Design of electronic anti-skid device for fast truck based on single chip microcomputer

Anti-skid control is an effective means to ensure the braking safety of railway vehicles and improve braking efficiency. In particular, as freight speeds increase, once skidding occurs, it may cause greater harm. Therefore, anti-skid devices should be installed on my country's fast freight cars [1]. Qiche Equipment Co., Ltd. and Meishan Rolling Stock Co., Ltd. installed mechanical anti-skid devices on the prototypes of fast freight cars they developed to solve the problem of wheelset skidding. However, mechanical anti-skid devices have the disadvantages of low sensitivity, unstable performance, inability to monitor the adhesion state and adjust the brake pressure in real time [1]. In addition, the key sensitive components in mechanical anti-skid devices are prone to wear after long-term use, resulting in a gradual decline in their performance. Electronic anti-skid devices use computer control technology to detect the axle skidding state in real time, adjust the brake cylinder pressure to prevent the axle from skidding, and can determine whether the axle is skidding based on multiple slip criteria. They have higher accuracy and are widely used in railway passenger cars, EMUs and other locomotives.

The use of electronic anti-skid devices on fast trucks requires solving the power supply problem. Southwest Jiaotong University has designed a wind power supply device that can be suspended at the bottom of the vehicle body [2]. Based on this, this paper conducts corresponding research on the use of electronic anti-skid devices on trucks and proposes an anti-skid controller based on digital signal processor (DSP-TMS320F2812) control. Fuzzy control is used as the algorithm to realize axle slip detection and anti-skid control.

1 Principle of electronic anti-skid device for fast trucks

Due to adhesion restrictions, vehicles are prone to skidding during braking, and even wheel scratches. At present, electronic anti-skid devices are installed on railway locomotives with speeds exceeding 120 km/h in my country to prevent axle skidding. In principle, the electronic anti-skid devices for fast freight vehicles and passenger cars are the same: the anti-skid device collects the speeds of the four axles on the vehicle in real time, obtains the vehicle reference speed through comparison, and then calculates the vehicle slip rate and deceleration, etc., and judges whether the axle is slipping based on the multi-slip criterion, thereby adjusting the brake cylinder pressure to prevent wheel skidding. The principle of the electronic anti-skid device designed in this paper is shown in Figure 1.

2 Hardware Design

The anti-skid device needs to collect the speed signals of each axle. The output signal of the speed sensor is generally in the form of speed pulses, and the anti-skid control can be completed after corresponding data processing. Higher computing speed and data processing capabilities can shorten the response time of the anti-skid device and improve the response accuracy. This paper selects TI's 32-bit fixed-point DSP-TMS320F2812 as the core controller of the anti-skid device host. The capture unit of its event manager can capture the jump of external pins, which can be conveniently used for the measurement of speed sensor signals. The chip has a data processing capacity of 150 MIPS and integrates a rich set of on-chip peripherals. It has both digital signal processing capabilities and powerful event management capabilities and embedded control capabilities. It is particularly suitable for the measurement and control domain that requires large-scale data processing [3]. In addition, the main modules used to complete the anti-skid control include: the timer unit and CPU timer of the event manager (EV), the peripheral interrupt expansion module (PIE), the analog-to-digital conversion module, the SPI communication interface, the watchdog, the general I/O, the external interrupt interface and the memory interface.

In order to reduce the signal interference between the system input, output and core control unit and facilitate maintenance, the anti-skid device hardware adopts a modular design as shown in Figure 1, which is divided into a signal conditioning module, a core control unit and a drive module.

The core control unit includes the minimum system, which includes TMS320F2812, clock and reset circuit, power supply and filter, JTAG, etc. In addition, the electronic skid breaker should also include other necessary application circuits, such as fault code power-off storage, display and clearing. This paper expands the EEPROM storage of the DSP's serial peripheral interface (SPI) and selects X5045 as the fault code memory, which has a storage capacity of 4 KB and can be erased and written 1 million times. For easy maintenance, this paper designs a fault code display function and selects MAX7219 as the display driver of the two-digit digital tube, which is controlled by the DSP's I/O port. In addition, the three external interrupt ports of TMS320F2812 are directly used as key control interfaces to control the display and clearing of fault codes.

The signal conditioning module and the drive module are important components of the electronic anti-skid device. The signal of the speed sensor is connected to the 4-way capture port of the DSP through signal conditioning (optical coupling isolation). The DSP collects the speed of each axle in real time and calculates the slip rate, deceleration, etc. The anti-skid device controls the solenoid valve drive circuit according to the operation of the axle, thereby controlling the anti-skid valve to charge and discharge and adjust the brake pressure.

3 Software Design

On the basis of clarifying the function of the electronic anti-skid device, the system adopts the idea of ​​modular programming. The overall program block diagram is shown in Figure 2. Since the PIE module of TMS320F2812 can support up to 96 interrupts, it brings great convenience to the modular programming of the system and enhances the readability of the system program. Among them, speed processing and skid detection control are the key modules to realize the system functions.

