Research On Key Technology For Permanent Magnet Linear Synchronous Motor

Several key technology of Permanent Magnet Linear Synchronous Motor(PMLSM) is studied in this thesis.First, according to the Y-vectogram of electromotive force, fractional slot of PMLSM with q<1 structure is studied. The reason why such structure can depress the cogging force is analyzed. The design feature of fractional slot with close slot-pole match is also summarized.Then, to analyze the electromagnetic characteristic of PMLSM accurately,2-D electromagnetic field model of PMLSM is set up by Finite Element Method(FEM), combining with auto-fitting meshing. The cogging force ripple is analyzed in different slot structure quantificationally and the way to reduce cogging force is also studied. The close slot-pole match is proved to be a very effective way to reduce cogging force.Intelligent neural network is used in structure parameter estimation. The estimator based on improved learning algorithm of BP neural network(BPNN) and radical basis functional network (RBFN) are set up, which use the FEM outputs as the inputs of the neural network. The outputs of two estimators are also compared. Results show the estimator based on intelligent neural network has high precision and efficiency. The RBFN based estimator is more effective and higher precision than that based on BPNN. The estimator of structure parameter can provide a useful consideration in structure design of PMLSM.And Then, based on FEM model, current of the fractional slot with q<1 structure is fixed by instantaneous current method. After the iteration of thrust force, the load characteristic and parameters are obtained. Comparing the experiments with FEM analysis, results show the FEM model has high precision and reliability. The obtained load parameters really have some impact in motion control of PMLSM.For the purpose of depressing cogging force further more, PI controller of PMLSM is set up. Experiments show the controller is good at reduce force ripple, which is a simple and effective way in generalization of PMLSM.To solve the problem of PMLSM on position estimation, some common way on position estimation of brushless motor in switching point is introduced including its advantage and disadvantage. Taken the structure and performance characteristic of PMLSM into consideration, G function is presented on position estimation. Through detecting three-phase and two-phase current of prototype, G function is set up based on experiments after simulation. The G function shows high sensitivity in switching point, which doesn't depend on the velocity of PMLSM. That is to say, it can estimate the velocity from Near-Zero to high speed, which is hoped to be used in position sensorless PMLSM.At last, control interface of PMLSM direct drive is developed, which shows simplicity in operation, convenience in use. Then, position accuracy and resetting accuracy are measured in experiments. The velocity of PMLSM under the PI control strategy in different load is also tested. Experiments show that PMLSM has high precision data, which has better application prospect in the field of high-degree accuracy and high velocity.