| Gas turbine as a new power technology, has represented a nationalindustrial development level, and plays an extremely important role innational economy. Condition monitoring and fault diagnosis of gasturbine, refer to monitoring various parameters of the gas turbine workingstatus, passing the data information converted to electrical signals orother physical quantity signal, processing and coming to the conclusionweather working state is normal, put forward the specific technicalcontent of maintenance, thus units can reliably work, reduce faults andmaintenance costs.In this article, the design and application of gas turbine faultdiagnosis system is divided into three parts, as follows:In the first part of the gas turbine condition monitoring, using thehypothesis test of mathematical statistics, through the real-time datacomparing with fault-free operation standard under same workingcondition, t computing statistics and through the t distribution statisticalfigure, we determine whether the gas turbine run normally and test by anexampleIn the fault diagnosis of the second part, for the first level of faultdiagnosis, that is, a reverse solution from a sign of turbine operation togas turbine characteristic parameters’ change value. Nonlinear BFGS optimization technology is used, with comparing with the linear solution,the steepest descent method and neural network method, to analyze theirrespective advantages and disadvantages. Then the paper describescomprehensively the current fault diagnosis of the second layer: patternrecognition, that is, a solution from the change of characteristicparameters value to actual fault type gas turbine. Through the analysis ofthe characteristics and application methods of all kinds of method, weacquire the advantages and disadvantages of various methods.In the third part of the assembly and debugging of hardware andsoftware, the virtual instrument technology is introduced in theapplication of engineering practice, and the former two partsprogramming processing, assembly into a gas turbine fault diagnosisinstrument, includes data collection, data processing, optimization of faultdiagnosis and result display modules. |
