Research On High-performance Induction Machine Vector Control And Its Operation At High Speed Range

Induction machine(IM)drive systems have been widely used in various areas,such as industrial and agricultural production,transportation,military defense,and daily life.With the development of technology,the applications in CNC machines and electric locomotives set a higher request to the performance of IM drive systems.However,there still exists a large gap in the research of high-performance inverter products between China and the developed countries.Thus,it is of great market value and strategic importance to develop high-performance inverter products with independent intellectual property rights.To improve the performance of IM vector control system,this paper focuses on the following four key techniques:(1)improving the dynamic performance and robustness of current loop;(2)enhancing the anti-disturbance performance of speed loop;(3)output torque maximization in the field-weakening region;(4)speed-sensorless stable operation in the field-weakening region.The specific contents are as followed:The dynamic performance of the conventional PI control-based current loop is limited.To solve this problem,a deadbeat predictive current control(DPCC)is studied in this paper.The command voltage expression of the conventional DPCC is first derived based on the IM discrete model.Then the transfer function from the command current to the actual current is obtained.The theoretical analysis shows that the DPCC is a two-sampling period deadbeat control.However,it is sensitive to the IM parameter variation due to its highly dependent on the IM model,which leads to steady-state current error,even instability.To address this issue,this paper proposes a second-order sliding-mode disturbance observer(SOSMDO)-based DPCC.The system disturbance is estimated by the proposed SOSMDO and then used as the controller feedforward to compensate the steady-state current error,which can effectively enhance the system robustness.By applying the Super Twisting algorithm-based SOSMDO,the observer can provide fast convergence rate,and the inherent chattering phenomenon in the conventional sliding-mode control is suppressed.The experimental results show that the proposed approach can improve the dynamic response and the robustness of the current loop.Although the SOSMDO-based DPCC can achieve ideal current control,it is complicated and cannot be analyzed by linear control theory.In terms of this issue,this paper develops a robust predictive current control(RPCC).According to the IM discrete model,a Luenberger observer is constructed to estimate the next moment stator current.On this basis,the command voltage expression of the conventional RPCC is derived.The theoretical analysis indicates that the RPCC can improve the system robustness at the expense of decreasing the cutoff frequency.However,the problem of steady-state current error still exists.For this problem,an extended state observer-based RPCC is proposed in this paper.The next moment stator current and system disturbance are estimated by the extended state observer simultaneously.By introducing the estimated stator current and system disturbance into the command voltage,the steady-state current error is essentially eliminated.The experimental results verify the effectiveness of the proposed approach.To enhance the anti-disturbance performance of the speed loop,this paper studies a torque feedforward compensation-based compound controller.The transfer function of PI control-based speed loop is derived.The frequency-domain analysis shows that the single PI controller cannot balance rapidity,stability and antiinterference of the speed loop.To solve this issue,this paper proposed a second-order fast terminal sliding-mode load torque observer.According to the torque feedforward compensation theory,the estimated load torque is used as the controller feedforward compensation.The stability of the proposed observer is demonstrated by Lyapunov stability theorem.Then the observer finite-time convergence and chattering elimination are theoretically analyzed.At last,the experimental results prove that the proposed approach can effectively enhance the anti-disturbance performance of the speed loop.To achieve maximum output torque for speed-sensorless IM drives in the field-weakening region,this paper investigates a field-weakening control based on a speed adaptive observer.First,according to the analysis of the system voltage and current constraints,the optimal current vector trajectory is derived for the output torque maximization in the field-weakening region.Then a robust field-weakening control is proposed,where a voltage loop is adopted to increase the utilization of dc-link voltage,leading to maximum output torque in the field-weakening region.Second,a full-order flux observer is designed based on the IM model in the two-phase static coordinate,and then the speed adaptive law is derived according to the Lyapunov stability theorem.However,the theoretical analysis shows that there exists instability problem in the conventional forward Euler discretization-based observer.For this issue,this paper proposes a modified Euler discretization,which can balance the observer stability,accuracy and computational burden.Finally,the experimental results verify the correctness of the proposed approach.