Discrete Command Filtering Fault-tolerant Control Of Asynchronous Motors Based On Dimension-reduced Observer

Induction motors(IM)are widely used in servo drive systems,modern industrial and agricultural production in virtue of its reliable performance,efficient operation,ruggedness and low cost.However,the driver systems of IM are the nonlinear multivariable objects with strong coupling,unknown load disturbances and uncertain parameters,the traditional control methods such as vector control,direct torque control,and others are insufficient to achieve good dynamic and steady performance.And,with the rapid development of digital computers and microprocessors,the existing industrial production control algorithms are mostly based on sampling control via digital controllers.Compared with continuous-time control methods,the discrete-time control methods are easier to describe the actual control procedure.In addition,the harsh operating conditions,load disturbances and grid interferences in the IM system may damage the power transistors in the inverter,resulting in output voltage distortion,affecting the performance of the motor and even causing serious damage to equipment and personal safety.Therefore,it becomes an urgent problem to study a fault-tolerant control method for IM discrete-time systems that can guarantee the expected position tracking performance even after the fault occurs.This dissertation combines command filtered backstepping,discrete-time control,neural networks and reduced-order observer techniques to propose a discrete-time faulttolerant control method for IM system considering voltage dip and bias faults.The main contributions of the designed method are as follows:(1)For a class of single-input single-output nonlinear discrete-time systems,based on the command filtered technique and neural network control,a discrete-time fault-tolerant controller considering actuator faults is constructed.Firstly,the command filtered backstepping and error compensation mechanism are combined to eliminate the influence of filtering errors caused by the command filters while avoiding the noncausal problem and achieve better control accuracy.Finally,the Lyapunov function is constructed and the stability of the system is analyzed.(2)A discrete-time dynamic surface fault-tolerant control method is designed for the voltage dip and bias faults caused by the open circuit of inverter power transistors in the IM system.First,the Euler formula is used to establish a discrete-time model of the IM system considering the voltage dip and bias faults.Second,the dynamic surface technique is used to overcome the noncausal problem and “complexity of computation” resulting from frequent differencing of virtual control signals.Then,the neural networks technique is used to approximate the unknown function with voltage dip coefficient and bias fault function.Simulation experiments demonstrate that the designed discrete-time dynamic surface faulttolerant controller can still achieve the expected position tracking performance well after the occurrence of faults,and suppress the adverse effects of voltage dip and bias faults on the control performance.(3)A command filtered fault-tolerant controller based on the reduced-order observer is proposed,considering the sensor accuracy degradation caused by vibration or humidity and the speed sensor cannot be installed due to cost constraints.The adaptive observer is designed to estimate the rotor angular velocity of the IM system,which saves the cost of the system and has more extensive engineering application value.Simulation results show that compared with the dynamic surface fault-tolerant control method,the proposed command filtered fault-tolerant control method has better robustness and better accuracy by virtue of error compensation mechanism,which can more effectively restrain the influence of voltage dip,bias fault and load disturbances.