摘要
用PID控制器控制有共轭极点和负实零点的二阶模型对象,当对象时滞比很小时,用现有准则如ITAE进行优化,很难保证幅值稳定裕度。在广义平方误差积分准则(GISE)的基础上,用对象响应特征时间来平衡准则中误差项与误差变化率项的数量级,并用一个恒大于1的时间函数作为误差项的权,从而提出了一种新的误差积分准则—改进的广义平方误差积分准则(RGISE)。仿真结果表明,新准则能够获得较大的幅值稳定裕度和较好的响应曲线。
In PID control system, when the plant model is second order with negative zero and the delay is very small, ITAE and GISE can not guarantee that the gain margin of the optimized system is large enough for engineering practice. To solve this problem, a new performance criteria called refined generalized integral of square error (RGISE) is proposed. It is based on GISE using a plant eigen time to balance error item and error changing rate item and a time function starting with 1 as the arbitrator of error item. Simulation shows that RGISE can guarantee a considerable large gain margin with acceptable dynamic response curve.
出处
《控制工程》
CSCD
2004年第1期52-54,共3页
Control Engineering of China
基金
国家自然科学基金资助项目(60174021)
国家科技攻关计划资助项目(2001BA204B01-02)
关键词
误差积分准则
PID控制器
优化
PID控制
参数整定
数学模型
PID controller
PID parameter optimization
integral performance criteria
normalized dead time
gain margin
