远程实验室仿真与算法生成的一体化软件设计

Integrated Software Design for Remote Laboratory Simulation and Algorithm Generation

为了解决使用第三方软件搭建远程实验室可能带来的依赖性强、灵活性差、效率低以及数据与隐私泄露问题,实现技术自主可控,以网络化控制系统实验室(NCSLab)的前端软件和硬件为基础,开发了一款面向远程实验室仿真与算法生成的一体化软件M2PLink(mathematics to practice link)。研究了从控制框图到控制代码的映射关系,并对各个控制模块进行了模块化设计。通过分析模块之间的依赖关系及算法结构,建立算法的输出链,并根据微分求解算法设计了多种求解器,最终基于制定的代码生成策略生成对应的控制代码。离线仿真实验和实体实验系统测试结果表明,M2PLink的计算精度满足要求,生成的代码适用于实体控制实验且系统效率得到了大幅度提升。目前,M2PLink已经部署在网络化控制系统实验室中,用于自动化专业实验教学,并且取得了较好的教学效果。

In order to solve the problems of strong dependence,poor flexibility,low efficiency and data security,and privacy disclosure caused by using thirdparty software,an integrated software M2PLink(mathematics to practice link)is developed for remote laboratory simulation and algorithm generation based on the frontend software and hardware of Networked Control System Laboratory(NCSLab).The mapping relationship from the control block diagram to the control code is studied,and each control module is designed with modular structure.By analyzing the module dependency and algorithm structure,the output chain of the algorithm is established,and a variety of solvers are designed according to the differential solving algorithm.Finally,the corresponding control code is generated based on the formulated code generation strategy.The system is tested by offline simulation experiment and entity experiment.The test results show that the calculation accuracy of M2PLink meets the requirement,the generated code is suitable for the physical control experiment,and the system efficiency is greatly improved.At present,M2PLink is deployed in the networked control system laboratory for the experimental teaching of automation major,and has achieved good teaching results.