Research On Control Algorithms Of Multi-Motor Driving Servo Systems

As the needs for systems with high inertia and power demand in industry and military increases,multi-motor driving servo systems has attracted more and more attentions.However,the non-linearity from backlash,fraction and asynchronous angular velocities among the motors,due to the transmission through cogs,will introduce side effects to the performance of control.In this paper,multi-motor driving systems are well studied to achieve load tracking and synchronization over multiple motors,by applying backstepping,robust adaptive control,decentralized control and cascade control etc.The main contents of this thesis are summarized as follows:(1)Based on dual-motor driving servo system embedded with Lu Gre model for frictions,the methodology of adaptive backstepping friction compensation with observer has been proposed.To take into account the unknown state of the friction model,the Lu Gre friction model based adaptive state observer has been designed,based on which an adaptive backstepping algorithm has been proposed.To address the dead zone issue and truncation in tracking with satisfactory performance.Due to the prediction of fictions exists in the process of algorithms designing,this controller has good robustness for friction nonlinearity.(2)For models of dual-motor driving system that features both non-linearity due to backlash and friction,and unknown disturbance,an extended state observer based robust adaptive controller has been proposed.First,the overall disturbance is defined by adding up each part of non-linearity with disturbances.A robust adaptive controller has been established,based on the extended state observer introduced to estimate the state of system and overall disturbance,and the adaptive rule accordingly designed.In the end both steady state behavior and transient performance in terms of L_?norm turned out to be satisfactory through analyses.The simulation suggests that such approach is able to address the nonlinearity issues and suppress external disturbances.(3)A decentralized,optimal controller for multi-motor systems with adjacent cross-coupling has been proposed.First,new metrics for defining synchronization error of multi-axis systems were proposed based on the adjacent coupling structure.The sliding mode controller was designed through combination of synchronous control and tracking,to achieve synchronization of multi-motor systems.After that,to deal with four-motor driving servo systems with interaction terms,a decentralized controller has been designed to minimize the cost function,under the assumption that backlash and frictions are of Lipschitz continuity.The simulation and the results from the experimental platform with four-motor driving systems have verified the effectiveness of this methodology.(4)As concerns the generalized cascade systems with multiple driving system and single driven system,generalized coupling error based on graph theory has been introduced and a new time-varying sliding mode controller which is able to eliminate the reaching phase and singularity.As a special case of generalized cascade systems,the original problem of complicated coupling was converted to the convergence of generalized coupling error,by designing the generalized coupling error to deal with coupling in tracking and synchronization in multi-motor driving systems.With echo neural networks,a non-singular,time-varying sliding-mode controller that eliminates the arrival stage has been designed,with which the convergence of error was accelerated while chattering could be suppressed.Last but not least,the algorithm is able to achieve convergence in limited time given saturated input.