Study On Independent Robust Control Of The Levitation Subsystem Of Bearingless Induction Motor

With the combined action of the magnetic field by the torque winding and the levitation force winding, bearingless induction motor(BIM) can realize the stable suspension of squirrel cage rotor,which is very suitable for high speed precision CNC machine tools and other high-tech fields, and has high research and application value. However, there is a complex electromagnetic relationship between multiple sets of windings in BIM, which makes decoupling control a key problem in the realization of stable suspension and torque output for BIM. Therefore, under the support of National Natural Science Foundation of China(61174055), with two degree of freedom BIM as the research object, a independent robust control method and a improved flux linkage identification method are proposed in this dissertation aiming at the defects of the traditional BIM suspension subsystem independent control method of the pure integrator problem of U-I model identification and poor robust performance. The main contents are as follows:In the theoretical basis of independent control of levitation system, a independent robust control method is proposed in order to improve the anti-interference ability of the system and enhance its robustness. Firstly nonlinear H∞-infinity control theory and Hamilton-Jaeobi-Issaes(HJI) inequality based principle are briefly introduced, and according to the bearingless induction motor system model, a robust controller based on HJI inequality principle of RBFNN is established to replace the traditional PID controller. Secondly, the PID current hysteresis loop is added in the suspension winding side control loop to speed up the control of the reaction. Thirdly an air gap flux observation is used to identify air gap flux value, so as to make the high performance independent control possible. Then the torque subsystem can be controlled by general inverter control to enhance the practicability of BIM.Finally, the control performance of the method is verified by simulation experiment.Aiming at the pure integral problem of U-I flux linkage identification model, a new type of air gap flux linkage identification method is proposed using the strong self learning and adaptive ability of the radial basis function neural network(RBFNN). Firstly, the air gap field oriented control of BIM is carried out, and the air gap flux data is obtained through the air gap flux linkage observer with high pass filter and coordinate transformation. Meanwhile the air gap flux is recorded in each state. Secondly through the analysis and processing of discrete data, several groups of stable data is selected to train the RBFNN according to the steps to meet the requirements of training to error. When training is completedthe RBF neural network can be used as the identifiers of the air gap flux. Finally, the performance of the BIM independent control system is analyzed by simulation.Finally, a digital control platform is designed and manufactured based on the TMS320F2812 as the digital controller subject to control the BIM prototype in the laboratory as the research object. The hardware platform mainly includes the main circuit, control circuit, protection circuit and so on. At last,relevant theory of the verification experiment is completed by compiling software program based on the hardware platform.