Research on New Wheel Hub Motor Drive Electric Vehicle Electronic Differential Control System

At present, some novel electric vehicles (EVs) adopt independent driving methods. The representative is the IZA electric vehicle launched by Tokyo Electric Power. The integrated technology is a direct driving method. Each wheel is a hub motor, no need to drive. The mechanism and the differential gear can distribute the power of the two motors according to the required power, so the efficiency of the whole system is improved. At the same time, an electronic differential drive control system is required for the drive unit.
There are few related studies available, most of which focus on the design of special motors with differential operation. Based on the analysis of the differential phenomenon of the car, this paper proposes a new electronic differential scheme for the electric motor of the hub motor for medium and low speed operation, and designs and implements the TMS320F2407DSP based on two events. The output of the two-wheel hub motor drive control system capable of controlling two motors) is widely used to determine the driving strategy of the vehicle. (See the Ningbo Science and Technology Commission Youth Fund-funded project. The speed ratio of the inner and outer wheels when the vehicle is purely rolling is the turn. Radius ratio, this model only performs static analysis, does not consider the influence of tires, and ignores the centrifugal force and centripetal force when the vehicle is turning. According to this model, the electronic differential scheme is proposed. This paper analyzes that it is not reasonable enough. In the case of a given corner, the four degrees of freedom of the four wheel speeds and the vehicle speed are 1, so if the speed control of the two drive wheels is performed at the same time, the actual system has slight errors, which will cause conflicts. The slip rate between the various wheels is controlled to be different or even slipping, resulting in system instability and affecting the whole
In summary, this paper believes that wheel speed driving electric vehicle electronic differential speed should not use wheel speed as a control variable.
2.2 New electronic differential control scheme The electronic differential scheme designed in this paper considers the change of the vertical load of the wheel during cornering so that the adhesion rate of the two driving wheels is equal, and the driving torque of the two wheels is distributed based on this. , so that the possibility of slipping of the vehicle is minimized. Considering the wind resistance and the effect of the lateral force of the tire, the motion state of the vehicle is simulated under the given total power output. As shown, the torque can be seen when the speed and the angle are large. The distribution ratio varies greatly, and the rollover torque generated by the centrifugal force of the vehicle body movement plays a decisive role.
The vehicle turning torque distribution ratio simulation result is further simplified, and only the influence of centrifugal force on the vertical load is considered. From the analysis of automobile dynamics, it can be seen that the lateral rolling moment generated by centrifugal force for static or pen turning is N3 when the wheel is driven after turning. It can be proved that for the turning body of <0.7 and the turning condition of <30, r= Estimated, the error is within 5%.
Therefore, under this condition, the available torque ratio is (6), and the simulation results are as shown. It can be seen that this simplified calculation can meet the engineering requirements under the condition that the medium and low speeds and the angle of rotation are not too large.
Simplified vehicle turning torque distribution ratio simulation result The control pedal input is equivalent to the torque control command, and the current control with linear adjustment negative feedback is used. The control block diagram shows the output characteristics as shown in equation (7). The mechanical characteristics are as shown. It is similar to the pedal control of an internal combustion engine car, and has a driving feeling similar to that of a conventional car.
Torque Control Block Diagram The overall block diagram of the DSP2407-based electronic differential control system designed in this paper is shown. The power circuit adopts the half-bridge modulation mode, which can reduce the switching loss of the inverter. The three-phase Y-connected brushless DC square wave hub motor adopts two-way conduction mode, that is, two power tubes are turned on every moment, every The 60° electrical angle is commutated once, and each power tube is turned on by 120° electrical angle. DSP uses Texas Instruments' motor microcontroller TMS320X2407, which uses high-performance static CMOS technology, which reduces the power supply voltage to 3.3V, reducing the power consumption of the controller. Two event manager modules 12-way full comparison PWM output It can realize the control of two motors. Because of the abundant resources in the DSP chip, such as the function modules such as AD conversion, the control circuit is greatly simplified. Due to space limitations, this article focuses on several aspects of design.
3.2 New half-bridge modulation phase current detection method For the system, two-phase three-phase six-shot operation mode is adopted, and PWM modulation adopts half-bridge modulation mode. When the PWM is inactive, the DC current is 0, so during the PWM valid period The current signal sampling can effectively detect the phase current of the motor. This paper realizes the single current sensor to detect the phase current simply and accurately through the improvement of software design. The timing diagram for AD conversion is briefly described as follows: The current sensor is placed on the DC terminal, and the PWM signal is asserted high by setting the DSP control word ACTRA/B. The general-purpose timer T1 of the DSP is set to the continuous addition and subtraction control system. Block diagram PWM cycle = continuous addition and subtraction counting mode PWM signal phase current diagram DC terminal; current "schematic DSP2407 controller is 3.3V, but its 5V interface circuit is inevitable, the existing 3.35V conversion chip price The counting method is a method of publi for every 51 periods of interruption, and the detected current value is closer to the actual current average. In addition, due to the characteristics of DSP2407, in order to reduce the time of ADC conversion, the output signal of the current sensor passes. After the conventional filter amplification, a first-stage emitter follower circuit is added, so that the output resistance of the signal terminal is small. At the same time, the sampling and holding module of the 2407 device ADC is adjusted by changing the ACQPS3ACQPS0 bit field and the CPS bit in the register of the ADCTR1. To adapt to changes in signal impedance, so as to ensure the sampling accuracy while selecting the setting of short conversion time to adapt to the PWM pulse Small and small.
