Research Of Switched Reluctance Motor Drives Modeling And Control

In the last decade, the switched reluctance motor (SRM) is rapidly gaining importance application in variable-speed drive systems. Compared with conventional motors, The SRM has many individual features. This dissertation is focused on the modeling and novel control strategies for SRM.An important characteristic of the SRM drive is its inherent nonlinearity, which makes it very difficult to derive a analytic mathematical model for SRM. Based on the measurement of the flux-linkage and torque characteristics of SRM, a fuzzy-neural network (FNN) model of SRM was presented, and then a novel procedure for the general simulation of switched reluctance motor drivers (SRD). Computed results are verified by experiment results.Because of the nonlinearity of SRM and its power converter, the optimization of control parameters is a very complex problem. This paper derives the nature characteristic of optimum switch-angle strategy of SRM from the energy conversion of SRM. Using the simulation method of SRD described, some results about the optimum switch angles have been drawn. According to the optimum switch angles control strategy, a SRD system based on digital signal processor is designed. Practical operation experiments are performed on a 3Kw SRM with 8 stator poles and 6 rotator poles. Results have shown that the system is a novel variable-speed electrical drive with simple and excellent performances.There are many implementation problems exist in the conventional SRM position estimation schemes. Hence, in this paper, a new SRM rotor position estimator based on simplified flux method is described, which presents a "fuzzy" position estimation and an online optimum switch-angle control for SRD. The experimental results shows that the sensorless position estimation strategy is simplified, practical and high efficient.The torque ripple in SRM are relatively higher compared to induction machines. In this paper, based on the experimental data of static torque characteristic, a fuzzy-neural network is applied to learning it's inverse model off line, then according to the predefined torque distribute function, optimal current profile is real-time gained by the FNN on line, which results in a linear, decoupled, low ripple control of torque. The performances of the proposed method is demonstrated by computer simulation results.For improve the adaptability and robustness of SRM torque control, based on a novel SRM torque estimation method, a closed-loop torque control for SRM has been presented, the simulation results have shown that torque ripple minimization can be achieved.