Study On Nonlinear Model Parameter Identification And Control Optimization Of Pumped Storage Unit Governing System

With the transformation of energy structure from fossil energy to non-fossil energy in China,as an important renewable and clean energy,pumped storage plants undertake important tasks such as peak shaving and frequency regulation in power grid.In addition,in recent years,a large number of intermittent energy sources,such as wind and solar energy,have been connected to the power grid,and their fluctuation has seriously threatened the stable operation of the power grid.Pumped storage plants,as an important energy storage technology,can effectively suppress the impact of intermittent energy sources such as wind and solar energy in the power grid.Therefore,speeding up the construction of pumped storage stations have very important significance for deepening the adjustment of energy structure and improving the safe and stable operation of power grids in China.However,the complicated layout of the conduit tubes,the long distance diversion pipeline,higher heads,larger unit capacity of the pumped storage paltnt have brought about great difficulty to the modeling and control optimization of the pumped storage unit governing system.Therefore,it is urgent to explore new theory and technology to solve the problem of optimal control.The accurate modeling of the pumped storage unit governing system is the foundation of system dynamic process simulation,stability analysis,fault diagnosis,parameter identification and control optimization.As a complex time-varying,non-linear,non-minimum phase system,accurate modeling of the governing system has always been the hotspots and difficulty of related research.There sittl has some drawbacks when the traditional PID controller faces to the complex nonlinear system,the controller parameters optimization and the novel controller design are the effective tolls to improve the control performance of the system.Under this background,aiming at the key scientific problems and technical difficulties faced by the modeling and control optimization of the pumped storage unit governing system,in this paper,a systematic and in-depth study has been carried out.Based on mathematical modeling,system identification and intelligent optimization algorithm,a accurate model of governing system is established.On this basis,the fractional order PID control,controller parameters optimization,multi-objective control optimization,fuzzy model identification and generalized predictive controller design of pumped storage units are studied,and some theoretical and applied results have been obtained.The main work and innovative achievements of this paper as following:(1)Based on the analysis of complex dynamic characteristics of the pumped storage unit governing system,the mathmatical models of the all sections of the governing system are established.The complete characteristics curves of the pump turbine show typical “S” characteristics area in the turbine condition,the turbine braking condition and the reverse pump condition caused by its reversible design.In the region of “S” characteristics,there exist strong guide vane opening curves cross,aggregation and twisting phenomenon in the two ends of the curves,so one unit speed value corresponds to three unit flow values or three unit torque.The “S” characteristics has brought more difficulties for the modeling and control of the pumped storaged unit.In order to solve these problems,some methods have been proposed,in this paper the Logarithmic projection transformation and the improved Suter transformation are used to deal with the complete characteristics curves.According to different research needs,linear simulation model,non-linear simulation model and numerical simulation model of governing system are eatablished,which lays a model foundation for parameter identification and control optimization.(2)In oder to establish the accurate model of the pumped storage unit governing system and obtaiin its accurate model parameters,the theory of the model parameter identification based on intelligent optimization algorithm is studied.A Mixed-Strategy based Gravitational Search Algorithm(MS-GSA)is proposed in this paper,in which three improvement strategies are mixed and integrated in the standard GSA to enhance the optimization ability.The first improvement strategy is introducing elite agent's guidance into movement function to accelerate convergence speed.The second one is designing an adaptive gravitational constant function to keep a balance between the exploration and exploitation in the searching process.And the third improvement strategy is the mutation strategy based on the Cauchy and Gaussian mutations for overcoming the shortages of premature.The MS-GSA is employed to solve the parameter identification problem of governing system.It is shown that the MS-GSA is able to identify the parameters of governing system effectively with higher accuracy compared with existing methods.(3)A fractional-order PID controller parameter optimization strategy for the pumped storage unit governing system based on improved gravity search algorithm is proposed.A fractional-order PID controller is introduced to control the pumped storage unit instead of the traditional PID controller.The proposed gravitational search algorithm combined with the Cauchy and Gaussian mutation is proposed and used to optimize the fractional-order PID controller parameters.The experimental results indicate that the proposed strategies has significant advantages over other controller optimization strategies and has great potential to improve the control quality of pump turbine governing system under multiple water heads.(4)Due to the influence of external environment and working conditions change,the fixed PID controler parameters can not guarante the optimal operation of units under different working conditions.Based on the theory of the multi-objective optimization,the control performance indexs under different working conditions are considered as the multi-objective functions,the fractional PID controller and the multi-objective optimization theory are combined to establish the multi-conditions adaptive optimal control strategy.Results have also proved that the multi-conditions adaptive FOPID controller significantly improves the adaptability of governing system to the environment and operation conditions(5)This paper proposes an adaptive Takagi–Sugeno fuzzy model-based generalized predictive controller.First,the T-S fuzzy model identification is applied to approximate the dynamic characteristics of the established accurate pump turbine governing system.The T-S fuzzy model show higher modeling accuracy from the experiments results.Second,the adaptive T–S fuzzy model composed of the offline model training and the online consequent parameters adjustment is used to obtain the controlled autoregressive integrated moving average model(CARIMA).Then,the generalized predictive controller is designed for the pumped storage unit based on the CARIMA.Compared with the traditional PID controller,it can effectively suppress the frequency oscillation in the“S”characteristic region under the low head no-load condition,and has better control performance.