The Study Of Characteristic Model Based Cross Feedback Control On Active Magnetic Bearing

The active magnetic suspension rotor has a good application prospect because of it's advantages of no friction and no lubrication compared with the traditional mechanical bearing.However,when the rotor is rotated at a high speed which could excite the first-order bending mode frequency,its stability is affected by the gyro effect.A classical decoupling control theory--PID-based cross-feedback control is proposed for this purpose.The cross-feedback term is introduced to suppress the gyro-coupling effect.However,it does not solve the weak anti-interference of PID control and problems of Magnetic field coupling affects.Therefore,this paper proposes a new cross-feedback control algorithm: characteristic model based cross-feedback control.Combining the characteristic model algorithm with adaptive function with the cross feedback control that suppresses the gyro coupling effect.Firstly,the sensor,power amplifier and single-degree-of-freedom magnetic bearing-rotor mathematical model are established.The four-degree-of-freedom state space equation including inertial mass and gyro coupling torque is established by five-degree-of-freedom force analysis.And clear the specific parameter values of the state space equation.Then the principle of the characteristic model identification and the full coefficient adaptive control algorithm and related theoretical proofs are introduced,and the "carrier" relationship between the two is explained.In order to illustrate the feasibility of cross-feedback based on feature model,the post-discrete PID control quantity and the characteristic model control quantity are compared and analyzed,and a "bridge" of two control algorithms is established.Then the PID controller is replaced by the feature model controller,and the mathematical expression of the cross-feedback coefficient is calculated by introducing the cross-feedback term in the rotor motion differential equation.Based on the control schematic diagram,a cross-feedback simulation model based on feature model is built in Matlab/Simulink.The step performance of the single-degree-of-freedom and its anti-interference are used to illustrate the floating performance of the rotor under the characteristic mode control.The attenuation factor value of the cross-feedback coefficient is changed to observe the axial trajectory of the rotor near the first-order bending mode frequency,and the control effect of the cross-feedback control based on the feature model is qualitatively illustrated.Finally,the software and hardware design of the controller and the hardware and software design of the speed collection component are completed.The performance of the magnetic bearing system based on the cross-feedback control of the feature model is analyzed from the rotor floating and high-speed rotation experiments.Experiments with four different attenuation factors show that the magnetic bearing-rotor system achieves the best control effect when the attenuation factor is 0.5.Comparing it with the cross-feedback,it was found that the radial vibration vibration peaks decreased by 18.4%,20.5%,10.5%,and 11.3% near the critical speed.