Permanent Magnet Linear Synchronous Motor Servo Control System Based On Self-learning Of PMLSM Parameters

To add a mechanical transmission device to the rotary motor drive, rotary motioncan be converted into linear motion. However, the rotary motor drive mode on linearmovement can’t meet the needs of modern linear servo system which requires highperformance in speed, acceleration and location accuracy. In contrast, linear motordrive mode can acquire linear motion directly without any intermediate mechanicaltransmission device. And in particular, it has a very unique superior performance inthe field of high-precision linear servo control system, which has attracted wideattention of researchers around the world. Compared to linear motor research overseaswhich is currently in the application stage, our research on the linear motor is later. Tobridge the gap between domestic and overseas standard, research on key technologiesof linear drive has very important theoretical significance and application value.In this thesis, for the servo control and the internal parameters of linear motordifficult to botain, we researched and designed the servo control system of permanentmagnet linear synchronous motor (PMLSM) based on parameters self-learning.Self-learning was used to identify automaticlly the parameters of PMLSM. It wasfundamentally different from the self-learning control servo system. And to developea PMLSM servo drive controller prototype based on self-learning of PMLSMparameters.First, research of PMLSM in both domestic and overseas and PMLSM motivatecontrol strategy was described briefly. And then according to analysis of PMLSMstructure and operating principle, a control scheme of PMLSM servo system based onself-learning of PMLSM parameters was designed. PMLSM had two modes ofoperation which were self-learning operation and normal operation in the solution. Inself-learning operation mode, the system could identify the PMLSM parametersautomatically to adjust automatically the parameters of the controllers. In normaloperation mode, the system had a thrust, velocity and displacement of three optionalcontrol ways for the thrust, velocity and displacement control of PMLSM.Secondly, a PMLSM mathematical model was established, and a self-learningmethod of the parameters of PMLSM based on vector control was proposed. Toidentify automaticlly the parameters of PMLSM by giving different voltage space vector. The d axis inductance and q axis inductance and resistance parameters ofPMLSM mover was identified based on steady-state wave of DC current response andits dynamic response characteristic. Permanent magnet flux linkage of PMLSM wasidentified combined with the voltage balance equation when PMLSM moved in a lowsteady speed. And according to the PMLSM movement equation, the quality of moverwas identified based on the rapid sampling of d and q axis current as well as themover’s speed. Then the control structure of PMLSM servo system was designed inwhich PI control algorithm was adopted in current regulator and speed regulator whileBang-Bang/PD control algorithm with variable structure was used in positioncontroller. The parameters of the controllers would be adjusted automatically based onthe identification values of PMLSM parameters from the self-learning operation modeof PMLSM. MATLAB simulation results showed the self-learning algorithm of thePMLSM parameters had high identification accuracy, and the PMLSM servo systemhad a good control result.Thirdly, both hardware and software of PMLSM servo drive controller withparameters self-learning were designed. The former consider the embedded32-bitSTM32F103VET6control chips as the core of hardware system, which consisted ofcontrol circuit, invert main circuit, system power supply circuit, phase currentsampling circuit and bus voltage sampling circuit. The latter was designed accordingto modular design mode which can achieve the core control algorithm of PMLSMservo drive controller by interrupting program with a timer, and the SVPWMmodulation wave generating, signal detection processing, self-learning operation andcontrol algorithm of position, velocity and current regulator, and even the serialcommunication with the host computer was included in the software design process.Finally, an experimental device of PMLSM servo drive controller withparameters self-learning was made, according to the software and hardwaretechnology solutions mentioned above. The experiments on LM110-1N-010AN linearmotor platform made by YOKOGAWA achieved expected results, which verified thefeasibility of the PMLSM servo system designed in the thesis.