基于大数据技术的雅砻江流域水电站安全研究

In recent years, the Yalong River Hydropower Development Company urgently needs to develop the signal processing and the data analyzing technologies related to the security of hydropower stations and the river basin in order to effectively improve the security level of the whole basin and the prominent economic benefits. Through close cooperation with the Yalong River Hydropower Development Company on the core scientific problems relating to the hydropower safety, the present project is intended to carry out research in the following aspects: 1. Based on the high-performance time-frequency signal analysis and processing methods, the in-depth analysis of the historical operational data of hydropower units in the Yalong River Hydropower Development Company will be carried out, and the primary influencing factors for the safe operation of the units will be identified and quantitatively evaluated. 2. Relying on the data analyzing technologies, the underlying correlation among the signals from different sources of hydropower units will be revealed together with its potential influences on the unit safety for the purpose of enhancement of the safety management levels of the whole company.

近年来，雅砻江流域水电开发有限公司迫切需要发展涉及电站及流域安全的信号分析和数据分析技术，从而有效提升全流域的安全水平并获得重要的经济效益。本项目拟围绕雅砻江流域水电机组安全运行所涉及的核心科学问题，与雅砻江流域水电开发有限公司合作开展以下研究：（1）采用高性能的时频信号分析与处理方法对该公司水电机组的历史运行数据进行深入分析，识别对机组安全运行具有重要意义的核心影响因素并进行定量评价；（2）依托数据分析技术，深入挖掘电站机组信号所蕴含的丰富信息及其对机组安全的潜在影响，从而全面提升该公司的安全管理水平。

对水电机组开展运行状态监测和大数据智能分析，可以提高机组的运行可靠性、安全性和稳定性，从而提升我国水电站的智能化建设水平。本项目结合先进的信号处理、深度学习和人工智能算法，对水电机组运行状态监测、信号特征提取、故障诊断和非线性信号预测进行了研究。研究成果如下：揭示了水电机组启动过程中的流激振动以及压力脉动特性；构建了适用于水电机组摆度信号降噪、轴心轨迹识别、振动故障诊断以及非线性振动信号预测的智能分析流程及模型，并采用水电机组实测信号进行了验证分析。