On-Line Adjustment Control Of Two Motor Synchronization System Based On Neural Network Inverse

The paper focuses on the multi-variable decoupling control of synchronization system composed of two AC induction motors and transducers-decoupling the synchronically modulating velocity system into velocity and tension subsystems by the application of ANN inverse system method.Moreover,a strategy of the connection value amended on-line is proposed.On the basis of theoretic analysis of theα-th ANN inverse system,the method are proposed to fabricate theα-th ANN inverse system of original system. And then theα-th ANN inverse system of the two AC induction motors and the transducers constructed.Cascading the ANN inverse system with the two-motor synchronous system,a pseudolinear system is completed.So the multivariable, nonlinear and coupling two-motor system is decoupled into two independent linear subsystems-speed subsystem and tension subsystem,and then a linear close-loop adjustor is designed to control each of the subsystems,and the method of the connection value amended on-line and simulation results are given.Secondly,the generalized reversibility of this two-motor system was testified.On the basis of theoretic invertibility analysis of that system,the methods are proposed to fabricate the neural network generalized inverse system of original system.The network gets the initial centers through a k-means algorithm,and gets the weights through recursive least squares(RLS) in off-line training.In on-line training,the network updates the network parameters using an improved BP algorithm.The neural network generalized inverse system composed of two AC induction motors and the transducers is constructed.The digital control arithmetic programmed according to the ANN and the synthesis controller of pseudolinear integrated system in PLC.Finally,the controlling experiments are performed on the platform of two-motor synchronous system.Experimental results show that the decoupling control based on ANN inversion system which connection value can be amended on-line decouples the system into the velocity and the tension subsystem completely.The ANN can be adjusted in the control process.The system has good static and dynamic decoupling performances.The good applicability of ANN along with the strong stability and robustness of the system can be achieved by using the proposed method.