Spatial Load Forecasting And Error Evaluation Method For Urban Network Planning

With the continuous development and construction of the city,the population flow is accelerated,and the spatial and regional characteristics of the urban network load are increasingly prominent.In order to ensure that the urban grid planning and design can meet the needs of load capacity and spatial distribution,it is particularly critical to improve the accuracy of spatial load forecasting results,evaluating the actual impact of forecasting errors on urban network planning objectively and effectively is also important.Firstly,the related theories of urban power grid load are summarized : the classification method of urban power grid load is sorted out and its influencing factors are analyzed.The temporal and spatial characteristics of spatial power load in urban power grid are expounded,and the realization process of spatial power load forecasting and how to complete spatial load forecasting in power geographic information system are introduced.Then,aiming at the problem that the existing spatial load forecasting methods focus on mining the time sequence regularity of load and insufficiently consider the spatial variation of load,a spatial load forecasting method considering temporal-spatial correlation is proposed.based on the cell,the spatial distribution of urban network load is located.Combined with the spatial location and shared boundary of type I cell,the measurement method of cell spatial correlation is given according to the geographical regularity,and the Pearson correlation coefficient is used to measure the temporal correlation of each cell load.On this basis,a spatial load forecasting model including a data outlier detection module using grid local outlier factor algorithm and a temporal-spatial rule mining module using temporal-spatial convolution algorithm is constructed.The example analysis shows that the method proposed in this paper can improve the accuracy of urban network spatial load forecasting and is effective.Finally,the existing research on spatial load forecasting error evaluation has the problem that the spatial distribution of errors is not considered or the evaluation is not accurate due to insufficient consideration.A spatial load forecasting error evaluation method based on error amplitude-space characteristic analysis is proposed.Based on the theoretical basis of the interpretation of the spatial load forecasting error amplitude-space characteristics,the error amplitude-space characteristics are characterized by constructing the error amplitude-space offset model and drawing the error amplitude-space superposition curve cluster,and then the error evaluation index that can comprehensively reflect the influence of the dual attributes of the error amplitude size and spatial distribution on the power grid planning is designed,and the actual impact of the error on the power grid planning is used as a benchmark to test the performance of the error evaluation index.The example analysis shows that the error evaluation method proposed in this paper realizes a more comprehensive evaluation of the spatial load forecasting error from the two dimensions of amplitude and space,and is closer to the actual situation of the influence of the error on the power grid planning.