基于NWAPSO参数全局寻优的LS-SVM电站风机性能预测  

Performance Forecast in the Fan of Power Station with LS-SVM Based on Parameter Optimization by NWAPSO

作　　者：朱予东[1] 王星久[1] 王天龙[1] 胡佳琪[1] 

机构地区：[1]华北电力大学电站设备状态监测与控制教育部重点实验室,河北保定071003

出　　处：《应用能源技术》2011年第10期46-48,共3页Applied Energy Technology

摘　　要：风机的性能曲线是风机选型和优化运行的重要依据.通常该曲线通过试验试验数据和性能图表上的数据进行曲线拟合获得.由于该曲线非线性很强,传统方法复杂昂贵,而且拟合精度不高。针对以上不足,提出了一种基于非线性权重自适应粒子群优化(NWAPSO)参数全局寻优的最小二乘支持向量机(LS-SVM)风机性能预测方法。通过最小二乘支持向量机建模,并应用非线性权重自适应粒子群优化算法对模型参数进行全局寻优,得到具有较高精度的风机性能曲线。计算结果表明,根据本文方法建立的模型很简洁,只需要知道少量的训练样本就能建立,可以比较精确的预测风机性能,具有较显著的工程应用价值。The performance curves of the fans are essential basis of type selection and operation optimization of fans. Generally, these curves are obtained by curve fitting with data from experiments or performance diagrams. Nevertheless, the curves are highly nonlinear, therefore, the traditional methods are expensive and the fitting accuracy is sketchy. In allusion to the sketchiness above, a new algorithm with the least squares support vector machine （I.S-SVM） based on parameter optimization by adaptive particle swarm optimization algorithm based on nonlinear inertia weight （NWAPSO） is pointed out. The calculation model of the fan is set up based on the LS-SVM at first, the parameters of the model are then globally optimized with the NWAPSO, the performance curves of the fans are finally obtained comparatively accurately. The calculation result indicates that the model with LS-SVM based on parameter optimization by NWAPS0 is pithily and can be set up with a small amount of train examples and the performance of the fan can be forecast accurately, the engineering value is remarkable.

关 键 词：非线性权重自适应粒子群优化 最小二乘支持向量机 风机 性能预测 

分 类 号：TK414.2[动力工程及工程热物理—动力机械及工程]