Monitoring System for Testing the Performance of an Electric Vehicle Using Ultracapacitors

Transcription

1 Monitoring System for Testing the Performance of an Electric Vehicle Using Ultracapacitors Abstract Juan W. Dixon, Micah Ortúzar and Jorge Moreno A monitoring system for an Electric Vehicle, which uses an ultracapacitor bank, has been implemented. The purpose of this system is to test the performance of the vehicle under different road conditions (city road and highway), with and without the ultracapacitors connected. One of the most important objectives of the work is to test the efficiency gain of the vehicle, and the current flow coming from the batteries in both cases, to allow the evaluation of the battery life change. These results will lead to conclusions about the overall economical and technical advantage of the use of ultracapacitors in combination with batteries in electric vehicles. The efficiency gain is being monitored at the ultracapacitor level, at the battery level, and at the vehicle level. Related with the design of the ultracapacitor system, this has been optimized in weight and size, by using a watercooled IGBT power converter, and an aluminum inductor with air core. The ratings of the ultracapacitor bank are: nominal voltage: 300 Vdc; nominal current: 200 Adc; capacitance: 20 Farads. The amount of energy stored allows having 40 kw of power during 20 seconds, which is enough to accelerate the vehicle with minimum help of the traction batteries. The vehicle uses a brushless dc motor with a nominal power of 32 kw, and a peak power of 53 kw. The control system, based on a DSP, measures and stores the following parameters: battery voltage, battery state-of-charge, car speed, instantaneous currents in both terminals (battery and ultracapacitor), and actual voltage of the ultracapacitor. Once the tests on the streets are finished, data will be analyzed and compared with previous results, with and without the ultracapacitors connected. The increase in range with ultracapacitors has been estimated in 16% in city tests, but the difference looks almost negligible in highway tests. However, tests are still under way. The car used for this experiment is a Chevrolet LUV truck, similar in shape and size to a Chevrolet S-10, which was converted to an electric vehicle at the Catholic University of Chile. Copyright 2002 EVS19 Keywords: ultra capacitor, control system, data acquisition, efficiency, state of charge, regenerative braking. 1. Introduction Ultracapacitors are a new technology that allows storing 20 times more energy than conventional electrolytic capacitors [1,2]. Despite this important advance in energy storage, they are still far from being compared with electrochemical batteries. Even Lead-acid batteries can store at least ten times more energy than ultracapacitors. However, they present a lot better performance in specific power than any battery, and can be charged and discharged thousands of times without performance deterioration. These very good characteristics can be used in combination with normal electrochemical batteries, to improve the transient performance of an electric vehicle, and to increase the useful life of the batteries [3-6]. Fast and sudden battery discharge during acceleration, or fast charge during regenerative braking can be avoided with the help of ultracapacitors. Besides, ultracapacitors allow regenerative braking even when the batteries are fully charged. This paper describes the design and implementation of a control and monitoring system specially developed for an auxiliary ultracapacitor bank installed in an electric vehicle. The ultracapacitor has a capacity of 20 Farads, a nominal voltage of 300 Vdc, and a maximum voltage of 330 Vdc. The nominal current is 200 amps, and the maximum current is 400 amps. A Bi-directional DC-DC converter has been implemented for energy transfer between the lead acid battery pack and the ultracapacitor bank.

2 The control and monitoring system is based on a TMS320F241 DSP from Texas Instruments. It controls the DC-DC converter, acquires and stores data, and is capable of monitoring and sending all relevant data to a Personal Computer on a real time basis. A PC software was developed to display and store data, tune algorithm parameters and set different functioning modes. The DSP can work on automated mode, according to a pre-defined algorithm, or supervised mode in which it accepts commands from the software. 2. Ultracapacitor System Figure 1 shows a diagram of the ultracapacitor system implemented into the vehicle. The power circuit has three main components: the bi-directional DC-DC converter using IGBT s, the smoothing inductor LS, and the ultracapacitor bank. The equipment is connected in parallel to the main battery, which has 26 batteries in series (312 Vdc nominal). The capacitor voltage is kept above one third of the nominal voltage (around 100 Vdc), allowing it to store an amount of 220 Wh of useful energy. This apparently poor amount of energy can deliver 40 kw of power during 20 seconds, which is more than enough time for a good acceleration (or deceleration) without detriment in the battery life. The nominal power of the traction motor is 32 kw, and the peak power is 53 kw. 312 Vdc + BATTERY PACK (26 lead-acid batteries in series) I BATT I LOAD POWER INVERTER FOR TRACTION MOTOR Buck-Boost Converter I COMP T2 C T1 L Buck D2 Boost D1 + I CAP V CAP + 20 Farads ULTRA CAPACITOR BANK (132 ultra capacitors in series) Figure 1: Ultracapacitor System This configuration was installed in an electric vehicle implemented at the Department of Electrical Engineering of the Catholic University of Chile. The vehicle uses a high specific power brushless dc motor, with a nominal power of 32 kw and a peak power of 53 kw. The car transformed to an EV was a double cabin, Chevrolet LUV truck, with an actual gross weight of 1,920 kg (720 kg in the front and 1,200 kg in the rear). The total weight of the ultracapacitor system is 120 kgs. This vehicle is similar in size and shape to a Chevrolet S-10. Some of the performance characteristics of this EV are: maximum speed, 120 km/h; acceleration, 0-60 km/h in nine seconds; efficiency, 4 km/kwh (0.25 kwh/km). The car can carry five passengers comfortably seated inside the double cabin. The photographs of Figure 2 show: a) the vehicle used for installation of the ultracapacitor system; b) the location of the Buck-Boost converter in the front; c) the location of the main inductance L and ultracapacitors in the rear; and d) a detail of one of the five ultracapacitor boxes used in the implementation of the system.

