Research On Power System Optimal Dispatch Model And Method Considering Demand Response

Energy is an important material foundation for the survival and development of mankind.The depletion of fossil energy and the change of environment are two major threats to human sustainable development.In response to these challenges,the power system is undergoing tremendous changes both on the generation side and the user side.On the generation side,the centralized or distributed integration of renewable energy like wind power generation and photovoltaic power generation is an important trend in the development of power system.However,due to the intermittency and randomness of renewable energy,the uncertainty factors in the optimal dispatch of power system are further increasing.On the user side,in the traditional schedule of power system,the user side is considered to be inelastic,mainly by dispatching the units to meet the load demand.The development of the smart grid,especially the advanced metering infrastructure,provides technical support for the users to participate in the power system operation.Demand response resources gradually enter the researchers' fields of vision.To counter the changes mentioned above,it is urgent to study the optimal dispatch and demand response schedule model under uncertainty.In this paper,with the study of demand response optimal dispatch model as the starting point,in which the uncertainty of system is considered,the optimal dispatching model of power system with demand response under uncertainty and the solution method are discussed.The concrete research contents are as follows:(1)The incentive based demand response is studied,and the optimal schedule of interruptible load(IL)and direct load control(DLC)is taken as the research emphasis.In analogy to the conventional unit commitment,the mixed integer mathematical model is constructed respectively.In the interruptible load dispatching model,the uncertain factors such as the conventional unit fault and the load forecasting error are taken into account in the scheduling process,and the scenario analysis is used to analyze.In the joint optimal scheduling model of IL and DLC,the uncertainty of load forecasting is considered and the optimal scheduling model is established based on stochastic chance constrained programming theory.The chance constraint is transformed into its deterministic equivalent,and lagrangian relaxation method is used to solve the model.Scenario analysis and stochastic chance constraint programming theory are important methods to study the optimal dispatch of power system under uncertainty.(2)The price based demand response is studied,and time-of-use pricing(TOU)is taken as the research emphasis.The integration of wind power increases the uncertainties in the system operation.As a strategy of demand response,time-of-use pricing can change the behavior of consumers on a long time scale,and promote the consumption of renewable energy.TOU can be used as the basis of other consumption means.TOU mainly involves the division and the pricing of the peak,valley and flat period.In the study of TOU,various uncertainties such as wind power output,load forecasting error and price elasticity of demand are taken into account,and Mont Carlo simulation is used to generate the corresponding scenarios.Clustering analysis is used on the system net load scenario set to optimize the division of peak,valley and flat periods.In order to ensure the accuracy and reduce the computational burden,simultaneous backward reduction method is used to reduce the scenarios in the process due to the large scale of generated scenarios.(3)Based on the chance constrained programming,the optimal scheduling model considering the fault probability of conventional units and lines,load forecasting error,interruptible load default probability and line power flow restrictions is built to further study the modeling and solving method under uncertainties.The first part of the model objective function minimize the sum of the generating cost,the start-up cost and the compensation of IL without considering stochastic factors.Its solution is the basic dispatching quantity of each unit and interruptible load,and the sum of basic quantity should meet the forecasted demand of the system.The undispatched capacity and the uninterrupted part of IL can participate in the optimization of reserve.In the second part,the sum of the expected cost of unit reserve cost and the compensation of IL should be the smallest when the stochastic factors are taken into account.Because the chance constraints in the model can not be directly transformed into the deterministic equivalents,the Monte Carlo simulation is used to transform the chance constrained model into the expected value model with multiple deterministic scenarios.The stochastic factors are eliminated and the model is transformed into the deterministic one.(4)Since distributed generation and demand response are widely distributed,varied in types and different in parameters,it is difficult for the dispatching center to dispatch a large number of distributed energy resources directly.Virtual power plants can be used to integrate distributed energy resources and participate in the electricity market as a whole.In the model,the internal output uncertainty and load forecast errors of virtual power plant and uncertainty of the electricity market competitors are taken into account.The virtual power plant takes the maximum benefit as the goal and takes part in the power system operation as a whole.The dispatching center will proceed the market clearing with the goal of maximizing social welfare after receiving the biddings of each virtual power plant and conventional unit power plant,and the virtual power plants and the conventional unit power plants will adjust the internal resource operation after receiving the instructions of dispatching center.