减少测量误差影响的误差检测–K值控制法

本方法检测单闭环系统或弱耦合的多闭环多控制对象的复合控制系统的误差信号E(s)和输出值C(s),得到输出期望值R(s)；再加上K倍E(s)作为新的输出期望值R*(s),然后使用原算法对系统实施控制.比较传统的控制方式,通过检测系统的输出期望值R(s)和输出值C(s)得到误差信号E(s)从而实施控制,该方法减少了测量误差对系统误差的影响.同时,该方法可对多控制对象中的每一个控制对象的开环增益通过系数K进行单独调节,从而便于实现每个控制对象开环增益的最优化,减少了控制系统的调试难度,降低了系统的稳态误差.论文推导了测量误差对原控制系统的影响及该方法减少测量误差对系统误差影响的原理,并提供了该方法在有源滤波器中的应用实例,仿真和实验结果验证了该方法的正确性.

Error-detection-K-control for reducing effects from measurement errors

In this approach, the error signal E(s) and the output signal C(s) of a single closed-loop system or a multiple closed-loop system, and a compound system of multiple-object and multiple-control with weak coupling are detected to obtain the expected output signal R(s), and then we replace R(s) with the new expected output R*(s) which equals R(s) plus K times E(s) and perform the control with original algorithm. Comparing this control with the traditional control in which we obtain E(s) by detecting R(s) and C(s), we find the effect from measurement errors on system errors are reduced. Meanwhile, it makes possible to change the open-loop gain of individual controlled object by adjusting the coefficient K to achieve the optimization for the open-loop gain of individual controlled object, thus cutting down the adjusting and testing work and reducing the steady-state errors of the system. The effect from the measurement error on the original control system is derived analytically, and the principle of lowering effect from measurement errors is explained. An application of this approach to active filters is investigated, and the efficacy of this scheme is validated by simulation and experimental results.