Technical introduction and development analysis of automobile active collision avoidance system

At present, there are three types of active collision avoidance systems for research and development:

This article refers to the address: http://

(1) Vehicle active collision avoidance alarm CWS (collisionwarning) system, this system gives an alarm to the detected hazard situation, the United States has been commercialized on some heavy-duty trucks and buses.

(2) Vehicle adaptive cruise control ACC (adaptivecruisecontrol) system, this system can achieve active collision avoidance and cruise control under simple traffic conditions, some car companies have begun to adopt ACC technology on high-end models.

(3) Compound vehicle intelligent control system, which is based on complex traffic conditions, especially urban traffic environment, adopts ACC system and vehicle stop (go&go) system to improve the practicality of vehicle intelligent control.

At present, most of the researches on active collision avoidance systems at home and abroad focus on the longitudinal control of collision avoidance systems. The key technologies of the system include target vehicle identification and motion information acquisition, safety distance model and collision avoidance control system modeling in the driving environment.

Analysis on the development of automobile active collision avoidance system technology

Driving environment identification

The key to driving environment identification is the measurement of distance. Current techniques for distance measurement include ultrasonic measurement, infrared ranging, laser measurement, machine vision, and radar technology. The real-time and accurate radar measurement is better. The millimeter-wave vehicle radar system currently working in the millimeter wave band is considered to be a better solution to the problem of controlling detection under the active collision avoidance condition of the automobile. Many companies' ranging technology has come into practical use. The collision avoidance alarm system being researched by GM is based on Lidar technology; the MEAR laser ranging system developed by Wilder Blackbrück in Germany provides the driver with distance and relative to other vehicles and obstacles. Speed ​​data, while providing video images; Mercedes-Benz, Nissan and other car companies are using the millimeter-wave radar system produced by the German ADC company.

Security status judgment

Safety state judgment uses safety distance logic algorithm to establish different safety distance models based on factors such as kinematics analysis of vehicle braking process, ride comfort, workshop time interval and driver characteristics. Among them, the safety distance model based on the workshop time interval does not consider the factor that the driver actually maintains the workshop distance and the relative speed, and the judgment result does not conform to the driver's subjective feeling. Therefore, Japanese researchers have established a driver's preview safety distance model. However, under the typical working conditions of the collision avoidance system of the front vehicle brake, the acceleration of the model is fixed, and the judgment result of the model does not conform to the subjective feeling of the driver. Aiming at this problem, the driver's optimal preview acceleration model is studied.

Collision avoidance system control

The active collision avoidance control mode of the vehicle mainly includes upper control and lower position control. The former starts from a safe distance, and applies a control algorithm from the perspective of kinematics to obtain the deceleration that the vehicle should have in the current situation; the latter starts from the target parameters such as the vehicle target deceleration determined by the upper control algorithm, combined with the vehicle braking system model. The application of the control algorithm to achieve precise control of the throttle, brake, steering, etc., to achieve the goal of superior control requirements. The control algorithm has fuzzy control and neural network control, and is suitable for the sliding mode control algorithm and Back-stepping method of nonlinear targets.

Problems with active collision avoidance systems

The current research considers that the side targets and other types of distance targets are very few, and the system has problems such as redundant alarms, false alarms, and insufficient consideration of driver characteristics. Only the longitudinal control model is considered in the control model. For dangerous conditions such as high-speed steering braking and overtaking of vehicles, a dynamic model considering lateral control must be established. In the controller design of the model, the existing host controller design considers the driver's driving characteristics to be insufficient, and the lower controller is less robust and cannot meet the needs of the collision avoidance system, which needs further research and solution.