Intelligent Load Modeling Technology And Application Based On Power Big Data

In recent years,with the continuous operation of ultra high-voltage transmission(UHV)alternating current/direct current(AC / DC)transmission lines,the connection between power grids become increasingly close.High end measurement technology,information communication technology and automation control technology have been rapidly developed and widely popularized in power grid,forming a whole set of power big data,which provides massive data support for realizing online analysis of synthesis load characteristics.How to integrate all kinds of power grid data,take the power big data as the support,deeply excavate the effective power information,improve the accuracy and reliability of load model,and provide more reliable decision support for power grid intelligent dispatching,is an important step for the current load modeling work from theoretical research to practical engineering application.Based on the current theoretical analysis method of load modeling and the early engineering application and development results of the laboratory team,this paper first carried out the research on the load characteristics of a power grid industry and the composition of typical consumers’ electrical equipment,studied the composition characteristics of typical consumers’ electrical equipment,counted and calculated the load characteristics of four major industries in different seasons and different operation modes The dynamic load proportion of Hunan power grid provided a more real data reference for the determination of the dynamic load proportion of a typical industry in the synthesis load of a power grid,and provided a certain reference for the selection of the dynamic load proportion in the simulation calculation of Hunan power grid.Secondly,aiming at the difficulty of selecting typical consumers due to the lack of obvious power consumption characteristics,a machine learning based model was proposed Based on the historical data,the new daily load curves were screened,which improves the clustering quality,enhances the operation efficiency,and achieves the update of the typical consumers load characteristic database.Furthermore,aiming at the problem that the daily load curves has insufficient ability to describe the actual load characteristics,this paper put forward some suggestions.A classification and synthesis method of load characteristics of typical power industry based on consumers daily electricity consumption curves was proposed.Considering the instantaneous characteristics of sampling points and the average characteristics between two sampling points,the concept of daily electricity consumption curves was proposed.To a certain extent,the limitation of sampling interval of daily electricity consumption curves was avoided,and the influence of random load disturbance was greater,so as to improve the efficiency of clustering class.Finally,combined with the above research,the existing load modeling theory and engineering application,relying on the massive power grid big data,it realized the automatic download,processing and storage of multi-source heterogeneous load characteristic data of power big data,completed the online modeling of substation and online correction of model parameters,jointly built a complete substation modeling platform based on big data fusion,and realized the real-time update of substation comprehensive modeling parameter database and load characteristic database.According to the current load modeling theory analysis method and engineering application development results,this paper deeply studies the intelligent load modeling technology based on power big data,with rich power big data as the support,deeply excavates the effective power information,and multi-dimensional analyzes the change law of load characteristics,so as to adapt to the requirements of power grid intelligent dispatching and improve the accuracy and feasibility of load model It is of great significance to improve the reliability,stability of power grid and transmission quality of UHV AC / DC.Therefore,this paper has high theoretical research and engineering application value.