Research And Implementation Of Intelligent Substation Condition Monitoring Software Platform Based On Data Mining

Intelligent substation system produced by a variety of advanced technology and facilities now is playing an important role in today’s society. Especially the system of intelligent substation monitoring is widely used in large-scale intelligent substation building and rebuilding. Due to the lack of uniform technical standards and unified platform, resources cannot be shared effectively during the traditional substation monitoring system. Besides it cannot play its role in diagnosis effectively because of online monitoring data cannot combine with other important state amount. So it is necessary to develop a new intelligent substation monitoring software platform, it can play an important role in fully automated data collection, measurement, protection, control and testing, also it can improve the operational safety of substationData mining techniques has great advantages in the massive data treatment in the power system. Applying data mining method in system can play a role in system monitoring and diagnosis during the data analysis and processing. In view of this, based on the data mining technology, structured software engineering development method, takes IEC61850 communication protocol as the uniform communication standard, established the new intelligent substation monitoring software platform.Through the design of system of substation monitoring and control software, based on functional and data analysis requirements, core modules of the system is well analyzed. Carrier wave and serial communication are designed in detail. Also data mining module based on BP neural network and real-time database system are conducted in-depth research. Firstly, the communication module of the system is analyzed in depth, such as the creating of thread functions, data collecting and analysis and the realization of network communication. Secondly, data mining process of module is discussed from several aspects such as samples normalization, training of BP neural network algorithms and simulation experiment. Finally, the paper illustrates the concrete application of the substation monitoring software platform.