In order to reduce the tension fluctuation of long segment warp and improve the tension control performance,A neural nonsingular integral fast terminal sliding model control(RBF-NIFTSMC)method was proposed.Through the force analysis of the let off system,the time-varying tension control model of the loom was first established.According to the characteristics of the system,such as time-varying mathematical model,unmodeled dynamics and unknown external interference,a nonsingular integral fast terminal synovial control method is designed.The controller can reach the sliding mode surface in a short time without singular problems.The unknown uncertainty is adaptively estimated by using NIFTSMC theory,and the stability was then proved by Lyapunov theory.In order to further improve the sliding mode control performance,the sliding mode surface and variable step size neural network were fused to realize the time-varying sliding mode surface.The parameters of the neural network were optimized by the improved Aquila optimization algorithm.Finally,the simulation results show the effectiveness of the proposed control method.
关键词
张力/送经系统/滑模控制/神经网络/天鹰优化算法
Key words
tension/let off system/sliding mode control/neural network/aquila optimization algorithm
维普
万方数据