Intelligent Test Platform for Power Plant Relay Protection Based on Multi-Source Data Fusion

Power plants should carry out relay protection testing in a timely manner, but traditional testing methods have certain errors in the results to develop an intelligent testing platform for power plant relay protection based on multi-source data fusion. Design intelligent testing computers, testing signal encoders, and testing boards for key hardware implementation, and design interface implementation links between them; build a software framework for an intelligent testing platform for relay protection. Based on the content of this framework, use multi-source data fusion technology to identify, decompose, and fuse voltage data to obtain a voltage dataset. Detect abnormal sample data in the voltage dataset, and use voltage separation method to remove samples far from the basic threshold to obtain a new dataset. Determine the threshold for sample detection based on the confidence level of voltage samples within the basic threshold. Reduce the dimensionality of the sample set to obtain the compressed sample set, calculate the square of the power interval distance of each voltage sample in the compressed sample dataset, and distinguish samples with the square of the power interval distance greater than the threshold for sample detection as abnormal samples. Complete the design of the software armor and merge the software and hardware to achieve platform construction. The performance of this testing platform is determined by comparing the relative error between the actual measured voltage value and the set value. The experimental results show that the platform can accurately measure the data information of the power plant power system with small errors.