The input pin of the DSP sometimes has an internal pull-up or pull-down circuit, so that it does not affect the impedance calculation of the interface circuit, but affects the DC offset calculation. Several interface methods are given to simplify the analysis without regard to internal pull-ups or pull-downs.
3V interface circuit When the maximum supply voltage of TTL device is 5.25V, the high voltage of TTL output is 3.4V at rated current and 4.05V at no load; therefore, considering the maximum voltage difference between components, it is assumed that DSP supply voltage is The maximum allowable voltage of 3.0V is 3.3V, and the maximum differential voltage of logic high is 0.75V. If the current is limited to 75M, it is sufficient to add a 10k resistor between DSP and TTL, which produces a small RC delay ( 10knx5pF=5Qns), except for the CAN bus, this delay can be neglected, and a larger resistor can be used to reduce the current, but the delay becomes longer and the noise suppression capability is worse.
When the power supply is 5.25V, the 5VCMOS output is 5.25V when it is no-load, so the voltage difference is 1.95V when the logic is high. Therefore, it is necessary to add a voltage divider circuit. If the resistance is reduced, the input resistance is also small.
Because the output of the DSP is TTL-compatible, no special circuit is required. The high-low logic of TTL is 2.4V to 0.8V, while the output logic of 3.3V CMOS is 2.8V to 0.4V. There is a large degree in the middle, many The motor control chip is a 5V powered TTL output to a 5V CMOS input: a level shift is required between them. When R1 is 10k, the CMOS output is 0.2V to 3.3V. After D1 translation, the output is 0.8V to 3.9V, 5V CMOS input. The threshold voltage is 1V to 3.5V with a range of 0.2 to 0.4 in the middle. At the same time, there is a small delay.
3.4 Improved rotor position detection method The permanent magnet brushless hub motor of the system has a Hall sensor, which is convenient to use and low in price. However, for a motor with a large power, when the winding current is large, on the one hand, the magnetic field generated by the permanent magnet rotor will be affected to shift its spatial position. On the one hand, the magnetic field distribution near the position sensor is affected by the commutation current shock. In both cases, the signal of the Hall position sensor is in error, and even the interference does not work properly. Usually, the control scheme of such a motor is to connect the three-way position sensor output to the capture unit of the DSP device. The six-way position sensor signal of the two motors in the system needs to involve four timers and corresponding interrupts if the capture unit is used. This system discards this conventional method, and connects the position sensor output to the /O port of the DSP, reads the state of the I/O port in the timer underflow interrupt service routine that generates the PWM, and judges the corresponding position signals of the two motors, and before Sub-position signal comparison, adopting weak delay commutation and commutation locking technology, that is, not immediately commutating when the position change is detected, but continuing to perform several position detections in a small interval to further determine whether it is indeed in the position of commutation When the commutation operation is determined, the commutation is no longer performed in a minute interval regardless of whether the rotor position signal changes. This not only ensures the accuracy of the commutation process, but also simplifies the software design compared to the use of the capture unit. In the design, the PWM switching frequency is 15kHz, and the rated motor speed is 340r/min, so the timer underflow interruption interval is small enough relative to the minimum commutation time interval of the motor.
The main program part completes the system initialization, the processing of the two motor current AD sampling results, the two drive wheel speed calculations, the vehicle body speed estimation, the electronic differential algorithm and the implementation.
In the T1 timer underflow interrupt service routine, the two motor position signals are respectively read from the I/O port, and the above-mentioned weak delay commutation and commutation lock are completed, and the ACTRA/B control words of the two motors are set to start. AD sampling of the car body corner.
The T1 timer is interrupted ten times to start the current AD meter of the corresponding two motors.
Integrated module IR2130. When the protection signals of current, overvoltage and undervoltage of the two motors are generated, the output of the corresponding IR2130 is blocked by hardware, and the PDPINTA or PD-PINTB pin of the corresponding DSP is connected. In the related books of DSP, there is no specific use description for PDPINT power interruption. Based on the actual test, this paper summarizes the characteristics of PDPINTA/B power interruption of TI2407DSP, as follows: PDPINTA or PDPINTB pin signal is valid for falling edge, corresponding The PWM output becomes a high-impedance state. This high-impedance state can be released after the program is reset. At the same time, the falling edge of this pin signal generates an interrupt request at the same time. If the corresponding interrupt is not masked, the interrupt service routine is entered. After the corresponding fault occurs in the background processing, the program continues to run after the interrupt service routine is completed, but the changes to ACTRA/B will not affect its output. This design performs fault analysis in the power interruption service program and gives a fault indication. If the fault is not released, the loop is detected. After the fault is cleared, the program jumps to 0000H to reset.
The experimental result (b) is the current of the two motors during linear operation. Since the frequency of the current waveform can be converted to the motor speed, it can be seen that the torque and the rotational speed of the two motors are basically the same, and (c) the angle is 5°. When the current of the two motors, the torque of the outer motor is greater than the torque of the inner motor, and the rotation speed is higher than the rotation speed of the inner motor, achieving good electronic differential control.