Research On Model-Based Fault Diagnosis Of Steam Turbine Governing System

With the development of the living standard and industry level in our country .The demand for electric power is increasing, the bigger unit capacity is required. As a result of the structure of equipments was getting more and more complicated. In order to ensure the safety and success of the production process, the part of the control system of turbine is becoming more and more important; the higher turbine governing system reliability is needed. Therefore, fault diagnosis for steam turbine governing system is essential. This paper against the whole turbine governing system, takes the digital electro hydraulic control system (DEH) which is widely used in large-scale units as the research object.Firstly, the structure of the DEH system is introduced and analyzed. The mathematical models of controlled plant and executive body in DEH system is set up according to their working principle.Various kind of fault phenomenon might be appeared in governing system is analyzed. DEH hydraulic components ' jam fault is found as the cause of the multiple kinds of faults. A meticulous analyzes of jam fault is made. The model of jam fault is set up and simulated in the MATLAB. The methods of the jam fault's diagnosis are summarized. With the compare of the advantages and disadvantages, the thought of identification by assuring system parameters estimation is chosen.In consideration of system's strongly nonlinearity and the specific characteristic of the DEH system with jam fault, identification diagnosis base on genetic algorithm is determined to choose.The principle and operation process of genetic algorithm is introduced, and the early maturity phenomenon might happen between calculation is identified. Aimed at maturity's overcome adaptive multiple population genetic algorithms is calculated. At last, the jam fault diagnosis is simulated. The result is analyzed. The effectiveness of the fault diagnosis method mentioned above is verified.