Research On Matching Technique For EV Powertrain

With the advantages of high-energy efficiency and no tail emission pollution, an electric vehicle (EV) is an important way of resolving energy crisis and environmental pollution.Inadequate endurance mileage and long charging period are the main restriction factors to hinder the development of EV due to the poor energy density of battery and low efficiency of electric drive system. In order to improve the efficiency of EV drive system and prolong EV endurance mileage, the reasonable match of battery pack, driving motor and transmission system parameters of power-train should be carried out to suit for EV most common driving conditions, and proper control strategy should be developed.The main works around the matching techniques of EV power-train are described in the following sections.1. Development of test bench for the performance testing of power-trainA test bench for the performance testing of power train system of EV has been developed. The bench consisted of motor power supply and battery energy consumption system, driving motor and its controller, dynamometer and its control system as well as bench’s control and data acquisition system. The characteristic test for key parts of the power-train system, the performance test and matching effect evaluation for the power-train system and electronic control units’ validity verification of EV power-train could be conducted using the bench. The communication modes of equipments with that test bench constructed include CAN bus communication, RS-485series communication and RS-232series communication. In order to meet the demands of data concentrated acquisition and equipment remote control, a communication mode changing system has been developed for data sharing between different communication modes using processor infineon XC164CM and the test bench communication network is constructed based on CAN bus. To realize the communication functions of the test bench, the application layer protocol of CAN bus has been designed, the nodes’ source address and parameter groups are all defined clearly according to the protocol of SAE J1939. The functions of data acquisition and equipments’ control have been achieved by means of communication network. The control program flow of equipment control and data acquisition has been discussed with the example of battery discharge process through intelligent battery discharge unit, and the protection strategy of the test bench is also discribed.Tests, such as battery pack discharging and power train performance following the basic urban test cycle, are carried out at the bench, and the results show that the test bench works well and satisfies the requirements for EV studying.2. Characteristic analysis of key components of EV power-trainTo guide EV power-train’s optimal matching and developing suitable control strategy, the efficiency characteristics of battery pack, driving motor and other key parts of power-train should be understood. The coloumbic efficiency, the open circuit voltage and the voltaic efficiency of a320V/100A·h LiFePO4/C li-ion battery pack are investigated using the test bench. The results indicate that the coloumbic efficiency exceeds99%for different charge current and discharge current, the voltaic efficiency changes with the battery pack’s current and its state of charge (SOC). The battery has a high charging efficiency of over92%at small charge current and in the range of20%-80%of SOC, and a high discharging efficiency with small discharge current and high SOC. A model that descibes the relationship of the battery charge and discharge efficiency to current and SOC is established based on the measured data. The comparisons between modeled results and measured values indicate that the model is valid and the maximum relative error is within0.57%at some typical points which are selected according to vehicle opretion conditions.The efficiency characteristics of a32kW AC asynchronous motor during the high-frequency operation time area and the field-weakening region are tested and analyzed using the test bench. The test results show that the efficiency of a traction motor varies with the motor speed and torque. There is an area in which the efficiency is high and the most efficient region locates in the middle of the speed and torque range, corresponding to the power output between0.3and1.4of rated power(Pe). Contrast to the high efficiency region, lower power output at low-speed or light load leads to low efficiency, and the efficiency will also decrease significantly when the load power is much greater than the rated power. A model to describe the efficiency of the motor system is established using quartic function based on the test results and the model is validated by comparing measured results that are obtained at the rated torque of the motor to the simulation values, and the results show that the maximum relative error is within3.4%.The efficiency characteristics of electric driving system under driving and braking energy recovering process are tested and results indicate that there is an optimal efficiency region nearby the rated speed and middle torque. Prediction models of energy recovering efficiency and driving efficiency are developed based on measured data and the models’ validity are verified by testing results.3. Test and analysis of power-train running conditions in Jinan’s drive cycleThe driving cycle has specific characteristics in different city. To master the characteristics that in Jinan, a driving cycle of Jinan is constructed, and according to that the ordinary EV running conditions are contoured. The running conditions are helpful to optimally match power train and develop proper control strategy. An on-vehicle data acquisition unit is developed to acquire vehicle real-time running data, and the valid data was packed and transmitted to monitoring center based on GPRS network. Typical roads in Jinan are selected considering lane number, road grade, traffic flow density and other factors, and then the running data of pure electric microbuses driving in the roads is acquired for up to15days and2.6million groups data are collected. A driving cycle constructing method is proposed in which vehicle energy consumption status is included. Key factors that affected vehicle’s energy consumption, such as road grade, vehicle transient specific power, vehicle speed, and vehicle acceleration rate are analyzed. Taking vehicle’s energy consumption