基于EPICS和差分进化算法的同步辐射光束线智能调束系统

Intelligent commissioning system based on EPICS and differential evolution algorithm for synchrotron radiation beamline

针对无规律变化的光源光束点和手动调束费时费力问题,本文实现了基于实验物理及工业控制系统(Experimental Physics and Industrial Control System,EPICS)的同步辐射光束线智能优化调束。该系统基于差分进化算法,建立光束线智能优化模型,应用LabVIEW程序实现调束过程的自动优化。系统可以自由选择待优化的电机及其搜索范围,设置算法参数,跟踪进化进程,通过CaLab接口模块与光束线EPICS软件平台进行通信,控制电机运动。对该系统在上海光源衍射线站进行了在线测试,首次成功地在EPICS控制平台上实现了光束线的智能调束优化。测试结果表明:该智能调束系统能够较快地收敛于最优解,收敛时间大约30 min,较手动优化效率提高一个数量级以上。

[Background]In synchrotron radiation facilities,it is important to keep beamlines operating in optimal conditions.The debugging process is normally very time consuming due to the irregular light source beam point,and it is not easy to get global optimum.[Purpose]This study aims to develop an intelligent debugging system based on Experimental Physics and Industrial Control System(EPICS)and differential evolution algorithm for synchrotron radiation beamline commissioning.[Methods]First of all,based on the differential evolution algorithm,intelligent optimization model of beamline was established.Then the automatic optimization of beam adjusting process was implemented by using LabVIEW program and communication with the EPICS-based control system was achieved by CaLab interface module.Functions of the user interface,motion control,algorithm implementation and evolution processing were integrated in the LabVIEW program.Finally,this intelligent commissioning system was tested at the X-ray diffractive(XRD)beamline of SSRF(Shanghai Synchrotron Radiation Facility)for optimization of the beam flux at sample position by adjusting the beamline optical components.[Results&Conclusions]Online tests results show that this intelligent commissioning system converges to the optimal solution quickly,and the convergence time is about 30 min,more than one order of magnitude higher than manual optimization efficiency.