Fig. 1 is a schematic diagram of the structure of an automobile ESP, and its electronic components mainly include an electronic control unit (ECU), a steering wheel sensor, a longitudinal acceleration sensor, a lateral acceleration sensor, a yaw rate sensor and a wheel speed sensor. ESP is an important electronic control system to ensure driving safety, and the normal operation of its sensors is the basis of effective control. This paper introduces the characteristics of ESP sensor, designs the hardware interface and software interface of the sensor, and verifies it in the real vehicle test. Two. The introduction of common ESP sensors is shown in Figure 1 and Figure 2, and the common ESP sensors are as follows. 1. The steering wheel angle sensor ESP identifies the driver's operation intention by calculating the steering wheel angle and its change rate. The steering wheel angle sensor converts the steering wheel angle into a signal that can represent the driver's required driving direction. The steering wheel angle is generally determined according to photoelectric coding, and the code wheel installed on the steering column contains coding information such as rotation direction and rotation angle. The information on this coded disk is scanned by a proximity photoelectric coupler. When the ignition switch is turned on and the steering wheel angle sensor rotates for a certain angle, the processor can determine the current absolute steering wheel angle through the pulse sequence. Communication between steering wheel angle sensor and ECU is generally completed through CAN bus. 2. Yaw angular velocity sensor The yaw angular velocity sensor detects the deflection of the car along the longitudinal axis, which represents the stability of the car. If the deflection angular velocity reaches a threshold, it indicates that the car is in a dangerous working condition of measuring slip or swinging tail, and then ESP control is triggered. When the car deflects around the vertical axis, the vibration plane of the micro tuning fork in the sensor changes, and the yaw rate is calculated by the change of the output signal. 3. The acceleration sensor in the longitudinal/lateral acceleration sensor ESP includes a longitudinal acceleration sensor along the forward direction of the automobile and a lateral acceleration sensor perpendicular to the forward direction. The basic principle is the same, but it is installed at an included angle of 90. ESP generally adopts micromechanical acceleration sensor. Inside the sensor, a small piece of dense material is connected to a movable cantilever, which can reflect the longitudinal/lateral acceleration of the car. Its output is about 2.5V at rest, with positive acceleration corresponding to positive voltage change and negative acceleration corresponding to negative voltage change. Every 1.0 ~ 1.4V corresponds to 1g acceleration change. 4. Wheel speed sensor When detecting wheel speed signals in automobiles, the most commonly used sensor is the electromagnetic induction sensor. The general practice is to install the sensor in the non-rotating part of the wheel assembly (such as steering knuckle or axle head), which is opposite to the gear ring made of magnetic conductive material that rotates with the wheel. When the gear ring rotates relative to the sensor, due to the change of reluctance, an AC voltage signal is excited on the sensor, and the frequency of this AC voltage is proportional to the wheel speed. ECU uses a special signal processing circuit to convert the sensor signal into a square wave with the same frequency, and then calculates the wheel speed by measuring the frequency or period of the square wave. In the original ESP system, both the longitudinal/lateral acceleration sensor and the yaw rate sensor were realized separately. At present, the sensor cluster mode is basically adopted, and the three sensors are designed as a whole, and communicate with ECU through CAN bus. As shown in fig. 3, the sensor assembly is produced by Siemens and Bay Company. In order to add new ESP functions and better control the stability system of the whole vehicle, Bosch Company put forward the modular concept of software and hardware, and developed the third generation of highly flexible and low-cost chronic sensor assembly DRS MM3.x Third, the block diagram of ESP universal sensor interface design is shown in Figure 4. The signal of steering wheel angle sensor in the figure is processed by microcontroller and sent to ECU through CAN bus (B in Figure 4). The yaw rate sensor and the longitudinal/lateral sensor are designed in the same module because they have similar signal characteristics and installation positions (A in Figure 4). ESP requires high real-time signal of wheel speed sensor, so the signal is directly sent to ECU after conditioning (C in Figure 4). In figs. 4 (a) and 4 (b), a microprocessor is needed to process signals and transmit data through the CAN bus. SAK-C 164CI of Infineon company is selected in this paper. The chip is specially designed for automotive applications, with built-in AD converter, input signal capture and orthogonal decoder, which has fast operation speed and is very suitable for sensor signal processing of ESP. 1. Steering wheel angle sensor interface The output of the steering wheel angle sensor is an orthogonal coded pulse. Orthogonal coded pulses consist of two pulse sequences with different frequencies and a fixed phase shift (90) of a quarter period, as shown in fig. 5. By detecting the phase relationship between the two signals, it can be judged as clockwise and counterclockwise, and the signals are counted up/down accordingly, so as to get the current counted cumulative value, that is, the absolute steering angle, and the change rate of the steering angle, that is, the angular velocity, can be measured by the signal frequency. In addition, the output signal of the steering wheel angle sensor is zero. When the steering wheel is in the middle position, the signal is 0V, otherwise it is 5 V.. Through this signal, the absolute angle can be calibrated online. The interface circuit between C 164CI and the steering wheel angle sensor is shown in Figure 6. An orthogonal decoder with incremental coding is built into the chip. The decoder uses two pins of Timer3 (T3IN and T3EUD) as the input of orthogonal pulses. After the relevant registers are set correctly, the value of the data register of Timer3 is proportional to the steering wheel angle, so the steering wheel angle sensor used in this paper corresponds to 44 pulses per revolution, and the data register of Timer3 is T3. Through operation, the change rate of rotation angle can be obtained.
The microcontroller sends the calculated parameters to ECU through CAN. 2. Wheel speed sensor interface According to the signal characteristics of the wheel speed sensor introduced in the previous part, the interface circuit shown in Figure 7 is designed. The circuit adopts two-stage filtering and shaping to ensure that the wheel speed signal will not be lost at very low speed and avoid signal interference caused by suspension vibration. In the figure, the first-stage hysteresis comparison is introduced by resistor R2 and the second-stage hysteresis comparison is introduced by 74HC 14. 3. The yaw rate, the yaw rate of longitudinal/lateral acceleration sensor and the installation position of longitudinal/lateral acceleration sensor are basically the same, and the output is 0-5 V simulation. Because the signal fluctuation characteristics caused by car bumps are consistent, they are packaged in the same module. Its hardware interface is shown in Figure 8, which realizes hardware analog pre-filtering, suppresses high-frequency noise components in analog signals from sensors, and prevents aliasing during sampling. The operational amplifier uses LMX324 with full swing output. By adjusting the parameters of each resistance-capacitance element in Figure 8, the cut-off frequency and delay of the filter can be set. In the process of driving, when driving on a good road, the delay is as small as possible because of the good signal, while when driving on a bumpy road, it is hoped that the filtering effect will be better. However, since the frequency characteristics of hardware filters cannot be modified in real time once they are designed, it is necessary to design digital filters in software. Wiener filter, Kalman filter, linear predictor and adaptive filter are commonly used in digital filtering. Here, the first-order low-pass filter with small calculation and good real-time performance is selected. The choice of k depends on the current road condition, which is identified by the original signal before digital filtering. The microcontroller packages the filtered signal, original signal, K value and road identification result, and then sends it to ECU through CAN bus. Figs. 9a and 9b are a set of comparison curves of longitudinal acceleration sensors collected in real vehicle tests on bumpy roads.
4. Conclusion The structural characteristics and signal characteristics of common sensors in ESP system are discussed, and the signal processing interfaces of each sensor are designed, including hardware interface circuit and software processing scheme. An integrated module including yaw rate sensor and longitudinal/lateral acceleration sensor is designed, which transmits data with ECU through CAN bus and has good anti-interference and reliability.