Application Of Distributed Network Motion Control System In Fast Testing And Sorting Of Solar Cells

The rapid development of the photovoltaic industry poses an urgent need for rapid testing and sorting of solar cells.The motion control system is an important pillar of the solar cell test and sorting equipment.The completion of fast and stable transmission of the solar cells between the various organizations and the speed and accuracy of multi-axis coordinated motion are directly related to the efficiency and consistency of testing and sorting.The high bandwidth of Ethernet is sufficient to support high-frequency data exchange and become an ideal solution for the development of motion control networks.However,the use of Ethernet in the field of motion control networks also introduces some new challenges,which are mainly manifested as that some network problems such as the uncertainty of the network,the time synchronization of the network and the induced delay of the network,can make negative effects on high-speed and high-precision multi-axis coordinated control.This paper makes the analysis and researches aiming at the existing above-mentioned major problems based on the Ethernet used in motion control system.Firstly,according to the demand for high-speed and high-precision coordinated motion of multi-axis motion control in solar cell testing and sorting equipment,the problems that need to be solved are analyzed when the standard Ethernet is applied to the motion control system,which mainly involve the deterministic communication scheduling of Ethernet,the time synchronization of network nodes,and the motion control algorithm.And then a network motion controller that meets the system requirements is designed and implemented.(1)The motion control network must have certainty of communication,the time of sending and receiving information must be predictable.Based on the standard Ethernet,a real-time Ethernet model with deterministic communication capability is established.Without changing the protocol of the data link layer,a deterministic scheduling strategy of time-slot partitioning is achieved by modifying the application-layer software.(2)Time synchronization is the basis of multi-axis coordinated control in asynchronous control network.In this paper,time synchronization method based on IEEE 1588 after several major network time synchronization protocols are analyzed and compared.Aiming at the problem that the synchronization precision is affected by the timestamp and transmission line asymmetry,the Kalman filter algorithm is used to estimate Clock Offset and Clock Skew.Complete offset compensation and frequency correction,improve the accuracy and stability of time synchronization,optimize the structure parameters of time slot division,and increase the real-time performance of the system.(3)If the network motion control system to achieve the desired control effect,the control algorithm must be able to follow the delay changes.The mechanism of network delay and uncertainty is analyzed from the transmission of messages on network protocols and physical links.The influence of delay and jitter on the control performance is analyzed by transfer function.And analyzes the total delay of the system under the time division scheduling method.A statistical system delay model is proposed,which decomposes the system delay into the delay expectation?E and the delay jitter?j,where ?E is the delay expectation of the system ?j is random interference obeying 0-means distribution.The delay expectation?E is compensated to improve the following performance for each axis,the delay jitter ?j is considered as interference,and the effect of delay is minimized by adaptive changes in PID parameters(Kp,Ki,Kd).In view of the above goal,using the self-similar characteristics of delay,the system delay is assumed to be an autoregressive model of discrete time.The Kalman filter algorithm is used to identify the system delay,and the system delay prediction value is obtained.Combining with the strong robustness of fuzzy control and neural network self-learning idea,RBF fuzzy neural network PID is designed.The RBF fuzzy neural network PID compensates for the system delay and jitter suppression.PID parameters can follow the system delay changes adaptively adjusted.Achieve the effect of delay compensation and improve the performance of multi-axis coordinated motion.Based on the self-designed motion controller,the corresponding verification experiments are designed respectively for the above three aspects.The experimental results show that the time synchronization based on the Kalman filter algorithm has higher synchronization accuracy and stability.Communication cycle maintenance has a high stability,the time slot scheduling strategy is effective;Based on the time synchronization and time slot division,Based on the time synchronization and time slot division,the RBF fuzzy neural network PID algorithm based on compensating delay has faster response time,higher steady-state performance and synchronous following performance than the traditional PID algorithm.