Stochastic Dynamic Economical Dispatch Of AC-DC Interconnected Power Grid Based On Approximate Dynamic Programming Theory

With the successful development of ultra-high voltage and extra-high voltage transmission technology in China,and the rapid development of new energy power generation technology represented by wind power and photovoltaic under global environmental pressure and carbon emission pressure,the stochastic dynamic economic dispatch problem of AC-DC interconnected power grid is becoming more and more important to the security and reliability,cleanness and environmental protection,economy and efficiency,intelligence and openness of modern power grids.In this paper,the dynamic optimal dispatch of large-scale AC-DC interconnected power grid with new energy integration is studied,and the problem is divided into two parts: the optimal dispatching of the unit output in the provincial grid and the optimal allocation of the active power of the AC-DC parallel transmission channels among provincial grids.Firstly,the day-ahead scheduling of generation units of each provincial grid in the AC-DC interconnected power grid is being optimized.A virtual storage model of security-constrained stochastic dynamic economic dispatch problem of power system without pumped storage hydro stations is established.In this model,system's spinning reserve capacity is used to balance the random fluctuation of wind farm output and the change of load demand.The value function approximate dynamic programming method is used to solve the model,and this method realizes the decoupling of each scenario and each time interval of the stochastic dynamic economic dispatch model.Based on wind farms' power output forecast curve,error scenarios are generated by Latin Hypercube Sampling method.In order to train the value functions of each time interval,by scanning each error scenario,the quadratic programming model of each period should be solved successively.After getting the convergence value functions,the final generation dispatch scheme will be gained by solving the optimal dispatch model with the forecast scenario.At last,computational results on a real provincial power system demonstrate the feasibility and effectiveness of the proposed method by comparison with the scenario method and robust optimal dispatch method.After the optimization of power output scheduling of generation units of each provincial grid,a stochastic dynamic economic dispatch model for AC-DC parallel transmission channel's active power optimal distribution is established,and the state space approximate dynamic programming method is used to solve this model.The state space is obtained by using the three-stage method to solve the optimization model under the error scenarios,and the approximate optimal solution of solving the prediction scenario is obtained by evaluating the similarity of the solution between prediction scenario and state space from the decision variables,random variables,and objective function values.The economy of the system is further improved by optimizing the active power allocation of the AC-DC interconnected power grid's transmission channel,and the network loss of transmission channels is reduced.At last,test results on a real large-scale AC-DC interconnected power grid demonstrate the correctness and effectiveness of the proposed method.