具有神经元网络补偿的隐式广义预测控制的硬件设计与分析

针对隐式广义预测控制算法在被控系统复杂性增加时有效性降低、实时性变差的问题，本文将神经元网络作为该算法的补偿，提出了一种具有神经元网络补偿的隐式广义预测控制方法，并着重讨论了神经元网络的硬件设计和ＮＩＧＰＣ的实现。仿真结果表明：ＮＩＧＰＣ不仅保留了ＩＧＰＣ鲁棒性强等特点，更主要的是它提高了ＩＧＰＣ的的实时性，并使之具有良好的跟踪特性。

HARDWARE DESIGN AND ANALYSIS ON THE IMPLICIT GENERALIZED PREDICTIVE CONTROL WITH NEURAL NETWORK COMPENSATION

In order to improve the effectiveness and real-time property of the implicitgeneralized predictive control algorithm (IGPC) with the adding of the controlled system'scomplexity, the implicit generalized control with neural network compensation (NIGPC)is presented. The hardware design of neural network and implementation of NIGPC is emphatically discussed.The simulation results demonstrate that the NIGPC not only has strong robustnesslike IGPC, but chiefly improves the real-time property of IGPC, has better trace performance.