Realization of Automatic Control System Based on Artificial Intelligence

摘要

Artificial intelligence is also called machine intelligence. It is a comprehensive discipline developed by the penetration of multiple disciplines. Based on artificial intelligence, automatic control science has also gradually emerged and developed, and will become an inevitable trend in the development of industrial control technology. However, NCS itself has problems such as network induced delay, disordered data packet timing, and packet loss. This paper studies two artificial intelligence control algorithms, namely Smith predictive fuzzy PID controller and generalized predictive control based on dynamic BP network error correction, and applies it to network automation control system to reduce the impact of delay on system control performance: Finally An automatic control system for ink color presetting of color newspapers is designed, and fuzzy PID control algorithm is applied to the system to realize the control of the ink output. Aiming at the non-stationarity of the signal to be cancelled, based on the research of traditional nonlinear filtering, we design an artificial neural network as an adaptive processor, relying on its powerful functions similar to the human brain to solve the inherent defects of the conventional filtering mode. The introduction of neural networks has greatly improved the ability of the system to solve complex and highly nonlinear problems. The model designed in this paper has laid a good foundation for solving nonlinear system problems, and the algorithm has a moderate degree of complexity, which has high application value. Experimental research shows that the method is verified by MATILAB experimental simulation to improve the stability of the system and make the system have good dynamic performance and robustness.