电力物联网中大型变压器健康状态在线监测关键技术研究

With the development of modern power systems towards higher voltage and larger capacity, the requirement of reliability in power grids is becoming more and more high, which makes it become an important measure to real-time on-line monitor power devices through Internet of Things (IoT). This project will research key technologies in multiple-sourse state information collection and health state eveluation for large transformers, critical devices in power IoT. The research plans mainly include the following four aspects. First, it will present a sensor placement method in order to optimize the placement of sensors by analyzing the multiple physical field distribution during working and typical faults. Second, it will build the real-time task scheduling method and bus communication method in order to meet the real-time requirements of on-site uploading nodes and the timing predictability of field bus transmission respectively. Third, it will propose a multiple-source information fusion method at data levels, which will build the information association between on-site collection data and keep the temporal and environmental correction among information. Fourth, it will research a health state evaluation method with wide domain associate characteistics, which will analyze the health state of large transformers according to the space and time associated information. This project is to provide theoretical achievements, algorithms, and implementation technique for the large critical devices in power IoT. Furthermore, this project helps to improve the reliability and maintainability of power grids as well as reduce running and maintenance cost.

随着现代电力系统向高电压、大容量方向发展，电网可靠性的要求越来越高，使用物联网进行实时在线监测成为保证电网稳定可靠运行的重要手段。本项目以电力物联网中的关键设备大型变压器系统为研究对象，研究通过使用多源感知获取状态信息和评估其健康状况的关键技术。本研究包括传感器的布设方法，通过分析变压器正常工作和典型故障情况下多物理场的分布情况，以优化传感器的布设；实时的任务调度和总线通信方法，以保证现场回传结点中任务的实时性及现场总线传输中的时间可预测性；数据级多源信息融合方法，以建立现场采集的数据间的信息关联，保持信息的时间和环境正确性；和广域关联的健康状态评估方法，以根据空间和时间方面的相关信息分析变压器所处于的健康状态。本研究为电力物联网大型关键设备的在线实时监测提供理论成果、算法和实现技术，有助于提高电网的可靠性、可维护性并降低运维成本。

本项目针对电力物联网中的关键设备大型变压器系统的安全可靠工作问题，研究使用在线监测的方法实时获取其工作状态信息，以实现按需维护并准确地评估其健康状况。本研究分析了变压器工作过程中电磁场、声音、温度和振动等可以在体外实时探测的信号与变压器的故障之间的关系，得出了可以在优化的位置布设声音、温度和振动传感器，以监测变压器的铁芯、绕组、内部温度变化等异常；提出了一种CAN总线上消息的时隙分配和调度方法，保证了传感器探测数据传输的实时性及时间可预测性；定义和分析了探测数据在位置、时间和状态方面的关联性，从而为变压器的健康状态分析提供了依据；最后，提出了声音、温度和振动的故障树分析方法，从而可以实现实时的故障发现和维护，并提出了一种结合变压器的基础指标、油气指标和在线监测指标进行变压器健康状态评估的方法。本研究为电力物联网大型关键设备的在线实时监测提供理论成果、算法和实现技术，有助于提高电网的可靠性、可维护性并降低运维成本。