3 Figure 2: Installation of the ultracapacitor system. a) the EV, b) the Buck-Boost converter, c) inductance L and ultracapacitors in the back, d) one ultracapacitor box. 3. Evaluation Objectives During the development stage of new equipment, especially when high power devices are involved, a thorough testing and evaluation process is essential to ensure safe operation and optimal performance. Also, considering the fact that ultracapacitors are a relatively new technology and there are few comparable experiences, the actual contribution to efficiency improvement and battery life increase (which implies operational costs reduction) has yet to be determined. Thus the evaluation process objectives can be summarized as follows: Test and improve hardware and software safety features. Ensure energy management algorithm proper operation under all conditions. Tune algorithm parameters. Evaluate the vehicle s energy efficiency under specific driving conditions, with and without ultracapacitors support. Evaluate the effective efficiency increase, under every specific driving condition, with the use of ultracapacitors. Advise about possible hardware and/or software improvements in order to increase efficiency or safety. Considering these objectives, it was clear when the ultracapacitor control device was designed that a monitoring and data acquiring feature had to be included, in order to make the debugging and evaluation process safe and easy.

4 4. Control and Monitoring System In order to have the capability of working in either stand-alone or monitored mode, an embedded system was designed for controlling and evaluating purposes. The system, based on a TMS320F241 DSP from Texas Instruments, has to sense and process several analog and digital signals and execute a predefined algorithm to manage power flow from and to the ultracapacitor bank. For the monitored mode, a personal Computer software was designed which, connected through a serial port, sends commands and retrieves data for display and further analysis. The control system purpose is to command the energy transfer between the battery and the ultracapacitors, through the DC-DC converter, making most efficient use of the ultracapacitor bank storage capacity. A previously defined algorithm, programmed in the DSP, determines the amount and direction of power transfer under every different situation. The decision-making considers the batteries and ultracapacitors model parameters, which determines these elements efficiency under every circumstance. For example, during a regenerative braking a current of 150 [A] is delivered from the drive system for a couple of seconds, but the maximum current allowed into the battery by the control algorithm (for efficiency and battery deterioration reasons) is of 90 [A]. This means that the ultracapacitors will have to take at least 60 [A], but if the batteries have been recently charged the voltage will go suddenly over the gassing voltage if 90 [A] are injected into them. As this is also not allowed by the control scheme, the voltage rising will be prevented by taking more current into the ultracapacitors. So, decisions will depend on the theoretical model of the power circuit and the actual state of every relevant component (battery, ultracapacitor vehicle speed, etc). However theoretical models are not 100% accurate and may neglect certain elements characteristics, hence the data acquiring feature importance lays on the possibility to empirically determine the system s efficiency, correct possible model flaws and even determine the optimum algorithm corrections using optimization models (optimum control for example). Also there are safety issues that need to be closely monitored during the test and evaluation process to avoid dangerous situations (ultracapacitors overcharge, short circuit, etc). To prevent any malfunction the control algorithm was compartmented in small software routines that can be tested individually. This is done by the operator in the monitored mode through communication commands sent from the monitor software Control and Monitoring System Hardware Layout Signals required to perform algorithm calculations are: battery voltage, battery current, drive train current, ultracapacitor bank voltage, ultracapacitor bank current, DC-DC (Buck-Boost) converter temperature, battery state of charge and vehicle speed. Some of these signals are available from the main inverter microprocessor, which are sent in analog format; the rest of the signals are acquired from specially installed sensors and the Ah counter. The control System final products are two PWM signals, which commutates the two IGBT s in the Buck-Boost DC-DC converter. The PWM values are calculated as part of a closed loop PI control, comparing a preset current reference and the measured current from the DC-DC converter. The power transfer algorithm calculates the preset current value for the current control, considering the battery state of charge, the battery voltage, the ultracapacitor charge; the vehicle s speed and the power drive system current. The main components and their connections are shown in figure 3.

