Direct Adaptive Control of Process Control Benchmark Using Dynamic Recurrent Neural Network

Abstract

In this article, we develop a direct adaptive control scheme based on Dynamic Recurrent Neural Network (DRNN) for a process control benchmark. The DRNN is represented in a general nonlinear state space form for producing the control action that force the system output to a desired trajectory. The control algorithm can be implemented without a priori knowledge of the controlled system. Indeed, the weights of the DRNN controller are adjusted on-line using the truncated Back Propagation Through Time (BPTT) method. Unlike the approaches in the literature, the learning signal of the network weights is generated by a control error estimator stage in the developed controller. Finally, the developed controller is applied to a laboratory flow control system with two experimental scenarios.