Font Size: a A A

Research Of Nonlinear Generalized Predictive Control For Pumped-Storage Power Units

Posted on:2017-03-01Degree:MasterType:Thesis
Country:ChinaCandidate:Y F MaoFull Text:PDF
GTID:2322330503972577Subject:Hydraulic engineering
Abstract/Summary:PDF Full Text Request
Pumped-storage power plant can take on the tasks of reducing peak and filling valley, emergency reserve, frequency modulation, phase modulation, etc. in power system, which is the key to maintain the security, stability and reliability of the power system. Thus, the power grid has higher requirements to the reliability of its controller system. The strong nonlinear of the pump turbine brings enormous difficulty to the control task of pumped storage unit. Traditional control strategy and control law have been unable to meet the control quality demand of pumped storage unit, it is difficult to fundamentally solve pumped storage power unit control objects strongly nonlinear, degeneration and the adaptation problems among linear control laws. Predictive control is a kind of control theory and method which can update the control object model online, it would be suitable to solve the problem of the control of pumped storage unit. Therefore this article attempts in the modeling and Simulation of complex model for pumped storage units, and then analyze the structure characteristics of the model, and focus on designing a nonlinear generalized predictive controller to solve the pumped storage power unit control problem. The specific research includes the following aspects:1. Building the simulation model of pumped storage units control system. In view of the complexity of the diversion system of pumped storage power station, the characteristic method is adopted to deal with the water diversion system. For the complete characteristic curves of pump turbine, the improved suter-transformation method has been used to eliminate the ‘S' characteristics, the uneven distribution, cross and aggregation of the curves, to obtain good interpolation effect, provide convenience for later simulation calculation for convenience. In addition, the mathematical model of the mechanical hydraulic actuator and synchronous generator is established. After the model of each module is completed, each model is integrated, and the simulation program is designed for the model of the pumped storage unit regulating system.2. Design the nonlinear generalized predictive controller based on the pumped storage unit regulating system. Depending on the characteristic of the controlled system, the transient linear model of pumped storage unit regulating system has been given. Because the model parameters linearized system is unknown, the design of the nonlinear generalized predictive controller is based on the least squares parameter identification method. In addition, for the system in the initial state or in strong interference may occur the model mismatch problem, a penalty coefficient is proposed. In addition, for the difficult to determine the initial value, the least squares method is improved by introducing a learning method. The performance of the designed controller is verified by a simulation experiment.3. Applied the nonlinear generalized predictive controller to the simulation experiment of the pumped storage power units. The simulation is conducted on MATLAB and Jiangxi Hongping pumped storage power plant has served as the experimental object. Experiments including starting-up disturbance experiment and no-load disturbance experiment, while the conventional PID control and Fractional Order PID control have been also introduced to compare the simulation results. The simulation results show that the nonlinear generalized predictive controller can well complete the start-up process and no-load disturbance process of pumped storage unit and has the gentle adjustment process, the robustness and stability is better and has a higher computational efficiency superiority.
Keywords/Search Tags:Pumped-storage power units, Regulating system, Nonlinear generalized predictive controller, Modeling, Simulation
PDF Full Text Request
Related items