aluminum frame prevents any interference generated byworking servos, engine (gearbox). We also add new

167978-1-61284-361-2/11/$26.00 c2011 IEEE


2/5

controller and IMU system. The appearance of new versionof mobile platform is shown in figure 2.

Figure 2. Mobile platform.

Where: 1- aluminum frame, 2- GPS, 3- antenna 2.4Ghz, 4-main controller, 5- IMU, 6- batteries.

III. CONTROL SYSTEM

Figure 3 shows a block diagram of the main controller.

The heart of the autonomous vehicle is mounted in the car-

driver will be built using a microcontroller AT91SAM7S,

managing servo, all sensors, power supply and wireless

communication.

To build a control system for mobile platform we usesmicrocontroller AT91SAM7S256 (ARM7 Core) it was

delivered by the ATMEL. He is responsible for collecting

information from sensors, processing them, the exercise of

control algorithms and generate servo control signals. In

addition, it enables wireless communication, reading values

of the individual, shared variables and modify them. Basic

features of the microcontroller is: RISC architecture,

maximum speed clock 60 MHz 256 KB FLASH program

memory, 64KB SRAM memory, interrupt controller, three

16-bits timers, four 16-bits PWM modules, three USART

interface, USB controller, I2C bus, two SSP, SPI bus, 11-channels 10-bits A/D converter [9].

The microcontroller of our choice has also some otherinteresting features, such as on board USB controller that

together with SAM-BA Boot Assistant provides an easy and

fast way of programming the ARM. We have also the JTAG

interface, which provides hardware debugging capabilities.

The main functions of the controller are: collecting

information from sensors, generating servo control signals

through the implementation of control algorithms based on

readings from sensors, providing two-way wireless

communication with a PC (telemetry, simulation), I2C

management. The process of autonomous control based on

sensor readings of the following:

GPS - provides information on the location of the car on the

globe, can also be a source of information about vehicle

speed and orientation relative to the direction north.

Electronic compass (IMU) - allows you to specify the

orientation of the car towards the north.Rotation sensor - provides information about the angular

velocity of the drive shaft speed of the car and indirectly,

used as a source of feedback for the speed.

Other sensors, such as: temperature sensors [1], ultrasonic

rangefinders, may be connected to an external derived I2C

bus driver.

Figure 3. Block diagram of the controller.

Main controller using the AT91 microcontroller is shown

in figure 4. The controller is equipped with a plate operator

panel with LCD, LEDs and push buttons for menu operation

serving. With them you can read some of the performance

of the program, namely: as GPS data, compass reading.

During normal operation, the buttons are not available

because the case is sealed. Then the display shows the next

cycle menus that contain the most relevant data. Sample

screens are shown in figure 5.

The aim of the work is to create an autonomous vehicle,but at the prototype stage and testing is necessary to reduce

its autonomy and ensure operator efficiency even take

control of the car, which for various reasons, may start to

behave unpredictably. Application of five-channel FM RC

controller allowed the design of switch mode AMS

(Autonomous/Manual Switch).

168 2011 12th International Carpathian Control Conference (ICCC)


3/5

Figure 4. Main controller with user interface.

Two channels, as in the case of a simple remote control, areused to transmit information about the location and direction

of the throttle servo. A third channel, acts as a switch to

allow the change of source signals servomechanisms

between the apparatus of the FM (manual) and

microcontroller (autonomous). In addition, upon failure of

the FM signal from the apparatus or GPS, followed by AMSemergency stop of the vehicle. AMS switch is built using 8-

bit microcontroller from Atmel - ATMega8. It is also

connected to the I2C bus of main controller.

Figure 5. User interface.

The controller consists of two mounted in a housing

designed for the project PCB and LCD. In total, this creates

a three-layer structure. Connection with external devices is

via connectors placed on the body. Furthermore, inside thecontroller can be mounted battery with dimensions

65x40x10mm. Visualization of the main controller is shown

in figure 6.

Figure 6. Visualisation of main controller.

In addition to the equipment fitted on mobile platform

available to the operator is base station, which includes:

remote control transmitter manual control andauto/manual switch, radio communication module XBee

PRO 2.4GHz sending/receiving data to/from the vehicle,

computer witch control application providing remote

viewing operating parameters of the control system mounted

on a platform. Figure 7 shows a system using wireless

communication modules, XBee PRO. These modules

operate in transparent mode. This means that any data sent

to the first module will be unchanged at the output of the

second module. Communication with the modules is done

using a UART protocol, which is easy to implement in the

microcontroller AT91SAM7S. From the PC side it is

necessary to convert the signals. After the application ofFT232RL FTDI and appropriate drivers on your computer,

you can create a virtual serial port.

Figure 7. Full duplex communication between PC and controller.

Communication between the controller and the computer

is realized in the form of data frames. An example of a data

frame containing the request and response are shown in

figure 8.Request:

Response:

Figure 8. Example of a data frames.

This frame allows you to read off the value ofn variables. In

the body frame includes a request ID variable in

hexadecimal, separated by commas. The answer of

controller is frame containing a comma-separated variables.

ARM7TDMI processor architecture allows you to install

an RTOS (Real Time Operating System). The operating

system allows the division of tasks performed by the

microcontroller processes. All tasks are performed

simultaneously which is referred to as multitasking. The

operating system also provides mechanisms for

communication between processes and synchronization

tasks. The core of the system takes over responsibility for

the allocation of CPU time for the process, taking intoaccount the priorities of processes and supporting a

notification of interruption.

The application of controller operates under the

FreeRTOS system [10]. It is optimized for embedded

systems with low hardware resources consists of only three

source files written in C. FreeRTOS also makes dynamic

2011 12th International Carpathian Control Conference (ICCC) 169


4/5

memory management, create lists and introduces features to

process execution pauses for a specified period of time.

