Delay Compensation Strategy of Networked Control System Based on Neural Network Predictive Control

The transmission delay of information in the network environment degrades the performance of networked control system, so it is necessary to compensate the delay in networked control system. This paper proposes a delay compensation strategy for networked control system based on neural network predictive control. Firstly, the BP neural network is optimized by the Cuckoo search algorithm to establish a delay prediction model to identify the transmission delay. Next, the multi-step delay prediction model is combined with the implicit generalized predictive control algorithm to solve the control increment at future sampling times. Finally, a delay compensation strategy is proposed by setting a buffer in the actuator, selecting appropriate control increments based on the delay situation to improve control accuracy and achieve delay compensation. The simulation comparison of networked gas boiler control system verifies the effectiveness of the proposed networked control system delay compensation strategy. DOI: 10.61416/ceai.v27i1.9287