Fault Warning of Hydro-Generator Based on Fisher Criterion

摘要

As the most mature technology and the most economical clean energy in the global energy supply, hydropower occupies an important position. There are many factors of safety and stable operation in the hydropower station, and equipment failure will bring huge economic losses, so the warning of equipment failure in the hydropower station has played a crucial role. This paper summarizes the principle of NSET modeling method for nonlinear state estimation. And the data is processed using feature selection based on Fisher criterion, then use NSET to simulate the hydro-generator, and select the characteristic state parameter of its key components as reference. When the residual value exceeds the pre-set, it issued a warning message, prompting the operator to check the equipment status.