Research On Model Predictive Fault-tolerant Control Of Five-phase Interior Permanent-magnet Synchronous Motor

The interior permanent-magnet synchronous motor(IPMSM)has been widely used in a variety of industrial applications due to its high efficiency,high power density and wide speed operating range.With the increased safety and reliability requirements in these areas,the fault-tolerant strategies for multiphase drives have attracted more and more attentions.In this paper,the five-phase IPMSM is taken for the research object,and the complex mathematical model of the motor and the working principle of the inverter are analyzed under the open-phase fault condition.A fault-tolerant strategy based on model predictive control(MPC)is proposed to improve the performance of the motor under single-phase fault,and realize the fault-tolerant control of double-phase fault.Furthermore,considering the increase of loss caused by current reconstruction and the particularity of five-phase IPMSM under the fault-tolerant condition,the implementation condition of the maximum torque per ampere(MTPA)control is analyzed,then the output torque and efficiency of systems can be improved effectively.The main research contents of this paper are as follows:(1)For the condition of single-phase and double-phase faults of five-phase IPMSM,the optimal solutions of fault-tolerant currents are calculated in order to maintain constant magnetomotive force as same as that in normal operation.The orthogonal reduced-order transformation matrices are obtained from the fault-tolerant currents.Then,the mathematical models of five-phase IPMSM under open-circuit faults are described,which lays the theoretical foundation for the MPC fault-tolerant control in the following paper.(2)The working principle of five-phase voltage source inverter under fault condition is analyzed with the orthogonal reduced-order transformation matrices;The voltage vectors are reconstructed and a z subspace is proposed to reduce the total current harmonic distortions under single-phase fault.And the distributions of space voltage vectors under double-phase faults are established to realize the fault-tolerant operation.Meanwhile,the predictive models can be obtained from the mathematical models by applying the Euler formula.Finally,the cost functions are derived from the degrees of freedom for fault-tolerant control;(3)According to the principle of virtual-signal-injection-based MTPA,the injection precision of virtual signal is improved,and the implementation conditions of virtual signal is obtained from the in-depth analysis of the working state of inverter under fault condition to realize the fault-tolerant MTPA control,which can reduce the amplitudes of phase currents and improve the output torque and efficiency of system effectively.(4)The experimental platform of prototype driving system based on d SPACE is designed.The development process of d SPACE and the main parts of its software and hardware are introduced.Then,the proposed fault-tolerant control in this paper are verified by experiments.