Research Of Parameter Identification For DEH And Designing Of Test Platform

Steam turbine digital electro-hydraulic servo control system (DEH) isthe key to the normal operation of turbine generator, how to improve thestability, reliability and control accuracy of DEH is persistent work andgreat challenge. Moreover, the hydraulic executive part (electro-hydraulicservo valve, oil motivation, etc.) is a key part concentrated withnonlinearity, the control quality and dynamic characteristics not only affectits own economy and safety, but also have a deep impressing on theoperating safety and stability of the power grid, enhancing the gridfrequency quality. Therefore, accurate and reliable modeling about theelectro-hydraulic servo control is necessary, it is important to undertakesome relative research about parameter identification.In this paper, firstly, we analysis the model, and establish a nonlinearmodel including dead zone, and through simulation, we analysis the effect,verify the feasibility. In order to eliminate the static error brought by thenonlinear link and to simply the experiment and improve the precision, wedevelop the software and hardware for the digital servo control testplatform, through USB means to realize the upper and down a machinedata communication, through the simple PC operation produced testrequirements excitation signal.Secondly, in order to achieve the identification accuracy and reliabilityrequirements, in this paper, the identification algorithm, a further study ismade about the particle swarm optimization algorithm (PSO). Since it mayeasily get trapped in local optima when solving complex problems, amulti-particle swarm optimization (MPSO) method based on double-layerevolution is studied in detail, besides we also make improvement about theinertia weight, the simulation results show that the improved particle swarm optimization algorithm achieves the requirements about parametersidentification in nonlinear models.Finally, we discusses how to use the improved particle swarmoptimization algorithm (PSO) in cite parameter identification, at the sametime, we make some research about the excitation signal, the condition andmethod of how to do the experiment, and the step signal can conform tothe requirements of the parameter identification, the possible error sourcesis also take into description.In conclusion, we mainly discuss the parameter identification of DEH,during modeling, we take nonlinear link into consideration and analysis theinfluence bought by nonlinearity. The digital servo control test system isdeveloped for the cite identification. Besides we applied an improved PSOinto the research, the simulation and experimental results show theaccurately about the established model and parameter identificationmethod.