核电数字化控制系统PID参数自动寻优软件研制

Research on Automatic Optimization Software for PID Parameters of Nuclear Power Digital Control System

在核电控制系统调试运维过程中,比例-积分-微分(PID)控制参数通常采用试凑法得到,调试时间较长。为了提高PID控制参数调试效率,本文以辨识模型为基础,采用多目标粒子群优化算法(MOPSO)研制了PID控制参数自动寻优软件。本软件利用参数模型辨识方法中的最小二乘法对系统进行辨识,保证了辨识的快速性与辨识结果的可靠性。该软件选取"华龙一号"工艺仿真模型中的PID单回路控制设备进行测试。测试结果表明,该核电数字化控制系统PID参数自动寻优软件性能良好,控制效果满意。

During the commissioning and operation of the nuclear power control system,the proportional-integral-derivative(PID)control parameters are usually obtained by the trial and error method,which results in long debugging time.In order to improve the debugging efficiency of PID control parameters,based on the identification model,a software for automatic optimization of PID control parameters is developed by using multi-objective particle swarm optimization(MOPSO).The software uses the least square method of parameter model identification method to identify the system,which guarantees the rapidity of identification and the reliability of identification results.The test results show that the software performs satisfactorily in single-loop PID control system of HPR1000 simulation model.