考虑PCA-RBF的变压器标底预测模型研究

Research on the Transformer Pre-Tender Estimate Forecast Model Considering PCA-RBF

针对变压器制造企业投标报价缺乏科学理论指导,准确性低的问题,以准确预估下一批招标的标底价为目标,提出了结合主成分分析(PCA)与人工神经网络(ANN)的变压器标底预测模型-PCA-RBF标底预测模型。模型采用PCA对高维原始变量进行降维处理,提取主成分(PC)作为径向基函数神经网络(RBF)的输入。PCA算法降低了原始变量的维数,去除了各变量间的相关性,在简化ANN结构的同时,提高了预测模型的精度。仿真结果表明,PCA-RBF变压器标底预测模型具有较高的预测精度。

Aiming at the problem of transformer manufacturing enterprises bidding is lacking of scientific theoretical guidance and low accuracy, for accurately predicting the next batch of tenders pre-tender estimate price,combining principal component analysis(PCA)and artificial neural network(ANN)of the transformer pre-tender estimateforecast model is proposed-PCA-RBF pre-tender estimateforecast model. The model uses PCA to preprocess the original multidimensional input variables, and selects the principal component (PC) as the input of the radial basis function neural network (RBF).The PCA algorithm reduces the dimension of the original variables, removes the correlation among the variables, and improves the accuracy of the prediction model while simplifying the ANN structure. The simulation results show that PCA-RBF pre-tender estimateforecast model has higher prediction accuracy.