基于改进支持向量机的抽汽管道阻力特性混合建模

The mixed model of the flow resistance in pipeline based on improved SVM

结合机理分析与统计智能算法,建立的热力系统子系统混合模型,可以充分发挥两种方法各自的优势。经现场实际数据分析,基于遗传算法改进ε-SVM的抽汽管道压损模型,符合理论分析规律,并且具有较高的精度。

The mechanism analysis and statistics intelligent algorithm was combined in this paper to establish mixed model of the thermal subsystems, which was a useful way to express the respective advantages of the two methods. Through the analysis of actual data, the pressure loss model of the steam extraction in pipeline based on support vector machine （SVM） improved by genetic algorithm, coincidences theoretical analysis rule, and has high accuracy to meet the need of the state reconstruction for the purpose of the energy consumption analysis.