单神经元PID控制算法在智能车控制系统的应用

Application of Single Neuron PID Control Algorithm for Intelligent Cars

针对传统PID控制算法对智能车的舵机和电机进行控制时,其控制参数很难随环境变化进行自整定,进而影响智能车的运行状态的问题,提出单神经元PID控制算法,该算法具有参数自整定能力强、现场调整参数少、调整周期短及稳定性好等优点。通过对智能车的系统结构和运动学模型进行分析,根据单神经元PID算法的控制策略,进行多种算法的仿真测试并将其应用在实际的智能车控制试验中。结果表明:单神经元PID控制算法能够使控制系统以较好的状态运行,实现智能车快速稳定的自主循迹行驶。

Aiming at the traditional PID control's difficulty in the self-tuning of control parameters which changing with the variance of circumstance,and its influence on running status of the intelligent car,the single neuron PID control algorithm was proposed. This control algorithm has strong self-tuning capability,fewer adjustable parameters,shorter regulation period and better stability. Analyzing both system structure and kinematics model of the intelligent car,simulating and testing several algorithms based on single neuron PID control scheme,as well as applying them to the control test show that,the single neuron PID control algorithm can keep control system running in best status and achieve fast and stable auto-tracking of the intelligent car.