Machine learning-based predictive control of nonlinear time-delay systems: Closed-loop stability and input delay compensation

The purpose of this work is to study machine-learning-based model predictive control nonlinear systems with time-delays. proposed approach involves initially building a machine learning (i.e., Long Short Term Memory (LSTM)) capture the process dynamics in absence time delays. Then, an LSTM-based controller (MPC) designed stabilize system without Closed-loop stability results are then presented, establishing robustness MPC towards small time-delays states. To handle input delays, we design predictor that compensates for effect used predict future states using measurement, and predicted initialize MPC. Stabilization time-delay both state delays around steady achieved through featured design. applied chemical example, its performance properties evaluated via simulations.