微型电脑应用

针对输电线路覆冰给电网企业带来的危害,以及传统预测模型精度不高的问题,提出一种改进的LSSVM覆冰预测模型。首先采用烟花算法对LSSVM模型参数进行寻优,并引入量子进化算法在全局搜索的优势对烟花算法进行改进,以得到LSSVM中最优的C和δ参数;最后,在实例分析部分,采用PCA主成分析法降维,提取出环境温度、相对湿度、风速、光照强度和负荷电流作为LSSVM的输入参数,得到PCA-QFA-LSSVM输电线路覆冰模型,并将改进算法与BPMM、SVM等算法比较,得到PCA-QFA-LSSVM在覆冰厚度预测方面与实际值的相对误差都在3%以内,表明该算法比传统的预测方法具有更为明显优势。

In view of the harm of transmission line icing for power grid enterprises and the low accuracy of the traditional prediction model, an improved LSSVM icing prediction model is proposed. Firstly, PCA principal component analysis is used to reduce the dimension of multi-dimensional parameters in LSSVM, and environmental temperature, relative humidity, wind speed, light intensity and load current are extracted as the input parameters of LSSVM. Fireworks algorithm is used to optimize the parameters of LSSVM, and the advantages of quantum progression algorithm in global search are introduced to improve the fireworks algorithm, so as to obtain the optimal C and load current in LSSVM δ parameter. Finally, combining the above steps, the PCA-QFA-LSSVM transmission line icing model is obtained. The improved algorithm is compared with BPMM, SVM and other algorithms, and the error between PCA-QFA-LSSVM and the actual value is the smallest, which shows that the algorithm has more obvious advantages than the traditional prediction method.