Research On Modeling Of Optimal Economic Operation Of Active Distribution Network And Solving Algorithm

Optimal operation of the active distribution network is one of the key technologies to achieve the economic and reliable operation of ADN.In ADN,the control variables have both continuous variables(output of distributed power generation unit,charge and discharge power of storage,etc.),and discrete variables(regulator tap position,switching status of switchable capacitor banks,etc.);both contain linear constraints(power upper and lower bound constraints,etc.),but also contains nonlinear constraints(power flow equality constraints,nodal voltage constraints,etc.).Therefore,the optimal operation of the active distribution network is a complex mixed integer nonlinear programming problem(MINLP).Because the solution based on the traditional interior point method is not ideal,the derivate-free intelligent algorithms with stronger adaptability are widely used.However,although the intelligent algorithm is widely researched and applied in the optimal operation of the active distribution network,there are still the following problems: a.lack of effective constraint handling mechanism,resulting in reduced efficiency of the solving process;b.due to the need for intelligent algorithms,a large number of power flow simulations,especially for three-phase modeling of the complex distribution network,reducing the speed of the solving process.Based on the above research background,this thesis has carried on the thorough research to the optimal operation model and the solving algorithm of the active distribution network,the main work is as follows:(1)In this paper,the optimal economic operation model of active distribution network is established considering typical active distribution network controllable resource equipment such as distributed power generation unit,energy storage equipment,voltage regulator and switchable capacitor bank and interruptible load.The model takes the lowest total operating cost of the system as the objective function during the whole scheduling period.(2)A constraint handling method based on particle swarm optimization algorithm is proposed.In the iterative process,the direct intervention of the particle position makes the particles in the population as far as possible from the infeasible domain into the feasible domain.This method improves the search ability of particle swarm algorithm in the feasible domain and speeds up the convergence rate.(3)A hybrid solution HAK-MFPSO combined with the Kriging surrogate model considering the constraint handling mechanism and the modified fuzzy particle swarm optimization algorithm is proposed.The power flow is calculated by the constructed Kriging surrogate model,which greatly accelerates the calculation speed.In order to guarantee the approximation accuracy of the Kriging surrogate model,a new dynamic update method of Kriging is proposed.