Research On Mechanical Resonance Suppression For Servo Systems

Servo systems are widely used in industrial production. In a servo system, mechanical resonance phenomenon is easily to be triggered as the transmission device between the motor and the load is not an ideal rigid body. Mechanical resonance can reduce the tracking performance of the system, and even make the system no longer stable. In this thesis, several methods to suppress mechanical resonance for servo systems whose transmission devices are not ideal rigid bodies are investigated on the aspect of control. Contents of this thesis are mainly arranged as follows:(1) The model of the two inertia system is established and the closed-loop domain pole assignment strategy by a controller in general form is discussed to suppress the mechanical resonance, where closed-loop domain poles are assigned according to the expected performance of servo systems and rules are given to select controller’s parameters with application conditions analyzed for each kind of pole assignment method. This thesis is mainly focused on closed transfer function whose order is three or four, and the fact is pointed out that the pole assignment based on IP controller is better than the pole assignment based on the traditional PI controller.(2) Aimed at the problem that IP controller fails to assign appropriate poles for two inertia system with arbitrary load rotating inertia ratio, the effect of acceleration feedback of motor side on the two inertia is discussed and it is put forward that acceleration feedback of motor side can change the equivalent load rotating inertia ratio of the system. At the same time, impact of acceleration feedback of load side on the two inertia system is also investigated, and the conclusion is conducted that acceleration feedback of load side will increase the equivalent mechanical resonant frequency, so mechanical resonance can be eliminated by adjusting the equivalent resonance frequency of the two inertia system out of the operating bandwidth.(3) Mechanical resonance suppression method based on low-pass filter is proposed for two inertia systems with low mechanical resonant frequency. The low-pass filter is designed to make the characteristic between the input and output of the designed reference model be considered as that of the motor side approximately, and the fast terminal control law is designed to guarantee the dynamic tracking performance of the servo system. This method has been proved effective in terms of both theoretical analysis and simulation results.(4) As far as the two inertia systems whose parameters are not fully known and whose load sides are affected by friction factor are concerned, the adaptive dynamic surface algorithm is studied to control the system. Continuously differentiable friction model is adopted to construct the state space expression of the two inertia system and the control laws on the basis of the adaptive dynamic surface are designed according to the state space expression constructed before, which suppress the impacts of the friction and mechanical resonance on the system and guarantee the dynamic tracking performance of the load side.The effectiveness of this control method is testified by the simulation at last.Finally, all mechanical resonance suppression strategies studied in the thesis are summarized and other approaches to eliminate mechanical resonance based on the previous research are prospected in the future.