Basic Theroy Of Embeded Control System For Active Magnetic Bearing And Research Of Key Technology

Active magnetic bearing (AMB) makes motor rotor suspend in space without any touch by magnetic field, so it is a new type of mechanical bearing without any frictional force between rotor and stator. This paper summarizes the recent AMB control technology from home and abroad, sets up mechanics model and electricity model of AMB system, designs and achieves full digital control system including hardware and software. This paper carries out deeply analysis of working course of AMB control system. According to special needs and actual condition of AMB system, the system structure of hardware and software will be adjusted and conformed to obtain better control performance and optimal system stability by control algorithm and embedded technology. It emphasizes to robust ability and optimizes testing experiments with normal algorithm and improved ways.It is studied that corresponsive position control algorithm combining with nonlinear characteristic of AMB control system, such as change-step PID algorithm, non linear back stepping to the control of magnetic, single-nerve cell PID self adaptive controller.It is realized full digital control system of AMB by hardware and software and set up full digital PWM amplifier by constant current control algorithm, so accomplish 10 ways power output by PWM signal source of DSP. By hardware and software platform, corresponsive control theory of AMB system is deeply researched comparing with improved ways, and get better technique project.By AMB hardware platform of full digital controller, different control software ways are adopted: former-back software design, main flow is given; programmed idea is inducted by C/C++ software. Based on uC/OSII RTOS(real time operation system), every step is achieved such as control module, task PRI assign, real time attempter, double communication disposal and fault diagnosis.On the test platform of AMB hardware system, it is obtained satisfied result in one-freedom and five-freedom experiments by above research and analysis.