Mechanical Resonance Suppression Based On Shaft Torque Limitation Control

Electrical servo systems often use elastic shafts and couplings to connectmotor and load, which can be simplified as a two-inertia elastic system. Theelasticity tends to cause mechanical oscillations, which will deteriorate torquetransmission characteristic and the life span of mechanical parts. Meanwhile, itwill lead to system insecurity, such as breaking the shaft if the actualtransmission torque is beyond shaft tolerance. Therefore, it is important toemploy some methods to suppress mechanical oscillations and control the shafttorque amplitude.Based on the two-mass elastic system, this paper has analyzed thephenomenon of mechanical oscillation and shaft torque exceeding its tolerancedue to elastic transmission. The method of actively changing parameters andstructure of the controller is used to avoid the phenomenon above-mentioned.First, the two-mass elastic system model is abstracted from electrical servosystem. According to its operation process, the mathematical expressions areestablished, from which the simulation system diagram is obtained. The systemtransmission functions and frequency characteristic are both obtained withoutspeed controller. The phenomenon of mechanical oscillation and shaft insecurityare simulated with the actual system parameters.Then, to solve the problem of mechanical resonance and shaft insecurity,three speed control strategies based on PI are designed. The first one usesengineering method which ignores the shaft elasticity to calculate the PIparameters. The second method is pole placement, whose idea is to optimizepoles of the system through the qualification of identical damping factor. Basedon the second method, shaft torque feedback is employed to modify theoscillation character of the system. Simulation results based on three strategiesare given.In addition, as PI control cannot limit the shaft torque freely, this paper hasdesigned the MPC (model predictive control) speed controller. Based on the stateequation and limitations of the system, Multi-Parametric Toolbox is employed todesign the controller. In order to realize the controller in a real-time system, partof the calculation process is done offline. Then, the constraints of shaft torquelimitation are analyzed and the simulation results at random shaft torque limitsare given.In the end, on experiment platform of the two-mass elastic system, thephenomenon of mechanical oscillation and shaft insecurity is firstly experimented at high and low speed. Meanwhile, the experimental results basedon engineering method, pole placement method, the shaft torque state feedbackand MPC at high and low speed are experimentally verified. The waveforms offour methods are compared and according to further exploration of characteristicof the four methods, the progressive relationship is obtained. MPC is verified asthe optimal method.