中国机械工程

• 机械基础工程 • 上一篇    下一篇

支持矢量回归机的参数优化及在智能减压阀压力预测中的应用

童成彪1,2;周志雄1;周;源1   

  1. 1.湖南大学,长沙,410082
    2. 湖南省特大口径电站阀门工程技术研究中心,长沙,410007
  • 出版日期:2016-07-25 发布日期:2016-07-22
  • 基金资助:
    国家科技重大专项(2011ZX07412-001-02)
    National Science and Technology Major Project(No. 2011ZX07412-001-02)

Parameter Optimization of Support Vector Regression and Its Applications to Pressure Prediction of Intelligent Pressure Reduce Valves

Tong Chengbiao1, 2;Zhou Zhixiong1;Zhou Yuan1   

  1. 1.Hunan University,Changsha,410082
    2.The Special Largest Size Valve Engineering Research Center of Hunan Province,Changsha,410007
  • Online:2016-07-25 Published:2016-07-22
  • Supported by:
    National Science and Technology Major Project(No. 2011ZX07412-001-02)

摘要: 智能减压阀可通过控制膜片缸压力实现出口压力的智能调节。膜片缸压力是智能减压阀控制器的控制目标,因此需要依据进口压力和出口目标压力对膜片缸压力进行预测。基于此,提出了基于人工化学反应优化算法的支持向量回归机(ACROA-SVR)参数优化方法,并将ACROA-SVR应用于智能减压阀膜片缸压力预测,采用实验数据将ACROA-SVR与基于遗传算法的SVR和传统SVR进行了对比,分析结果表明了ACROA-SVR的有效性和优越性。

关键词: 人工化学反应, 支持矢量回归机, 参数优化, 智能减压阀, 回归预测

Abstract: Intelligent pressure reduce valve might regulate the downstream pressure by controlling the diaphragm cylinder pressure. Diaphragm cylinder pressure was the control goal of pressure reduce valve intelligent controller, so it was necessary to forecast the diaphragm cylinder pressure according to upstream pressure and downstream pressure. Thus, the parameter optimization method for SVR was proposed based on artificial chemical reaetion optimization algorithm(ACROA-SVR). Furthermore, the ACROA-SVR was applied to the pressure prediction of the diaphragm cylinder pressure herein. By analyzing the experimental data, ACROA-SVR was compared with SVR based on genetic algorithm and traditional SVR. The efficiency and superiority of ACROA-SVR was shown by the analyzed results.

Key words: artificial chemical reaction(ACR), support vector regression(SVR), parameter optimization, intelligent pressure reduce valve, regression prediction

中图分类号: