Study On Parameters Identification And Control Laws Of Hydraulic Turbine Regulating System

An accurate mathematical model of hydraulic turbine regulating system(HTRS)not only provides the basis of hydro-generator unit simulation and hydraulic transition process,but also plays an important role in power system's state analysis and calculation,stability evaluation and controller optimization design.At present,the hydro unit are developing towards to the trends of large-scale and complex-structure,the built and building hydro station in China are with large capacity generator,long penstock,big water interia and hydraulic-electrical coupled properties.These properties bring the difficult to the modeling and controlling of HTRS,which cause great challenges for the safety and stability of power system.In this context,the characters of the HTRS are deeply analyzed on the basis of its working principle,and as a result the precise linear and nonlinear model of HTRS are developed according to the concept of different submodel modular.Based on these models,the scientific problems of parameter estimation,controller design and parameter tuning of PID controller has been proposed.Combing the evolutionary algorithms based estimator,unknonwn parameter observer(UPO),synchronization parameter observer(SPO),multi-objective optimization controllers,linear sliding mode and terminal sliding mode controllers,this paper studied deailed parameter estimation,controller gains tunning methods and advanced controller designs for the hydropower system,and built the basic framework of paremeter identification and controller design for HTRS system with respect to linear and nonlinear mathematical models.The main contribution and innovations of this research is organized as follows.(1)Based on modularization modeling method,we devided HTRS system into four parts,namely conduit subsystem,water turbine,generator and governor.The different model of each subsystem is constructed,and its merits,shortcomings and available applications are then analyzed.On the basis of these subsystem models,some integrated linear and nonlinear models of the HTRS are proposed,which provides a foundation for parameter identification and control rule design of hydropower plants and HTRS system in the following sections.(2)To estimate the unknown parameters of a specified linear model with respect to HTRS system,this study introduced a general intelligence algorithms based frame network,which converts the problems of parameter identification of the HTRS into solving the extremal optimization problem with a predefined error fitness function.Focus on the characters of multi-input and multi-output,time-varying and non-minimum phase of HTRS,an improved gravitational search algorithm(IGSA)approach is proposed and used to address the problem of unknown parameters in HTRS.Simulations under different working conditions show the validity and effetiveness of the proposed method.(3)In accordance with the inherent structural properties of a nonlinear HTRS model,this paper presents three new parameter identification methods,namely ant lion optimizer based estimators,unknown parameter observers and synchronization parameter observers.Among them,the latter two observers are the first applicaiton of modern observer technical methods in solving parmater identification problem of HTRS model.By carrying out experimental simulations on an identical model,we compred the differences of three proposed approaches in terms of the orders,the structure,the independence and the results of identification system,and analyze their advantages and disadvantages.At last,we discussed the application scope of each identification method.(4)Focusing on the single-objective evolutionary algorithm based optimal PID controller gains tunning method cannot satisfying the need of HTRS with multi-condition and time-varying parameters,this paper proposed multi-objective evolutionary algorithm based optima PID gains tunning methods,and introduced them into the application of linear HTRS model.In order to improve the performance of the controller further,we extended integer order PID controller to fractional order PID controller,and applied it into HTRS.Furthermore,in order to overcome the discontinuous control output saturation of the govenor,a hyperbolic tangent function approximation modeling method is used.(5)Considering that PID control law is not suitable for the complex nonlinear HTRS model,some sliding mode controller(SMC)laws were illustrated in this paper.In order to decrese the inherent chattering phenomenon in SMC,fuzzy sliding mode controller(FSMC)and intelligence algorithms based FSMC controller for the nonlinear HTRS models were proposed.In order to overcome the asymptotic convergence of conventional sliding mode contollers,we introduced the concepts of terminal sliding mode controller(TSMC),and proposed the traditional TSMC and fast TSMC controllers for the nonlinear HTRS models.Besides that,a satutaion compensator and dead zone approximator were also established in the TSMC controllers for the limitiaons of the controllers in HTRS.