The Control Method Of The Active Rotors System Supported On Magnetic Bearings

In this paper, the control system of the magnetic bearing is researched. The model of control system is founded from the disturb standpoint, and the characteristic of control system is analyzed. The theory of neural network is applied in the unit of single freedom magnetic bearing, the system of digital control is established around the core of the digital signal processor.Firstly, the principle of the magnetic bearing is recommended briefly, the magnetic force is analyzed and calculated. The feasibility of the closed loop control of the rotor system supported on magnetic bearings is demonstrated. The control object- rotor is not the response of the input single, is the restrain of the disturbance input because that the control system of the rotor system supported on magnetic bearings is different from the general system, and does not exist the input of the importation single and the input of the track single, the transfer function of the system is founded with the disturbance single as the input single.Secondly, the control of the neural network PID using in the rotor system supported on magnetic bearings is researched and simulated. The weigh-value of network is optimized using the method of grads-descending. This paper analysis the influence of the disturbance single upon the capability of the rotor system supported on magnetic bearings, and compares the difference of the control of general PID and the control of neural networkPID.Thirdly, the digital control system of the rotor system supported on magnetic bearings based on DSP is designed. The hardware structure of the control system is researched. The basic game of the control system of neural network of the magnetic bearing system based on DSP is planed.