Fault Tolerant Control Of A Lorentz Self Bearing Motor Considering Open Coil Faults

This paper presents a decoupling and fault tolerant control algorithm for a segmented arc, Lorentz self bearing motor with open faults. The particular motor uses 12 phases to produce forces in the x and y directions and the motoring torque. The algorithm uses the redundancy in these phases to adapt to open faults that result in failed phases. In this approach a model of the actuator torque-current and force-current relationship is computed for a given fault condition. Since this model includes the commutation sequence for the motor, it must be computed at each sampled time step. This model is then inverted onto itself which reduces the current gain matrix to the identity matrix and eliminates any cross-coupling effects. Any desired current gain matrix is then achieved by inserting it into the feed forward path. Results indicate a high level of redundancy in the actuator such that several faults can be accommodated simultaneously. The results also show that the cost of a fault may be a higher power usage and lower peak force/torque ability.