Interval Adaptive Predictive Control of Nonlinear Systems Based on LSTM Network
Shouping Guan, Qixin Zhao, Xinzong Wang · 2023
This paper presents a formulation of interval control for nonlinear systems based on Long Short-Term Memory network (LSTM) and using predictive control to solve the nonlinear interval control problem. The gradient descent with momentum is used as the learning algorithm to train the LSTM, which is used as the nonlinear system prediction model. The set interval is introduced into the optimized performance index and the gradient descent with momentum is applied to obtain the control variable in the performance index to realize the interval control, and the LSTM online learning algorithm is used to replace the feedback correction to realize the adaptive control. Furthermore, the system stability for the proposed formulation is proved based on Lyapunov theorem. The effectiveness of the proposed algorithm is proved by simulation and comparison experiments.