Reactive Power And Voltage Optimal Control Of Distribution Network With Renewable Distributed Generation Participating In Regulation

In recent years,with the large-scale distributed clean energy and diversified load access to the distribution network,its operational reliability and security have been challenged.The high-ratio RDG decentralized access brings about the operational state time-varying and unbalanced load distribution,which needs more effective coordination of multi-temporal scale optimization and control schemes to suppresse the voltage stability problem caused by uncertain factors and ensure the rapidity and precision of optimal control.Therefor,on the basis of traditional reactive voltage regulation and combined with new equipment access to expansion research ideas,the reach on " reactive power and voltage optimal control of distribution network with high proportion of renewable distributed generation participating in regulation" is carred out.Make full use of varieties of controllable sources in the active distribution network to achieve coordinated control and friendly interaction of photovoltaic,wind power,distributed energy storage,electric vehicles,and flexible loads,which can improve the quality of power supply and improve the reliability of distribution network operations.The main research contents are as follows:(1)A multi-time scale dynamic reactive power optimization model and method for distribution network with RDG participation is proposed.Aiming at the voltage stability problems caused by randomness of RDG output and load fluctuation in power distribution system,a multi-time Scale reactive power optimization model for distribution network based on model predictive control is established,which includes a day-ahead optimal regulation layer and a real-time rolling control layer.the day-ahead optimal regulation focuses on the economics of operation,large-Scale reactive power regulation is performed through the coordination and cooperation between different operating characteristics of reactive power devices,and sufficient reactive power reserves are reserved to respond to the dynamic regulation requirements;the real-time rolling control focuses on the stability of the system operation and is implemented based on the ultra-short-term forecast of wind power and load.According to the demand of reactive compensation,timely correction of decision-making feedback is implemented so as to realize forecasting information and control methods layer by layer,which can restrain voltage limit caused by uncertainties.(2)A real-time reactive power optimization method based on dynamic partitioning for distribution network with RDG participation is proposed.Most partitioning methods need to give the number of partitions in advance or give a parameter threshold to determine the number of partitions.Furthermore,the reactive power margin of each power supply in the system needs to be considered during the partitioning process.It is difficult to realize real-time dynamic partitioning without human intervention.To solve the above problem,combined with the dynamic reactive voltage regulation characteristics of renewable power sources such as wind power and photovoltaic,a dynamic partitioning method of real-time reactive power optimization for distribution network is proposed.Based on the radial structure characteristics of the distribution network,the initial segmentation is based on the nature of the branch end nodes,and the initial partition is performed with the maximum modularity function as the metric.The partition adjustment is performed according to the real-time operating status and the reactive reserve constraint,thus forming the optimal partition scheme and then performing partitioning real-time adjustment of the voltage.(3)A multi-time-space-scale reactive power optimization model and method for distribution network considering network dynamic reconstruction is proposed.The locality of voltage distribution,diversity of load/power characteristic and pulsation of active power determine that the voltage and reactive power regulation of the distribution network needs "multi-level coordination and step-by-step refinement" in the time domain and the spatial domain.A multi-time-space-scale reactive power optimization model for distribution network is established.The coordination of"global collaborative optimization","zonal self-regulation" and "global coordinated regulation" is carried out in the spatial domain,and network dynamic reconstruction is integrated into the global collaborative optimization.Optimize the "feasible domain";perform dynamic partitioning and self-regulation based on the modularity function and dynamic reactive reserve index to achieve "partition autonomy";the partition autonomy control results return to the global level for system-wide detection and coordinated control to improve control accuracy;In the time domain,the global synergistic optimization is Id cycle 1h granularity,the partition self-regulation is 5min granularity,and the model predictive control "predictive model","rolling optimization" and "feedback correction" is embedded in each layer to achieve space-time coupling the "partition layer by layer" optimization control.(4)The active/reactive power joint optimization model and method of distribution network with“source-network-load-storage”interaction in power market environment are proposed.It is an effective means to improve the absorption of renewable energy and to ensure the safe and economic operation of distribution network by breaking the barriers of relatively independent sources such as wind,photovoltaic and storage horizontally to realize multi-energy complementarity and by carrying out "source-network-load-storage" multi-link,multi-dimensional active/reactive joint scheduling and coordinated control at a deep level.A multi-time scale active/reactive joint optimization model for active distribution network with"source-network-load-storage" interaction in the electricity market environment is established,and the layer-by-layer progressive control is realized by stochastic model predictive control.The active/reactive controllable source and the communication status of tie switches are arranged in the daily optimization scheduling based on the forecast of electricity price and RDG output to guarantee operating reliability as well as maximize the profit.The profit space is further explored and the voltage limit is eliminated in the intraday rolling optimization based on the short-term forecast information.Operating safety is focused in real-time feedback correction,and voltage fluctuations are suppressed based on ultra-short-term prediction information for more refined control.(5)The active/reactive joint optimization model and method for distribution network with CSSIS participation are proposed.As a new type of load,electric vehicle has certain randomness in space-time distribution.Its disorderly charging management easily leads to "peak plus peak" phenomenon during peak load periods,which may increase the load peak-to-valley difference and the risk of local voltage over-limit.A double-layer active/reactive power joint optimization model for distribution network considering CSSIS participating in regulation is established.Focused on the distribution network level,the outer layer optimization emphasizies the CSSIS overall active and reactive power output plan to exert the functions of CSSIS peak load shifting and voltage dynamic adjustment.So the safety and economy of the distribution network operation is improved.The inner layer optimization focuses on the specific output plan in the CSSIS station to achieve energy compensation in the station,stabilize the EV load fluctuation,reduce the cost of EV electricity,and realize the cascade utilization of waste batteries.In order to reduce the regulation pressure of dynamic power supply in real time,the optimization method adopts robust optimization method to improve the system's ability to cope with uncertain factors and realize the“friendliness”of CSSIS access to the distribution network.The above research results can provide theoretical support for the research and development of the distribution system management platform of power distribution enterprises,and have broad application prospects in active power dispatching and reactive power optimization of power distribution networks.