Research On Grid-connected Optimization Of Distributed Photovoltaic Power Generation Based On Improved QPSO

With the continuous growth of global energy consumption and the increasing environmental pressure,distributed generation technology has achieved increasingly strong development and application.Among them,the photovoltaic industry benefits from government subsidies and policy support to develop particularly rapidly.In particular,China’s Xinjiang,Inner Mongolia and other areas with abundant sunshine and sparsely populated areas have the natural advantage of developing photovoltaic power generation technology.However,due to climatic conditions,seasonal changes and other aspects,distributed photovoltaic output has volatility and randomness,so distributed power is not fed into the grid as a dc power source of constant power.The different access modes of DG will cause different degrees of changes in the distribution network structure,and the impact will be very different.Therefore,how to reasonably plan the number,location and capacity of its grid connection,and give full play to the beneficial impact of distributed photovoltaics on the distribution network,is the focus of this paper.In this paper,the mathematical model of the solar cell is first constructed,and the IU characteristics of the battery are analyzed according to the mathematical model of the battery,and the output characteristic curves at different irradiation intensities and different temperatures are simulated.Then,taking the distributed photovoltaic access radiative distribution network as an example,the mathematical model analysis of distributed photovoltaic access under different numbers,locations and different grid-connected capacities can obtain different impacts on the power flow trend of the distribution network under the corresponding mode.Photovoltaic access can often raise the voltage of the parallel point and the surrounding low-voltage nodes to improve power quality and stability;but if the capacity is too large,the load of the access point cannot be fully consumed and may exceed the upper limit.The increase or decrease of active network loss is also likely to have a positive or negative two-way effect.Secondly,this paper constructs a distributed PV output clustering timing model that comprehensively considers geographical location,climatic conditions and seasonal characteristics.Taking Weng Niu Te Banner as an example,considering that the daytime irradiation intensity has time-varying characteristics,if the unit of 0.5 hours is used,the annual cycle of 365 days is used,there are 17520 data samples,and there are 17520 photovoltaic output scenarios,and it is not feasible to plan too complex.Therefore,this paper uses the DBSCAN clustering method to cluster the above scenarios to reduce data redundancy,and obtain the corresponding Weng Niu Te Banner photovoltaic output representative scenarios.Then,this paper proposes an improved quantum particle swarm algorithm(QPSO)and constructs an optimization model for distributed photovoltaic grid connection with the optimization goal of minimizing active network loss based on this improved algorithm.In the basic iterative search process of the traditional particle swarm algorithm(PSO),the weighted idea is combined to find the optimal position of the average individual mbest^’,and the elite particles are considered to have more discourse power to influence the overall particle swarm update.And update the external storage of the pareto solution set with crowded distance ordering to save storage units and improve global search capabilities.The improved QPSO algorithm is then tested with four test functions to verify its optimization ability and global convergence.Finally,taking the IEEE-33 node system as an example to simulate the MATLAB platform,the improved QPSO algorithm has reached a minimum value of 65.93 k W in the 27 th iteration,and the average node voltage has been increased to 0.9845 p.u..In order to further verify the feasibility of the algorithm for the actual distribution network,this paper simulates the simplified model of the 10 k V distribution network construction of Weng Niu Te Qi and uses this as an example to simulate and verify.The output results show that the photovoltaic access point and access capacity planned by the improved QPSO algorithm have reached a minimum value of 78.19 k W in the 22 nd iteration,and the average node voltage has increased to 0.9860 p.u..The above two simulation outputs prove that the grid-connected optimization model of the improved QPSO algorithm constructed in this paper has significant optimization performance in terms of loss reduction and voltage quality improvement.