and road grade into consideration,27characteristic values of kinematic sequences are proposed. Candidate driving cycles are constructed according to the analysis result of kinematic sequences by using tools such as principal composition analysis and clustering analysis etc. And the actual driving cycle for Jinan is selected from candidate driving cycles by comprehensive consideration of the correlation coefficient, relative error and key parameters probability distribution. By statistical analysis for Jinan’s driving cycle, the area of the greatest operation time is within the vehicle speed of10km/h～40km/h, wheel torque demands of-200N·m～300N·m and power demands of-2kW～3.5kW.4. Research on software in the Loop simulation technology for EV power-trainSoftware in the Loop simulation system for EV power-train is developed for the purpose of matching the parameters of power-train. And the simulating system is established using MATLAB/simulink, which includ driving cycle model, vehicle kinematics model, vehicle controller model, traction motor model, motor controller model, battery pack model and selection model which used to select key parts of power-train. The simulating system is validated by comparing the modeled value with the measured data and by comparing the modeled value with the result that are obtained from simulation model developed using ADVISOR software. The parameters of a motor, a battery pack and a manual transmission are selected for a power-train by simulation using the simulating system. The efficiency characteristics of the traction motor and the battery pack are measured on the EV test bench. The electrical driving system and the mechanical transmission system are matched using a chassis dynamometer and the results show that the most efficient region of the power train overlaps the greatest operation time area. With a7.5kW AC asynchronous motor, a192V/100Ah LiFePO4/C battery pack and a constant gear ratio6.18, the EV endurance mileage reaches169km under40km/h testing by constant speed metering method, and reaches160km, costing12.01kW·h/100km, according to basic urban driving cycle.5. Development of controller and control strategy for electric driving systemA motor controller unit is designed using Infineon TC1782F microcontroller and Hybrid PACK1IGBT module, and the motor control strategy is developed based on the vector control. And in the strategy, coordinate transformation module, rotor flux angle calculation module, voltage space vector location calculation module and basic voltage vector acting time calculation module are included. In real-word operation, some factors that affect EV drive performance, such as DC voltage ripple, rotor resistance varying with motor temperature, field weakening control at high-speed range, proportional-integral (PI) parameters tuning in the motor speed control loop, voltage supply and discharge current of the power system are analyzed. Tests about motor torque response to a step change in the torque demand and closed loop speed control performance are carried out and the results show that the motor controller could work effectively.EV’s driving modes are established, mode identifying and mode switching pattern are constructed using Matlab/simulink&stateflow. The driving control strategy of EV’s power-train is developed, in which the signal of accelerator pedal is smoothed and filtered; the incremental PID control based on speed deviation and the optimal efficiency path control are adopted in steady mode and transient mode respectively; power output is limited in failure mode. Under transient conditions, in order to further improve power-train’s efficiency, an optimizing power-train’s control strategy is proposed based on the power-train efficiency model. Simulating results and testing results indicate that the control strategy is valid.With the tests of AC asynchronous motor and a LiFePO4/C battery pack on the test bench, the relationship of energy recovery efficiency to motor speed, brake torque, battery pack’s SOC, battery pack’s temperature is studied. The restriction of motor’s temperature to the maximum brake torque setting is discussed. For recovering energy effectively, under the premise of driving comfort, the dynamic matrix control (DMC) algorithm is adopted to develop the strategy of coasting condition according to energy recovery efficiency model and conventional vehicle’s coasting resistance. To ensure vehicle safety, a braking control strategy is designed that distributes the required braking forces among regenerative braking and frictional braking referring to wheel slip rate.The test results indicate that driver’s driving characteristics have great influence on EV energy consumption. The difference of energy consumption driving by different drivers is tested and driving characters of affecting the difference are discussed according to test results of electric vans running in Jinan city. Factors influencing EV energy consumption, such as vehicle acceleration rate, speed, braking deceleration rate and motor power overload are tested and analyzed. By adding a control mode, in which motor torque and over-speed are limited, and therefore the driving parameters to that energy consumption is sensitive are optimized. The test of optimized electric vans is carried out, and results indicated that energy consumption of optimized vans could be decreased as more as34.9%when compared with the original vehicle.6. Performance examination of matched EVChassis dynamometer test and vehicle road test are carried out using the matched EV. Results of chassis dynamometer tests show that the vehicle maximum speed satisfies the EV’s design index requirement of umax>80km/h, and the endurance mileage reaches177km, costing10.71kW·h/100km, over the urban driving cycle. The vehicle test indicates that time used to accelerate the vehicle from zero speed to60km/h is10.88s when the motor torque is limited to the range less than120N-m, and the time also meet the demands of vehicle acceleration performance. The test results about driving modes’management strategy show that the operation mode can identify operation status accurately, operation mode can switch smoothly and the control strategy works well. The safety protection function for power-train is tested on chassis dynamometer and the results indicate that the battery management system and the motor controller can provide effective protection for the key components of power-train according to user’s settings.