摘要
研究了基于PID神经元网络的智能车多变量控制系统。智能车的转向控制与速度控制相互关联、相互影响、且都具有时变性,针对智能车在行驶时要求电机的动态响应速度要快、舵机的动态响应时间要短的特点,提出了将PID神经元网络(PIDNN)控制器及其算法应用到智能车的控制系统中来对传统PID控制进行改进。PIDNN控制系统不依赖智能车电机与舵机的数学模型,能够根据控制效果在线训练和学习,调整网络连接权重值,最终使系统的目标函数达到最小来实现智能车的精确控制。Matlab仿真测试表明,PIDNN控制系统的响应快,超调小、无静差,与传统PID控制算法相比,大大提高了智能车控制系统的性能。
Research on intelligent vehicle multi variable control system based on PID neural network. The steering control and speed control of interrelated connected to each other, and influence each other, and has time-varying , for smart car when driving mo-tor faster dynamic response, the steering gear of the characteristics of dynamic response time is shorter, the PID neural network （PIDNN） controller and its algorithm are put forward to improve the classic PID control. PIDNN control system does not rely on the mathematical model of intelligent car motor and steering gear, can according to the control effect of online training and learning, and adjust the network connection weights, eventually make the system objective function to minimize to realize accurate control of the smart car. Matlab simulation tests show that the PIDNN control system fast response, small overshoot, no steady-state error, com-pared with the traditional PID control algorithm, greatly improves the performance of the intelligent vehicle control system.
出处
《自动化与仪器仪表》
2014年第1期6-9,共4页
Automation & Instrumentation
基金
国家自然科学基金资助项目(51165024)
关键词
PID神经元网络
智能车
控制系统
PID neural network
Intelligent vehicle
Speed control system
