基于主成分分析-BP神经网络的风电备件需求预测

风电机组具有结构复杂,运维困难,且长期处于恶劣的工作环境的特点。风电备件的需求预测有助于为风电场配备最合适的备件数,以确保风电场的平稳、高效运行。本文以某企业近五年的齿轮需求数据为例,应用主成分分析(PCA)降低几个影响因素的维度,得到复合变量,然后应用BP神经网络算法,构建预测模型；比较ARIMA模型、BP神经网络预测和PCA-BP神经网络预测的结果。表明：主成分分析法(PCA)能显著降低神经网络预测误差,预测的精度为93.94%,高于BP神经网络预测的88.39%和ARIMA模型的85.31%,所以PCA-BP神经网络模型的预测精度准确且有可靠结果,能够适用风机备件的需求预测。

Research on wind power spare parts demand forecasting based on PCA-BP neural network

In order to ensure the smooth and efficient operation of wind farms, spare parts for wind turbines with complex structures and difficult operation and maintenance are accurately predicted. This paper takes the gear demand data of an enterprise in the past five years as an example, and uses principal component analysis (PCA) to reduce the dimensionality of multidimensional data to obtain comprehensive variables, after which the prediction model is constructed by BP neural network algorithm, and the results of ARIMA model, BP neural network prediction and PCA-BP neural network prediction are compared. The results show that the principal component analysis (PCA) can significantly reduce the neural network prediction error, and the prediction accuracy is 93.94%, which is higher than the 88.39% of BP neural network prediction and 85.31% of ARIMA model, so the PCA-BP neural network model has higher accuracy and reliable results, and can be used for the demand prediction of wind turbine spare parts.