|
| 针对阀门黏滞特性的多阈值 Knocker 控制算法研究 |
| Research on Multi-Threshold Knocker Control Algorithm for Valve Sticking Characteristics |
| 投稿时间：2022-01-04 |
| DOI： |
| 中文关键词: Knocker 二叉树模型 黏滞 樽海鞘群算法 |
| 英文关键词:Knocker binomial tree model sticky state bottle sheath swarm algorithm |
| 基金项目:湖南省自然科学基金资助项目（2020JJ6078）；湖南省学位与研究生教育改革研究基金资助项目（2020JGYB210） |
|
| 摘要点击次数: 292 |
| 全文下载次数: 419 |
| 中文摘要: |
| 为了研究阀门的黏滞特性引起的控制变量振荡问题，结合阀门运动特性建立了二叉树黏滞模型。通过在Knocker补偿器算法上引入Smith预估器抵消反馈的滞后影响，建立了樽海鞘群算法优化的多阈值Knocker补偿控制策略。实验结果显示，对比单阈值Knocker策略改进的Knocker补偿策略在减少控制阀振荡上有更好的效果，IAE和ISE指数比改进前分别减少了73.7%和80.7%，使控制的稳态误差保持低于1%的同时，大大减少了阀杆的动作次数。 |
| 英文摘要: |
| In view of a research on the oscillation of control variables brought about by the viscous characteristics of valves, a binary tree viscous model has thus been established based on the valve motion characteristics. By introducing Smith predictor into the Knocker compensator algorithm to offset the hysteresis effect of feedback, a multi-threshold Knocker compensation control strategy is established optimized by the bottle sheath swarm algorithm. Based on an extensive comparison experiments, the results show that the improved Knocker compensation strategy is characterized with a better effect on the reduction of the control valve oscillation than the single-threshold Knocker strategy. IAE and ISE indexes of the former are reduced by 73.7% and 80.7% respectively compared with the latter, thus keeping the steady-state error of the control below 1%, with the action times of the valve stem greatly reduced as well. |
|
|
| 关闭 |
