Modeling and Adaptive Tension Control of Chain Transmission System with Variable Stiffness and Random Load

In the fully mechanized mining face, the proper chain tension of chain transmission system is the key to reduce the failure rate of scraper conveyor and improve the service life of expensive chain. However, the interference of random load and variable stiffness in the chain transmission system of the scraper conveyor make it very difficult for tensionable tail drive (TTD) to control the chain tension. In this research, the mathematical model of the TTD based on catenary theory is established, and the dynamic characteristics are analyzed to investigate the influence of load and stiffness on the tension control. An adaptive neural command filtering backstepping algorithm with parameter identification is proposed to control the chain tension of the chain transmission system. The combination of stiffness identification with the linearized tension model weakens the influence of variable stiffness and load on the tension control. Based on state observation and stiffness estimation, radial basis function neural network is introduced into the tension controller to compensate the load randomness caused by tail sprocket drive. Comparative experimental results are completed to verify the high-accuracy tension tracking performance and a strong robustness of the proposed controller in the random load mode.