Study On Hybrid Dynamic Control Technology For Motor Servo Systems

With the development of microprocessor and power electronics technology,servo control system based on analog devices is now giving way to digital servo system with high-computation power microprocessors such as single-chip microcontroller,digital signal processor(DSP),and etc.More advanced power devices are used in servo control systems to enhance the reliability and performance of the system.At the same time,the developed vector control and direct torque control of AC motors greatly simplify the AC motor control system and promote the AC motor control technology to a new height.AC motor servo systems are now widely used in daily life and industrial areas.Aiming to realize the efficient control of motor servo systems,the thesis focuses on the study of hybrid dynamic control(HDC)technology.In a hybrid dynamic control system,discrete event logics are integrated with continuous variable dynamics control(CVDC)techniques.By event handling and control law scheduling,the performance limitations of conventional CVDC can be overcome,and high performance control can be achieved.The first HDC algorithm studied in this thesis is the explicit model predictive control,which belongs to the switched-type hybrid dynamic control techniques.The algorithm offline calculates the control law for each patch of the state space,and then stores all the state-space patches and their corresponding control laws in a matrix.Online calculation only needs to determine the patch which the current state falls into and then execute the corresponding control law.As the system state changes,the control law switches by the rules in real time.This method reduces the online computing time and improves the computing speed.However,the offline calculations call for an accurate model to obtain accurate results.If there is a mismatch in model,the control effect might be degraded in practical application.Next,the time optimal model predictive control(TOMPC)is studied in this thesis.It is a two-layer optimal control scheme,belonging to the hierarchical HDC control category.The first layer of TOMPC is the traditional model predictive control problem,and at each time step the second layer minimizes the prediction horizon of the first layer,i.e.,the settling time.The traditional model predictive control problem can be solved efficiently by using the on-line active set method.Since the first layer has a priority solution,the TOMPC will have the optimal solution if the second layer has been solved.The feasibility of the control scheme is guaranteed.Based on the mathematical model of permanent magnet synchronous motor(PMSM)in the dq coordinate system,the vector control scheme is adopted in this thesis,with the direct-axis current being set to zero.Thus,a second-order system model is derived for PMSM position servo system.The two aforementioned hybrid dynamic control schemes are used to realize position control.Digital simulation is first conducted in MATLAB.Then,a PMSM testbench is constructed with a 32-bit float-point chip TMS320F28335DSP as the core hardware,together with motor drive circuit and external hardware circuits.Control algorithms are programmed using the Code Composer Studio(CCS)software system,and experimental results under different position targets verify the effectiveness of HDC controllers.