Research On Grid-Connected Capacity Of Distributed Photovoltaic Considering Random Characteristic Of Power Quality

With the new power system with new energy and the implementation of a series of national incentive policies,the installed capacity of distributed photovoltaic(PV)with the advantages of cleaning,efficiency,economy and flexibility increase rapidly.If the integration capacity of distributed PV exceeds the actual hosting capacity of the distribution network,problems such as loss rising,voltage overlimit and harmonic pollution are caused.It threatens the safe and economic operation of the distribution network.In addition,distributed PV is affected by meteorological conditions such as light.The characteristics of intermittence and randomness are highlighted.The distributed PV power with a close geographical location has a spatial correlation.The uncertainty of the distribution network is aggravated and the PV hosting capacity is affected.To guide the distributed PV to be connected in a reasonable and orderly manner,it is necessary to study the integration capacity of distributed PV.A more scientific and reasonable integration scheme is obtained.Therefore,the power quality of the distribution network integrated with distributed PV and probabilistic harmonic power flow(PHPF)calculation are studied in this paper.Then,the distributed PV hosting capacity is analyzed.The specific research contents are as follows:(1)Aiming at the power quality of the distribution network integrated with distributed PV,the influence mechanism of distributed PV grid-connected on voltage quality is studied based on the network collection method of the distributed PV firstly.In addition,aiming at the harmonic characteristics of the distributed PV,the generation mechanism of the harmonic is analyzed.The harmonic coupling admittance matrix models of distributed PV based on harmonic state space and measurement data are established.Then,the influence of distributed PV on the loss is clarified.The harmonic loss model of the distribution network based on the AC/DC resistance coefficient and harmonic loss factor is established.Finally,the power quality variation characteristics caused by distributed PV grid-connected are analyzed by the simulation and measured data.It provides theoretical support and data reference for the study of the PV grid-connected capacity.(2)Aiming at power quality probabilistic evaluation of the distribution network considering distributed PV correlation,the harmonic source characteristic equation and the harmonic network equation based on the network element model considering different harmonic frequencies are established firstly.The deterministic harmonic power flow is analyzed by the decoupling method.Then,the statistical characteristics of state variables such as fundamental voltage,harmonic voltage and harmonic loss of the distribution network are extracted by the improved multi-point estimation method is proposed.On this basis,the probability distribution function is fitted by the improved maximum entropy based on nonlinear transformation.Finally,the proposed PHPF calculation is verified based on IEEE 33-bus system.The results show that this method can evaluate the power quality probability state which has input random variables with correlation and non-normal distribution characteristics.It provides a basis for mastering the power quality random characteristics of the distribution network under different PV grid-connected capacities.(3)Aiming at the study on distributed PV hosting capacity considering the random characteristic of power quality,the node fundamental voltage under different distributed PV permeability is randomly simulated based on the Monte Carlo Simulation firstly.The maximum capacity of distributed PV is analyzed from the perspective of the overall distribution network.Then considering the chance constraint such as the node fundamental voltage and the total harmonic distortion of the node voltage,a calculation model aiming at improving the PV grid-connected capacity and reducing the loss is established.It is solved by an improved adaptive genetic algorithm.The IEEE 33-bus test system and a practical test system of a 10kV distribution network in Shandong Province are used to verify the effectiveness of the proposed method.The influence of chance constrained confidence level,PV correlation and statistical characteristics of PV power on the distributed PV hosting capacity are studied.The results show that this method can provide guidance for reasonable integration of the distributed PV and improve the power quality.