Research On Induction Motor Speed Sensorless Control System Based On Fractional Sliding Mode Theory

Induction motors(IM)have been widely used in transportation,industrial manufacturing and other fields due to their simple structure,stable performance and low cost.IM control systems are attracting extensive attention in research and application.IM vector control(VC)systerm based on the field oriented has brought perfect dynamic and static performance.The speed sensorless VC of IM simplifies the system structure,reduces the cost and improves the reliability further.However,IM is a strongly coupled nonlinear system,and its control performance is affected by the variation of motor parameters,load disturbance and other factors.Therefore,new speed sensorless IM control strategy has become a hot research topic.Fractional calculus has the characteristics of flexible control and strong robustness to parameter changes.Meanwhile,sliding mode control(SMC)has the distinguishing features of fast response,strong robustness to parameter changes and disturbances,simple structure and easy implementation.Fractional order sliding mode control(FOSMC)combines the advantages of these two control methods to further improve the performance of traditional SMC and become a new branch in the field of SMC.Based on the intensive study of fractional calculus,SMC and FOSMC theory,a flux fractional order sliding mode observer(FOSMO)and speed estimation method is proposed.At the same time,the dynamic fractional order sliding mode(DFOSMC)speed controller with load observation is also designed.Finally,the designed observers and controllers are applied to the control of mine traction motor for mine locomotive with complex and harsh running environment,which effectively improves the control performance and reliability of the system.The main contents of this thesis include:(1)Based on the analysis of the traditional current observer of flux and voltage observer of flux,a hybrid observer of flux based on the switching of speed correlation function is proposed.The smoothly switching between the two observer is realized according to the change of speed,which has better observing precision in the full speed range.However,the hybrid observer of flux belongs to open-loop,and the switching speed value needs to be determined during the observation process,which not only affects the observation performance,but also has poor versatility in practical applications.For further study,rotor flux FOSMO with improved reaching law is proposed in the thesis,which combines fractional order integral theory with SMC theory.FOSMO not only guarantees the global robustness of the system,but also suppresses chattering in sliding mode effectively.By analyzing the relationship between the reaching law and the sliding mode chattering,the flux FOSMO with different reaching laws are designed and their stability are also proved.The simulation results show that the rotor flux FOSMO based on the faln function reaching law suppresses the chattering of the observer obviously and improves the rotor flux observing accuracy,which lays a foundation for high-performance VC and speed estimation.(2)Speed estimation is the core part of motor speed sensorless VC.In this thesis,the motor speed estimation is carried out by means of the flux FOSMO,and the speed sensorless VC of the IM is realized.The simulation analysis of the IM running under different working conditions is carried out,and the effectiveness of speedless VC based on FOSMO is verified experimentally.(3)Speed controller is the key part that affects the performance of IM speedless VC system.Integer-order proportional-integral(PI)control is linear in nature and has poor adaptability to control system parameters.In this thesis,the fractional calculus and FOSMC theory are applied to the speed controller of the IM sensorless VC system.The intelligent fractional order PI~?(IPI~?)speed controller is designed to improve the control performance of the system effectively.And then,in order to further improve system control performance,a DFOSMC speed controller with load observation is proposed.Compared with integer order integral sliding mode(IOSMC)controller and FOSMC controller,the simulation results show that the DFOSMC speed controller with load observation not only has stable output torque,but also has better dynamic and static performance and robustness.And the feasibility of DFOSMC is verified by experiments.(4)In order to verify the engineering practicability of this research,the research results of this thesis are applied to the traction motor control of mining locomotive,speed sensorless VC is applied to the mine traction motor control field.And the fractional order PI~?and the DFOSMC algorithm with load observation are applied to the speed controller respectively.The performance of the mining locomotive under different working conditions is analyzed by simulation.Furthermore the feasibility of the speedless VC based on PI~?control for mining traction motor is verified by experiments.