3.1 Calculation of slip rate and deceleration
The forward speed v of the axle is proportional to the pulse frequency f emitted by the axle sensor:

Where D is the wheel diameter, and Z is the number of pulses emitted by the wheel per revolution. As long as the speed sensor pulse frequency is measured, the speed of each axle can be calculated. Since the speed sensor frequency varies widely during vehicle braking, in order to ensure high measurement accuracy at both high and low frequencies, this paper uses the variable-period M/T method to measure speed, as shown in Figure 3. The maximum error of frequency measurement is a reference pulse signal, where the frequency of the reference pulse (37.5 MHz) is equal to the system clock frequency divided by the frequency division coefficient. Therefore, as long as the number of reference pulses corresponding to the speed pulse is measured, the frequency of the speed pulse can be calculated. Capture CAP1, the 2-pin time base is timer 1, and the counting period is set to 1 ms; capture CAP5, the 6-pin time base is timer 3, and the counting period is 1 ms. Turn on the timer period interrupt and capture interrupt, the number of reference pulses corresponding to N speed pulses:

[page]

The accuracy of speed measurement directly affects the control result, so the anti-skid device has higher requirements for speed measurement. This paper uses a function signal generator to generate a pulse signal, simulates the speed output of the speed sensor, and uses an electronic anti-skid device to measure its frequency to verify the speed measurement accuracy of the electronic anti-skid device. Table 1

The experimental results are given. Among them, fosc is the frequency of the signal generated by the function signal generator (DF1405), vosc is the speed corresponding to 200 pulses per rotation of the axle, fDSP is the frequency measured by the anti-skid device, and vDSP is the corresponding speed value. The results show that no matter at high or low speed, the absolute error of the speed measurement of the electronic anti-skid device does not exceed 0.05 km/h, and the relative error does not exceed 0.05%, which can meet the requirements of the anti-skid device for speed measurement accuracy.

3.2 Slip detection and control

This paper applies fuzzy control algorithm to realize anti-skid control. In order to improve DSP operation efficiency and shorten system response time, the system adopts offline query method to realize fuzzy control; a two-dimensional fuzzy controller is designed through MATLAB/Simulink simulation; the basic domain of slip rate is [0, 0.25], and the basic domain of deceleration is [-4, 4]. In the actual control process, as long as the quantized values ​​of slip rate and deceleration are measured, the current control quantity can be obtained by table lookup method. The control output of the electronic anti-skid device is the inflation and deflation time (0-500 ms), positive value means inflation, negative value means abandonment, and 0 means pressure maintenance.

Set the CPU timer 0 interrupt cycle to 5 ms (i.e., the single inflation and deflation time is 5 ms), and set the CPU timer 2 interrupt cycle to 100 ms[4] (i.e., the coasting state detection cycle). In actual control, the fuzzy control quantity OP is maintained within [-50, 50].

4 Anti-slip simulation test

In order to verify the control effect of the anti-skid device, this paper conducted an anti-skid simulation test in the laboratory, as shown in Figure 4. The test uses LabVIEW software as a platform to simulate the vehicle speed signal and the slipping axle speed signal. The control of the brake cylinder pressure is achieved through the control simulation of the anti-skid device on the volume chamber pressure. The two axle signals obtained by MATLAB/Simulink simulation are read on the computer through LabVIEW programming, and the NI 6008 data acquisition card is controlled to generate two voltage signals proportional to the speed. The voltage signal is converted to a pulse signal whose frequency is proportional to the speed through the voltage-frequency conversion circuit. The speed pulse signal is connected to the CAP pin of the DSP through the optocoupler isolation of the signal conditioning module. The voltage-frequency conversion is implemented using the AD654 chip. AD654 is a low-drift, high-linearity, low-cost voltage-frequency conversion chip that requires only a few peripheral components to achieve voltage conversion.

The control object of this test is the volume chamber, whose pressure is controlled by the inflation valve and the exhaust valve. The pressure of the volume chamber is connected to the data acquisition card through the pressure sensor, and the pressure of the volume chamber is displayed in real time on the computer through LabVIEW programming.

The test simulates the situation that the axle slips twice during the braking process of the vehicle, and the results are shown in Figure 5. When the anti-skid device detects that the axle is sliding, the pressure of the volume chamber is adjusted in real time to prevent the axle from continuing to slide. After there is no sliding, the pressure of the volume chamber is restored to ensure the smooth implementation of the braking force.

Anti-skid control is one of the key technologies for fast truck braking, and is of great significance for ensuring braking safety and improving braking efficiency. The fast truck electronic anti-skid device designed in this paper based on DSP can measure the axle speed with high precision, and uses fuzzy control as the algorithm to detect the axle sliding state according to multiple slip criteria, and adjust the brake cylinder pressure in time to prevent the axle from continuing to slip, thereby ensuring braking safety, shortening the braking distance, and improving braking efficiency. The simulation test shows that the anti-skid device has the characteristics of fast response, good real-time performance, and high accuracy.