Research On Flux-weakening Control Method Of Electro-mechanical Servo System Under All Working Conditions

With the rapid development of power electronic technologies,mechanical engineering and advanced control theories,the electro-mechanical servo system has drawn extensive attention and applications in many high-tech technical fields,especially in precision instruments,med-ical devices,industrial robots,aerospace and many other fields that require high control ac-curacy.Under the background of ”Industry 4.0”,people’s demands for the performance of electro-mechanical servo products are constantly increasing.It is required to be competent for various working conditions in industrial applications,and has superior field weakening control performance.Research on flux-weakening control method of electro-mechanical servo system under all working conditions has become a hot issue in related fields all over the world.This dissertation takes the electro-mechanical servo system as the research object,mod-els its complex nonlinearity,grasps several difficulties of its flux-weakening control under all working conditions,and designs a novel control method that meets the demands of the modern electro-mechanical servo system.The main works are summarized as follows:1.In order to design a set of control methods that meet the various demands,this paper first considers various working conditions and models the electro-mechanical servo system compre-hensively and detailedly.Considering the generality,this dissertation establishes a mathemat-ical model of the interior permanent magnet synchronous motor(IPMSM).According to the model,the problems caused by the nonlinearity,cross-coupling and parameters time-varying are analyzed for the current controller design.Then,according to the different load charac-teristics of the electro-mechanical servo system,the modeling of the load is divided into three cases: rigid load,flexible load and sudden increase and withdrawal added to these two kinds of loads.Additionally,it is pointed out that the torque nonlinearity,parameters time-varying and various forms of external disturbance must be considered in the design of speed control.This dissertation also models the nolinear friction torque of the flexible load,and points out that it will cause the position ”flat-top” phenomenon and speed ”dead-zone” phenomenon,and finally cause the decrease of positioning accuracy.2.Apart from the id=0 control framework,the flux-weakening control framework is usu-ally applied to electro-mechanical servo system to tap the potential of the motor extreme speed and torque.This dissertation introduces the basic concepts and basic principles of the flux-weakening control,and compares several commonly-used flux-weakening control frameworks.Moreover,it is pointed out that the effects of nonlinearity,cross-coupling and parameters time-varying under the field weakening control framework are greater than the id=0 control frame-work,and they must be specially considered when designing the current control method.3.Considering the nonlinearity and cross-coupling of the current dynamics under the flux-weakening control framework,an integral nonlinear predictive control(NPC+I)is designed for the current loop.First,a quadratic cost function with the goal of minimizing the d- and q-axes current tracking errors is designed.Then,the output prediction is completed through the current dynamics nominal model.Next,the receding optimize is done to obtain the current control laws.Last,the steady-state performance is ensured through an integral compensation.After the stability analysis,comparative simulations and experiments are carried out.The results show that NPC+I suppresses the effects of nonlinearity and cross-coupling.But the performance gets worse with the parameters time-varying.4.The problem of parameters time-varying under the flux-weakening control is more serious than the id=0 control,so the NPC+I can not meet the demand.In this dissertation,the design of the current loop control is further improved for the time-varying problem,and a generalized proportional integral observer(GPIO)based nonlinear predictive current control(NPC+GPIO)is proposed.The NPC+GPIO regards the parameters time-varying as a kind of internal disturbance of the current dynamics,and then designs a GPI observer to estimate it.Then,the time-varying estimation obtained by the GPI observer is employed to improve the output prediction.Last,receding optimization is finished to obtain the current control law.After some stability analysis,a comparative experimental verification is carried out,which proves the effectiveness of the proposed method for improving the current loop control performance in the flux-weakening control framework.5.Based on the superior performance of the inner-loop current control,this disserta-tion continues to solve the problems of the outer-loop speed control.Aiming at the problems of torque nonlinearity,parameters time-varying and various forms of external disturbance of the speed dynamics under all working conditions,this dissertation promotes the application of NPC+GPIO to speed control to solve these problems.First,a quadratic cost function is designed with the goal of minimizing the speed control deviation.Then,the parameters time-varying of the speed dynamics is dealed with and treated together with the load torque as a lumped distur-bance.The disturbance is then estimated by a GPI observer.Additionally,the output prediction is performed based on the disturbance estimation.Last,the receding optimization is done to obtain a speed control law.After the stability analysis,the experimental verification at differ-ent operations of speed and full working conditions is carried out.The results show that the NPC+GPIO has good performance under the three working conditions of no-load,inertia plate and sudden increase and withdrawal load.However,the NPC+GPIO still has position ”flat top” phenomenon and speed ”climbing” phenomenon under the working condition of electro-mechanical actuator.6.Aiming at the nonlinear friction of flexible load,this paper proposes a nonlinear speed predictive control based on the friction compensation and GPI observer(NPC+GPIO+FRIC).The difference between this method with the NPC+GPIO is that this method models the fric-tion specially.Considering the accuracy and complexity,the STRIBECK piece-wise linearized friction model is employed.Then,the least squares matrix algorithm which has small online calculation is employed to identify the friction model.Additionally,the friction model error is also added to the lumped disturbance,which is estimated by the GPI observer.Last,the non-linear predictive speed control is redesigned based on the friction compensation and the GPI observer.After the stability analysis,the experimental verification is performed again on the electro-mechanical actuator platform.The position ”flat top” phenomenon and speed ”climb-ing” phenomenon are almost completely eliminated.