5 Sensors: Current Voltage Temp. Gating Signals PWM Buck- Boost Converter Digital Signal Processor TMS320F241 Ah Counter Data Serial Comm. Motor Control µ Processor EVPH 332 Figure 3: Control and Monitoring System layout. Every signal acquired by the DSP, except for the Ah counter s battery state of charge, is in analog format. Although the A/D converter sampling rate can be set up to 83 KHz for 6 signal conversion, a 12 KHz rate is used. A digital filter applied to avoid aliasing Monitoring and Control System Implementation A circuit board was designed containing the DSP, the required signal amplifiers, communication circuits and IGBT gate power supplies. As the DSP goes through a reset state during the turn on sequence, both PWM signals qualify as digital ones during this process. To avoid shot circuit in this circumstance, a supply voltage and reset supervisor circuit was included in the design. Figure 4 shows the circuit board with its components and connectors. This board is enclosed in a protective box.

6 Figure 4: DSP control circuit board and connectors. A separate flash memory was installed in the circuit board to store the algorithm parameters and some data regarding the vehicle that can t be reseated (battery state of charge for example). This feature allows algorithm parameter tuning and storing in nonvolatile memory without having to reprogram the processor. However there are changes that need to be made to the DSP s program during the test and evaluation process in order to correct errors and improve performance. To make this in a simple way without removing the DSP or the circuit board from the vehicle, a flash programming feature through the serial port is used. The operation only requires changing a dipswitch position and resetting the processor. The dipswitch and reset button are conveniently located on one of the boxes side panel al shown in Figure 5. Figure 5: DSP control panel and operation indicators. The panel LED indicators show if the system is powered and what operation is performing, charging ultracapacitors, discharging them or just resting. If an invalid operation has occurred the Fail light glows indicating debugging is needed.

7 The monitoring software was programmed using a Visual Basic compiler. Figure 6 shows the software s main window. The screen represents a control panel in which the user has complete control of the DSP and is capable of storing and displaying data. Figure 6: monitoring software control panel. 5. Preliminary Tests Results The still undergoing evaluation process contemplates several tests to correct possible problems and ensure software algorithms proper operation. But during this process hardware parts and power circuit design is also being tested. Some of these test results are presented in this chapter. The Buck-Boost converter s inductance was one of the sensible hardware design parts, even more considering it was hand assembled using an aluminum metal sheet and air core. To test the actual inductance value a constant PWM test was made. The monitoring software was used to perform the controlled test, worst possible condition values were set (duty cycle= 0.5, and the greatest voltage difference possible between the batteries and the ultracapacitors), and current ripple was measured using a Hall effect current sensor. Figure 7 shows the results obtained under these conditions.

8 Figure 7: Ripple current during acceleration and deceleration It is clear in figure 7 that the worst possible condition current ripple value is a little over 4 amps p-p. This is a good performance considering that the Buck-Boost converter can transfer up to 400 amps. A step reference change test was made to the current control. This test s results are shown in Figure 8. Figure 8: Step response for Buck and Boost operation. The test was performed applying a step change from zero to nominal current values (+200 amps, -200 amps) using the monitor software to command changes manually. The response was satisfactory achieving steady state in about 60 ms. The currents where measured using a Hall effect current sensor. The Estimated Series Resistance (ESR) is one of the ultracapacitor parameters that have more significance to this application, because it determines the ultracapacitor losses and efficiency under different operation conditions. To calculate the actual value of this parameter a constant current test was performed. Considering the fact that usually ultracapacitor currents vary very slowly in this kind of application, the ESR value obtained can be assumed valid in every condition for this particular equipment. For this test the monitor software was used, setting a constant current to and from the ultracapacitors of 200 amps. Data was also acquired using the monitor software through the serial port. Figure 9 shows the ultracapacitor pack voltage during the test.

