神经网络在大坝安全监控数据分析系统中的应用

Application of Neural Network in Dam Security Monitor System

研究神经网络在水电站大坝安全监控工程观测数据分析系统中的具体应用。基于水电站大坝安全监控项目的实际需求,以BP算法为基础的前馈神经网络用于东北水电站大坝安全监控数据分析软件系统中,通过与多元线性回归和逐步回归处理方法相比较,神经网络方法能够有效地提高观测数据的拟合精度和预测质量,使用数据分析软件系统对东北某水电站的历史数据处理,结果证明了神经网络用于水电站大坝安全监控项目数据分析具有独特优势和深远的研究前景。

This paper studies on the application of NN（ neural network） in the observation data analysis system security monitor project of the hydropower station dam. Based on the practice requirement of the project, FNN（feedforward neural network） based on the BP algorithm is applied to the security monitor data analysis software system of hydropower station in the northeast. By comparing the multiple linear regression and step-by-step regression, it proves that NN can improve the precision and predict quality of the observation data effectively. The result of using data analysis system to handle the bypast data of a hydropower station in the northeast shows that the application of NN to the data analysis system may have idiographic superiority and far-reaching research foreground.