Design And Implementation Of Cloud Platform For Short-term Load Forecasting Of Regional Power Grid

Short-term power load forecasting is an important research direction in the field of power system research.Accurate and efficient forecasting has a very positive effect on the safe and stable operation of the power grid.However,there are many influencing factors of power load,and the collection and extraction of various influencing factors are usually done manually by hand,with huge workload and low efficiency.Relevant scholars have carried out a lot of work around short-term load forecasting research,made a lot of research results,and also proposed many forecasting algorithms.However,most of these methods have certain restrictions on applicable conditions.In addition,as the volume and dimensions of power data continue to increase,short-term load forecasting research faces new difficulties and challenges.This thesis has carried out research on this,and designed and implement a cloud platform for short-term load forecasting in regional power grid.This thesis firstly introduces the research background and significance of short-term power load forecasting,and describes the current status of domestic and international research on load characteristic indexes,load forecasting methods and load forecasting software.Combining literature analysis and field research,this thesis adopts B/S architecture system and LAMP(Linux,Apache,My SQL,PHP)development framework,integrates three mainstream neural network load forecasting models as load forecasting algorithms,uses python for load forecasting basic data collection,and uses grid enterprise cloud as the cloud platform of the system.Then,this thesis analyzes the functional and non-functional requirements of the system for the actual needs of grid load forecasting,and designs the system architecture and functional modules on this basis,i.e.load forecasting,load data management,load characteristics management,and personnel management functional modules.In this thesis,detailed functional design,algorithm flow design,and data model design are carried out for each module,and the system is deployed and implemented on the private cloud server of electric power.After the cloud platform was built,the system was tested for functionality and compatibility and security,and the test results showed that the platform can accurately predict short-term loads and has good security and compatibility.The cloud platform system for short-term load forecasting of regional power grid designed in this thesis realizes automated and scientific load forecasting of power grid,informationized management of load history data and load characteristic data,assessment and management of load forecasting personnel,improvement of load forecasting work efficiency,improvement of power grid automation management,and powerful support for accurate,efficient and scientific load forecasting of power grid.