Magnetic bearing with compensator continuous slidi

px477801

Compensator with a continuous sliding mode control of magnetic bearings

. 3 A numerical example consider the uniform homogeneous magnetic rotor axis cross section , length = 1_2 alum diameter d = 20 in Figure 2. The finite element method . 『 disk is divided into 12 units , 13 nodes, 26 degrees of freedom, the simulation data shown in Table 1 . System has four controller , were placed in the nodes 3,59,11 . Placed in the same position displacement and velocity sensors. The modal truncation method, is 4 -order modal analysis. Taipa quiet one by one T ---? ... ... Figure 2, unit 3 rotor unit nodes ● Figure 5 of Lin Tong et al: flutter compensator with continuous sliding mode control of magnetic bearings . 63. Table 1 Magnetic parameters of the project bearing numerical permeability of vacuum vo / Hm ' air-gap cross-sectional area S / nP air gap length of the ho / m per unit length. / M coil winding turns N / turn bias current D / A per unit length of the diameter ddm / m × 10978 × l005 × 1001l000.50.021.2 / s Figure 3, the control axis position control system design parameters are as follows : Q = cling [1.1,1,1,1 o6, 10,106,1 o6, 10], R = diag [103,1 o3.103. 1 Shanghai ], P1 = [10,10,10,103, P2 = dlag 【 1,1.1,1], P = 【 Pl'P2], D1 = [0,0,0.0], D2 = diag [10.1,1,102], D = [D?, Dj], L12 = [ a 10 , a 1o, a 10, a 10j, L22 = diag 【 a 10, a 103. A 10. A 10]. = diag [0.1.0.1,0.1,0.1]. Assuming uniform rotor eccentricity 5/an, working speed is 3000r/min, the initial offset 0.3rnrn, initial velocity is 0, consider the A matrix , B matrix has , zSB perturbations ,[link widoczny dla zalogowanych], the simulation curve shown in Figure 3 Instructions. Node 3 is given here only the simulation results. The results of the other nodes is similar . It can be seen from the figure as A, B matrix generated perturbations , only a slight impact of convergence time , the system perturbation has a strong adaptability. If using non- sliding mode control with compensator , calculations show that the system parameters in a small perturbation , such as dA = 0.04A or dB = 0.04B to cause results to diverge.