Font Size: a A A

Study On Application Of System Identification Based On Neural Network In Hydropower Plant Simulation Training System

Posted on:2008-12-06Degree:MasterType:Thesis
Country:ChinaCandidate:Y J WangFull Text:PDF
GTID:2132360215989711Subject:Electrical engineering
Abstract/Summary:PDF Full Text Request
With fast developing of computer technology, automation level in the run of hydropower plant is continually improving. Simulation training is very effective in improving the ability of operators and automation engineer. This dissertation focuses on the application of neural network in simulation training system.Turbine is a nonlinear and non-minimum phase system, and it is hard to build an accurate model for it. This dissertation points out the defects of the widely used simple linear model, improve the present model. This dissertation builds the model of different arithmetic digital PID controller by dispersing it in order to get the model more similar to actual system. This dissertation establishes single hydro generator set model of active power closed loop control system. Through the simulation, the model's validity is proved.The structure and parameter between different hydro generator sets are not the same, so it is difficult to describe it with one model and to determine the parameter of it. This dissertation uses neural network identification to get the model of hydro generator sets. Trained with the data from the nonlinear model, this dissertation uses neural network to identify the hydro generator set. In the training process, the adaptive learning rate and error batch-mode process are introduced to shorten the convergence period. The supervised training alleviates over-matching. This dissertation establishes closed loop control system with neural network model identified. This dissertation introduces the operational method in MATLAB of system identification based on neural network in detail.This dissertation designs the graphical user interface of simulation training system with VC and the program by which the user interface can exchange data with MATLAB program. It get easy to embed the simulation of hydro generator set active power closed loop control system into the configuration software of hydro power plant combined automation control system which is designed with VC. A user can finish the simulation of hydro generator set active power closed loop control system and get the simulation response just by inputting the parameters of PID controller.
Keywords/Search Tags:Neural network, System identification, Hydro generator set, PID control
PDF Full Text Request
Related items