Research And Design Of Multifunctional Concentrator For Smart Grid

In the traditional electricity consumption information acquisition network,the network node’s resource is not fully utilized,and it is impossible to analyze the power consumption data,resulting in difficulty in meeting the rapid progress of smart grid construction.In this thesis,designed a multi-function concentrator for smart grid.The concentrator can make full use of its own resources to realize the power consumption data acquisition,storage,analysis and processing.The specific studies are as follows:1.According to the functional requirements of the multi-functional concentrator,the concentrator’s overall solution was designed,including the hardware design and software design of the concentrator.In order to realize the power consumption data analysis and detection,the data storage capacity of the concentrator is increased by expanding the memory,and the database is used to achieve efficient access to the power consumption data.2.The detection of nontechnical power loss in the electricity network is implemented at the multi-functional concentrator.According to the energy metering principle,the nontechnical power loss can be divided into the under-voltage method,the under-current method,the phase-stealing method,and other methods of stealing.By setting a threshold,the under-voltage method can be used to steal and shift power.Detection of power stealing and undercurrent stealing;other stealing methods will cause the load to deviate significantly from the normal load,and the grey correlation analysis will be used to obtain the date load data with higher similarity in temperature as the reference data set.Through the distancebased outlier detection method,the detection of abnormal power consumption behavior is realized.3.According to the characteristics of changes in the power load,the similar-day method based on the grey correlation analysis and the least square method combined forecasting model are used to achieve short-term load forecasting.Based on the characteristics of the two predictive models,on the working day,by calculating the similarity between the forecast date and the reference date,a similar day is used for load forecasting;at rest days,the real-time segmentation of electricity consumption data is used to adopt the least-squares method.Realize the forecast of electricity load.Combined with the predicted load,the number of meter readings is increased for adjacent periods where the load change is large,in order to achieve an accurate grasp of changes in the power load.4.Through the design and test of the multi-functional concentrator,the concentrator can realize the basic acquisition and storage of electricity data;According to historical electricity data and field tests,the abnormal load detection and short-term load forecasting are verified.The results shows that the concentrator can meet the requirements of anomaly detection and short-term load forecasting of electricity data.