基于神经网络技术的泵站机组性能预测

Performance Prediction for Pumping Station System Based on Neural Network Technique

性能曲线是反映泵变工况运行情况的一类曲线,通常该曲线均是通过实验或是以已有数据为基础的性能换算而获得,但前者费用昂贵,后者准确性差 为此,通过对BP人工神经网络模型的分析和研究,提出了利用BP神经网络技术进行泵站机组泵的性能预测的新方法,并以16CJ80型全调节轴流泵为例,进行了泵的性能预测,经济、可靠地获得了泵的性能曲线 简述了该BP神经网络所存在的缺陷及其改进的有效手段 这一技术的成功应用提高了泵站机组的可靠性、运行质量,同时也推动了神经网络等新技术。

Based on the analysis and study of BP neural network model, a new method for predicting the performance of the pumping station system is presentedThe performance of a 16CJ80 series fulladjustment axial pump is studied as an example and its characteristics is economically and reliably obtained. The effective ways to improve the defects of the BP neural network is summarized The successful application of this technique improves the reliability and running quality of the pumping station system