Research On Frequency Regulation Strategy Of New Energy Power System Based On Distributed Model Predictive Control

In the modern power system,the frequency stability is highly associated with the power quality of the grid.The real-time changes in load demand can cause fluctuations in the frequency and result in deviations of the inter-change power between generation areas.Load frequency control(LFC)plays a key role in guaranteeing the frequency stability of the power system.Its major task is to adjust the grid frequency by regulating the output power of the generation units,and to keep the inter-change power around the scheduled value.In line with the "dual carbon" target,the new power system is featured by the high penetration of new energy and the heavy integration of power electronic equipment.Compared with the traditional power system dominated by synchronous generators,the new power system is characterized by low inertia,weak damping,and poor frequency support capability.With the increasing penetration of new energy in the power grid,the power system has become a geographically dispersed large-scale uncertain system.The randomness and intermittency of new energy will seriously affect the frequency stability of the power system,making it increasingly challengeable to apply the load frequency control in the power system.The distributed model predictive control(DMPC)has emerged as a promising strategy for LFC problem in recent years.It decomposes the integral control system into multiple subsystems and achieves the LFC objectives of the entire power system based on the information exchange between local models,which can ensure the global control performance while reduce the online computational burden.In the meantime,DMPC strategy can effectively utilize the new energy to realize frequency regulation,therefore reducing the dependence on traditional coal-fired power generation and enhancing the environmental friendliness of the power system.Aimed at the large-scale interconnected power system with wind power penetration,the frequency regulation strategies based on DMPC are studied in this thesis,where the advanced tube-based robust control and the economic model predictive control are incorporated into DMPC for improving the reliability,stability,and economy of power system operation.The main contributions are as follows:(1)The load frequency control model for conventional generation units including thermal power units and hydro power units is established based on the mechanism characteristics.The virtual inertia control unit is introduced into the wind farm for improving the inertia response capability of wind turbines.On this basis,the load frequency control model for the wind farm is established.which enables the wind farm to directly participate in the frequency regulation of the power system.(2)A two-layer robust distributed model predictive control strategy is proposed for the power system where wind power and hydro power are settled in the same region.The control objectives of tracking the load change and rejecting the wind power disturbance are realized in the two layers of the robust DMPC separately.The inner-layer MPC tracks the load variation while ignoring the disturbance,and the outer-layer MPC suppresses the effect of wind power disturbance using a disturbance observer to estimate the real-time wind power signal.This two-layer control structure can effectively improve the control smoothness while reducing the frequency fluctuation of the power system.(3)A tube-based coordinated distributed model predictive control strategy is proposed for the power system with high wind power penetration from an independent region.The controllers of each region solve their local nominal optimization problems in parallel under a decentralized structure,and achieve coordination through a distributed feedback control law.Simulations show that the proposed tube-based DMPC control strategy can well coordinate the output power of each region and reduce the grid frequency deviation.Furthermore,the impact of uncertain wind speed disturbances on the whole power system can be effectively restrained.(4)For optimizing the dynamic economic performance indices concerning the wind turbine maintenance cost and thermal generation cost,a robust distributed economic model predictive control(DEMPC)strategy is constructed for the interconnected power system with high wind power penetration.The robust invariant set and the terminal inequality constraint are introduced to ensure the recursive feasibility of the nominal optimization problem and the stability of the closed-loop system.The simulation results under various intensities of wind speed fluctuation show that the proposed robust DEMPC can evidently improve the dynamic economy of the power system while significantly enhance the robustness against wind speed disturbance.