Neural Network Sliding Model Control of Radial Translation for Magnetically Suspended Rotor (MSR) in Control Moment Gyro

For a magnetically suspended control moment gyro (MSCMG), the high-speed rotor is actively by magnetic bearings of 5-DOF, but nonlinearity suspension force one main reasons for poor accuracy radial translation (MSR). To solve this problem, here, characteristics are analyzed, and nonlinear dynamic model MSR established. A sliding mode (SMC) based on neural network presented, basis function (RBF) adopted to approximate displacement stiffness current weaken chattering in SMC improve MSR. The stability analyzed Lyapunov functions, rules updating weights presented adaptive algorithms. Compared with these existing classic methods, simulation experimental tests performed single-gimbal MSCMG an angular momentum 200 N.m.s indicated that MSR’s can not only make its more stable also position precision higher.