Figure 9 shows the diagram of mechanisms for exchange ofdata between the key elements of the microcontroller

software to which they are: processes - constantly running in

a loop programs, drivers - used, among others by processes

to communicate with microcontroller peripherals.

Autonomous vehicle program consists of 6 processes

which have been divided between tasks. The process of

"GPS" is responsible for operating the GPS receiver. When

you receive the correct line of NMEA 0183 (every 5ms) is

processed, and read data update the corresponding variables.

The process of "IMU" collects data from the IMU device

driver via UART0, every 20 ms is sent to query the value of

the magnetic compass reading. The process of "interface" isused to refresh the driver of the operating panel is currently

displayed on the LCD. The process of "LED" acts as a

signal generator, reporting via LEDs on the correct

behavior. The process of "control" uses data to assess the

state in which currently there is a car (including position,

speed, azimuth), and then using the implemented algorithms

update a servo settings. The process of "communication" is

the basis of telemetric system of the car.

Figure 9. Division of microcontroller programs and threads used forperipherals and their drivers.

IV. R ESULTS

Test drives using the built controller were implementedin several stages. The first was the verification of the systemof GPS and radio communications transmissionmeasurements to the base station. At this stage, was carriedout manual control of direction and throttle. The results ofthis test are shown in figure 10. Test results for the manualcontrol shows a high accuracy to obtain information from

GPS.The second test phase was to incorporate automatic

direction and throttle control. Information on the target

points WP1WP7 was recorded in the memory of themicrocontroller. In an exemplary control algorithm usingtwo independent controllers. First, the block diagram shownin Figure 11, controls the servo direction. It seeks to

minimize the difference between the azimuth obtained in thecalculations in the perception stage, and a given azimuth.

Figure 10. Results of test GPS system.

The latter is the direction from which the car should move toreach the waypoint in a straight line. This system is a PIDregulator, extended with additional capabilities.

Figure 11. Block diagram of the azimuth control system.

The second controller shown in Figure 12 controls the speed

servo uses a PI regulator. Preset speed depends on the

quality of data obtained in the stage of perception. It is

estimated based on the number of satellites used by the GPS

receiver to measure the position.

Figure 12. Block diagram of speed control.

If their number is less than 4, the quality is considered

satisfactory and speed reference is 0. Otherwise, it is

different from zero and can take two values. The value of

fast (1m/s) applies only when the car is already headed

toward the via point and not located closer than 2m from it.If the distance is less than 2m predetermined value of speed

is 0.25m /s. This helps the controller to turn the direction of

the correct execution.

Example of autonomous waypoints navigation (Fig. 13)

shows positions and orientation taken from GPS and

170 2011 12th International Carpathian Control Conference (ICCC)


5/5

electronic compass. During drive we also recorded

measurements such as velocity (GPS, encoder), azimuth

(GPS, electronic compass), servos (Figure 14).

Figure 13. Example of waypoints navigation.

V. SUMMARY

The proposed solution allows to control an autonomous

mobile platform using a GPS and electronic compass.

Further development of the mobile platform will beimplemented will support the addition of sensors (eg laser

scanner, ultrasonic range finder), allowing to overcome

obstacles and to develop new control algorithms. After

applying the sensor system may need to change the glow

engine in the electric motor with reverse gear.

Figure 14. Examplary measurments recorded during autonomous drive(Azimuth GPS, electronic compass; Speed GPS, encoder; Servos).

ACKNOWLEDGMENT

This work has been supported by Ministry of Science andHigher Education In the years 2010 2012 as developmentproject OR000132 12.

REFERENCES[1.] Babiarz A., Jaskot K.: Temperature control system for

glow engine, Zeszyty Naukowe Politechniki lskiej,seria Automatyka z.150 , s. 115-121, Gliwice 2008.

[2.] Babiarz A., Jaskot K.: Autonomous Mobile Platform,International Carpathian Control Conference ICCC

2010, s. 179-183, Eger, Hungary.[3.] Babiarz A., Jaskot K.: The inertial measurement unit for

detection of position. Electrical Review, ISSN 0033-2097, R. 86 NR 11a/2010, pp. 323-334.

[4.] Baker C. R., Dolan J. M.: Street Smarts for Boss -Behavioral Subsystem Engineering for the UrbanChallenge, Robotics & Automation Magazine, IEEE,Carnegie Mellon Univ., Pittsburgh, PA, March 2009

[5.] Behringer R., Maurer M.:Results on Visual RoadRecognition for Road Vehicle Guidance, Proceedings ofthe 1996 IEEE Intelligent Vehicles Symposium, Tokyo,Japan 1996

[6.] Braun T., H. Schfer, K. Berns: Topological Large-ScaleOff-road Navigation and Exploration RAVON at the

European Land Robot Trial 2008, The 2009 IEEE/RSJInternational Conference on Intelligent Robots and

Systems, St. Louis, USA, October 2009[7.] Harkins R. et al.,Design and testing of an autonomous

highly mobile robot in beach environment, Proceedingsof the World Congress on Engineering and Computer

Science 2008 WCECS 2008, October 22 - 24, 2008, SanFrancisco, USA

[8.] Jaskot K.:Implementation of GPS information to control

of UAV model, ZN Politechniki Rzeszowskiej,Mechanika z.71 Awionika, Rzeszw 2007, ISBN 0209-2689.

[9.] AT91 ARM Thumb-based Microcontrollers, ATMEL[10.] FreeRTOS ARM7, www.freertos.org

2011 12th International Carpathian Control Conference (ICCC) 171

Autonomous Mobile Platform

Documents