Study On Key Technologies Of Multi-axis Coordinated Motion For Networked Motion Control Systems

Motion control(MC)is an important pillar of modern electro-mechanical equipments.The coordinated motion(CM)for multi-axis is the key elements for equipments to implement complicate functions and to improve product quality.With the rapid progress of the modern equipments in the direction of super-speed and high-accuracy,firmer requirements are proposed for motion control systems(MCS).The wide application of networked motion control systems(NMCS)not only brings great convenience for multi-axis MCS,but also puts forward challenges for CM with high speed and high precision.To deal with the challenges NMCS faced,this thesis focuses its goal on the relative problems for multi-axis NMCS to improve its CM precision.Firstly,since network time delay(NTD)has great influence on the performance of NMCS,researching on its action characteristics of NTD to control systems is the basis to solve multi-axis synchronization problems.Based on the analysis of NTD arising mechanism and its characteristics, a NTD model is presented with statistics principle,which divided the NTD into two parts of delay time expectationÏ„_D and delay time jitterÏ„_J.WhereÏ„_D is the mathematics expectation of NTD andÏ„_J is considered as bounded stochastic disturbance with zero mean value.From the perspective of transfer function,the influence of NTD on control signal is analyzed,and by combining together networks with control target,the discrete time mathematical model between them is built on the condition of time delay randomly,which provided a theoretical basis to analyze the influence of NTD on control systems.Aiming at the instance that the communication delay is less than a control cycle of motion control network,based on the relative definition about network delay statistics model,a new discrete time model is presented,which proved that the being of network delay changes the discrete state equation's real inputs into three items combination of u(k),u(k-1)and u(k-1)-u(k), and the coefficient matrices are B⁰,B¹ andâ–³B respectively.Where the value of B⁰,B¹ is correlated withÏ„_D and can be considered as known parameters;â–³B is correlated withÏ„_J and is an unknown parameter,and can be processed with u(k-1)-u(k)simultaneously as disturbance.The model provides a theoretical basis to improve the performance of NMCS by compensation of network delay and interference suppression.Aiming at the synchronized performance of multi-axis NMCS,the multi-node synchronization mode of NCSs is studied and the node driven mode,signal sampling mode and time delay of feedback channel are discussed. The extended state equation for closed loop multi-axis NMCS is presented, which built a foundation for analysis and synthesis of the control systems.Secondly,the network nodes synchronized precision plays an important role in the improving of multi-axis synchronized motion accuracy,and it is an effective way to improve node synchronization precision by network time delay compensation.In NMCS,precise time synchronization among motion controller node, actuator nodes and sensor nodes is the basis for multi-axis synchronization. According to the different characteristics ofÏ„_D andÏ„_J,a new method is proposed.The synchronization precision of network nodes is further improved by compensation ofÏ„_D,while considers the influence ofÏ„_J as a disturbance and eliminates it by disturbance suppression.According to the objective above mentioned,a network delay identifier is designed based on genetic algorithm,which presents an online identification method for networks with time delay greater than one cycle,and the simulation experiments demonstrated that the identifier provides very high identification precision.The network time delay precise measurement is studied,the measurement data of time delay is processed based on the correlative theory of mathematical statistic,which provided a basis for the implementation of time delay compensation.Through researching of the network time delay compensation method,a compensation strategy based on timer is presented.By the method of delay compensation,the network delay from controller node to actuators and sensor nodes will have the same mathematic expectationÏ„_D and it will improve the synchronization precision further and simplify the motion control model.The network time delay jitterÏ„_J cannot be obtained precisely.In the object discrete model,the influence of jitter on control systems is considered as an attached disturbance,and a grey estimator is worked out based on grey system theory,which is used to identify the disturbance item.The simulation result proved that the disturbance item can be restrained very well by compensation control based on the identified results of the grey estimator.In order to ensure the satisfaction of modeling condition of the feedback message's delay in networked multi-axis control systems,a RBF neural network predictor is designed.The information in feedback channels that exceed the scheduled time deadline is predicted and the feedback information with uncertain time delay are changed into estimated information with determined time and bounded error by the predictor,to ensure the delay of feedback messages in their setting time range.The simulation showed that the predictor had favorable prediction precision and multi-step prediction ability.Furthermore,in networked motion control systems,the multi-axis coordinated motion precision are not only determined by the time synchronization precision of each network node but also closely related to the dynamic characteristics of motor driven device and motion executing mechanism and the parameters matching status among axis.Therefore,the multi-axis synchronized motion control need to be further studied on the side of control algorithms.The improvement of tracking accuracy for single-axis driven system can improve the coordinated motion accuracy for multi-axis indirectly.According to the purpose,a self-adaptive fuzzy control algorithm for AC servo system is proposed,which combines with the merits of fuzzy control and the study ability of self-adaptive control system,and has strong ability of withstanding parameter perturbation and restraining disturbance.Simulation results demonstrated that the algorithm gives a good tracking precision.Coupling control is an effective control method for improving multi-axis systems' coordinated motion precision and disturbances rejection ability. Aiming at the multi-axis coupling control,a novel coordinated control algorithms based on virtual reference axis is presented.A multi-axis coupling controller with sliding model control theory is designed to achieve multi-axis velocity and position coupling control.The algorithm's convergence and stability is justified rigorously with mathematics,and the simulation results show the effectiveness of the control algorithm.In computer numerical control(CNC)systems,the cross-coupled control algorithm can improve the contour accuracy effectively.But it is very difficult to estimate the contour errors of complex curves and that restricted the application of cross-coupled control algorithm seriously.According to the output characteristics of modern CNC,a new space contour error algorithm is established,which uses the "cutter locations" outputted by interpolator of numerical controller to computing the contour error vectors and no mathematical model of the machined curve is needed.The calculating precision of the model is stable and it can be applied to any space curve.Based on the contour errors calculating model above mentioned,a double closed loops controller consisted of contour loop and position loop is designed, to implement the closed loop control for each single axis' position and their synthetic motion contour of servo system,which improved the multi-axis position coordinated precision efficiently.The simulation results show that the controller can improve contour accuracy obviously in the condition of unchanging the single-axis servo system's position tracking precision, especially for the parameters of servo axles are relative large mismatched,the proposed controller can improve contour precision significantly. At last,the response rate of servo systems has great importance in improving the dynamic synchronized performance for multi-axis system,and direct torque control(DTC)can improve motor's torque response rate and dynamic performance greatly,but it has the shortcoming of greater torque ripple.This thesis presented a new DTC algorithm based on the function relationship between the electromagnetic torque and torque angle of permanent magnet synchronous motor(PMSM).The torque error is directly mapped into the conduction time of the power switches of three-phase voltage source inverter to achieve arbitrary space voltage vector output by the proposed algorithm,and a perfect circular track of stator flux linkage vector can be generated,therefore it can reach the target of reducing torque ripple and improving torque output precision.The method is simple in computation and can avoid complicated sector discrimination and logical calculation in traditional DTC and SVM(Space Vector Modulation)control methods.The algorithm can achieve less than±0.5%torque ripple by the simulation results.Because the map between the torque and torque angle takes the relationship as nonlinear,which makes the normal torque controller with fixed parameters cannot give stable and uniform control precision in the full torque working range.Therefore,a parameter self-tuning PI torque controller is designed based on RBF neural networks.The simulation results indicated that the controller can keep lower torque ripple in the full torque output range of a PMSM,and a high torque control precision can be achieved as the output torque varies between 0～T_emax.