9 V COND [V] V COND [V] CHARGE DISCHARGE I=-200 Adc I=200 Adc t [s] t [s] Figure 9: Constant current tests to determine the ESR value. The characteristic voltage step variation was between 25 [V] and 28 [V] for this test, this value represents a mean ESR value for each capacitor between 0.95 [mohm] and 1.06 [mohm]. The manufacturer of this ultracapacitors declares an ESR for direct current of 1 [mohm], which is very close with the value calculated in this test. Initial ultracapacitors charge, when the vehicle is being started, is not trivial because current through it is no easily controlled when it is discharged. The charge pattern is shown in Figure 10. Ultracapacitor Charge Buck-Boost Current [A] Ultracap. Charge [Wh] Time (s) Figure 10: Ultracapacitor charge pattern. The control algorithm determines the initial charge required in the ultracapacitors to assist the batteries during acceleration. According to the charge reference, the control normally applies a PI gain to the charge error determining the reference passed on to the current control, this is when the ultracapacitor pack voltage is over 50 [V]. When the ultracapacitors voltage is below 50 [V] current is not easily controlled through it, because the capacitors are seen as a short circuit from the batteries side. That is why a charge pattern was implemented for initialising the system. The pattern consists on a constant duty cycle PWM signal applied until the ultracapacitors reach a voltage of 50 [V], then a constant current of 30 [A] on the ultracapacitor side is fixed until they reach 80 [V]. After this, normal charge control operates, reaching almost immediately the ultracapacitor current limit of 150 [A] (which is seen as a ramp in the Buck-Boost Current as the Battery-Ultracapacitor voltage ratio changes). Then the battery current limit of 70 [A] is reached and a constant current can be seen in the pattern. When the charge reference is reached a reset is triggered on the integral part of the charge control (PI) to avoid long oscillations.

10 6. Conclusions An embeded system, based on a TMS320F241 DSP, was designed and implemented to manage power flow between an ultracapacitor bank and lead-acid bateries on an electric vehicle. The ultracapacitor prupose is to allow higher accelerations and decelerations of the vehicle with minimal loss of energy, and minimal degradation of the main battery pack. The control system is also capable of acquiring and displaying data on a real time basis. During the evaluation process, which is still under way, compartmented control has prooved usefull for determining maximum current ripple, current control step resonse and ultracapacitors ESR, without major software modification. A charge pattern has been applied to the charge control for system inicialisation, avioding current oscillation duet to ultracapacitor low voltages. It is expected that data acquisition and further analysis will help optimize the power management algorithm and lead to conclusions about the actual contribution of ultracapacitors to improve efficiency and increase battery life. 7. Acknowledgments The authors want to thank Conicyt through the Fondecyt projects N and , for the financial support given to this work. 8. References [1] Maxwell. Ultracapacitors Data sheets and technical information for 1,000 and 2,500 Farads, [Maxwell publications] [2] Jun Takehara, and Kuniaki Miyaoka, EV Mini-Van Featuring Series Conjunction of Ultracapacitors and Batteries for Load Leveling of its Batteries, Technical Research Center, The Chugoku Electric Power Co., Inc., Kagamiyama, Higashihiroshima-City, Hiroshima 739, Japan. 14 th Electric Vehicle Symposium, 1996 [on CD ROM]. [3] F. Caricchi, F. Crescimbini, F. Giulii Capponi, L. Solero, Ultracapacitors Employment in Supply Systems for EV Motor Drives: Theoretical Study and Experimental Results, University of Rome. 14 th Electric Vehicle Symposium, 1996 [on CD ROM]. [4] A. F. Burke, Electrochemical Capacitors for Electric Vehicles. Technology Update and Implementation Considerations, University of California at Davis, EVS-12 Symposium Proceedings, pp.27-36, [5] Hans Kahlen, Energy and Power from Supercapacitors and Electrochemical Sources, 16 th International Electric Vehicle Symposium, Beijing, China, October 1999, paper PA 95 en CD-ROM. [6] B. J. Arnet, L. P. Haines, Combining Ultracapacitors with Lead-Acid Batteries, 17 th International Electric Vehicle Symposium, Montreal, Canada, October 2000, paper 2B-2 en CD-ROM. 9. Affiliation Juan W. Dixon, Ph.D. Department of Electrical Engineering, Catholic University of Chile, Casilla 306, Correo 22, Santiago, Chile; Phone , Fax , jdixon@ing.puc.cl J. Dixon got the Ph.D. degree from McGill University, Montreal, Canada; From 1977 to 1979 he was with the National Railways Company (Ferrocarriles del Estado) Since 1979 he is Associate Professor at the Catholic University of Chile Micah D. Ortúzar Department of Electrical Engineering, Catholic University of Chile, Casilla 306, Correo 22, Santiago, Chile; Phone , Fax , mortuzar@puc.cl M. Ortúzar is working towards the Master of Science degree from Catholic University of Chile, Santiago, Chile;

11 Jorge A. Moreno Department of Electrical Engineering, Catholic University of Chile, Casilla 306, Correo 22, Santiago, Chile; Phone , Fax , J. Moreno is working towards the Master of Science degree from Catholic University of Chile, Santiago, Chile;