ESP (Electronic Stability Program) is a landmark invention in automotive electronic control. Different research institutions use slightly different names for this system. For example, Bosch initially called it Vehicle Dynamics Control (VDC), but now Bosch and Mercedes-Benz call it ESP; Toyota calls it Vehicle Stability Control (VSC), Vehicle Stability Assist (VSA), or Electronic Stability Control (ESC); and BMW calls it Dynamics Stability Control (DSC). Despite the different names, they all add a lateral stability controller to traditional automotive dynamics control systems such as ABS and TCS. By controlling the distribution and amplitude of lateral and longitudinal forces, they control the vehicle's dynamic motion patterns under any road conditions, thereby improving the vehicle's power performance in various operating conditions.
This article provides a circuit analysis of the sensors and interface technologies in ESP:
Circuit principle
Steering wheel angle sensor interface
The steering wheel angle sensor outputs orthogonal coded pulses. These pulses consist of two pulse sequences with varying frequencies and a fixed phase offset of one-quarter of a period (90°), as shown in Figure 1. By detecting the phase relationship between the two signals, the clockwise and counterclockwise directions can be determined. Based on this, the signals are incremented/decremented to obtain the current cumulative count, which is the absolute steering wheel angle. The rate of change of the angle, i.e., angular velocity, can be measured by the signal frequency. Additionally, the steering wheel angle sensor has a zero-position output signal. When the steering wheel is in the center position, this signal outputs 0V; otherwise, it outputs 5V. This signal allows for online calibration of the absolute steering angle.
Figure 1. Waveform of the pulse sequence from the steering wheel angle sensor
The interface circuit between the C164CI and the steering wheel angle sensor is shown in Figure 2. The chip has a built-in incremental encoding quadrature decoder. This decoder uses two pins of Timer 3 (T3IN, T3EUD) as inputs for quadrature pulses. After correctly setting the relevant registers, the value of the Timer 3 data register is proportional to the steering wheel angle, thus facilitating angle calculation. The steering wheel angle sensor used in this paper corresponds to 44 pulses per revolution. Assuming the Timer 3 data register is T3, the absolute angle is...
Figure 2 Steering wheel angle sensor interface circuit
By performing differential calculations, the rate of change of steering angle can be obtained. The microcontroller then sends the calculated parameters to the ECU via CAN.
Wheel speed sensor interface
Based on the characteristics of the wheel speed sensor signal described in the previous section, the interface circuit is designed as shown in Figure 3.
Figure 3 Wheel speed sensor interface circuit
The circuit employs two stages of filtering and shaping to ensure that the wheel speed signal is not lost at extremely low speeds, while also avoiding signal interference caused by suspension vibration. The first stage of hysteresis comparison is introduced by resistor R2, while the second stage of hysteresis comparison is introduced using a 74HC14.
Yaw rate, longitudinal/lateral acceleration sensors
The yaw rate and longitudinal/lateral acceleration sensors are installed in basically the same location, and their outputs are all 0V-5V analog signals. Because the signal fluctuation characteristics caused by vehicle bumps are consistent, they are packaged in the same module. The hardware interface is shown in Figure 4, which implements hardware analog pre-filtering to suppress high-frequency noise components in the analog signals from the sensors and prevent aliasing during sampling.
Adjusting the parameters of each resistor and capacitor in Figure 4 allows you to set the filter cutoff frequency and delay. During vehicle operation, on smooth roads, the signal is strong, so a small delay is desirable. However, on bumpy roads, a good filtering effect is needed. Since the frequency characteristics of hardware filtering cannot be modified in real-time once designed, a digital filtering stage needs to be designed in software. Commonly used digital filters include Wiener filters, Kalman filters, linear predictors, and adaptive filters. Here, a first-order low-pass filter with low computational cost and good real-time performance is chosen.
Figure 4. Interface circuit for yaw rate and longitudinal/lateral acceleration sensors
This paper discusses the structural characteristics and signal properties of commonly used sensors in ESP systems, and designs the signal processing interfaces for each sensor, including hardware interface circuits and software processing schemes. An integrated module incorporating yaw rate and longitudinal/lateral acceleration sensors was designed, which transmits data with the ECU via a CAN bus, exhibiting good anti-interference performance and reliability.
