CN-121457733-B - Regional power consumption intelligent prediction method based on multi-source heterogeneous data

CN121457733BCN 121457733 BCN121457733 BCN 121457733BCN-121457733-B

Abstract

The invention relates to the technical field of electricity consumption prediction, in particular to an intelligent regional electricity consumption prediction method based on multi-source heterogeneous data, which comprises the steps of determining electricity consumption prediction influence tendency aiming at the next preset monitoring period based on an electric power change characterization value; and when the prediction abnormality of the power consumption is determined, the prediction parameters for determining the power consumption in the next preset monitoring period are corrected based on the regional industry change quantized value, the method comprises the steps of identifying a main analysis data set, determining an increase model training set or a separation prediction model based on the period quantized value of the main analysis data, and carrying out targeted analysis prediction on the power consumption according to the specific condition of the power elasticity coefficient of the region to be predicted, thereby improving the prediction efficiency of the power consumption in the region.

Inventors

WANG ZHENG
YU JIA
HU JINJING
CUI SHIXIONG
LIU SHUANGXU
LIU ZHENGYUAN
CHEN FENGTAO

Assignees

国电华研(北京)电力咨询有限公司
国网辽宁省电力有限公司朝阳供电公司

Dates

Publication Date: 20260508
Application Date: 20251113

Claims (3)

1. The intelligent regional power consumption prediction method based on the multi-source heterogeneous data is characterized by comprising the following steps of: collecting electricity consumption in each history preset monitoring period in the area to construct an electricity consumption data set; Determining a power change characterization value based on the historical power spring force coefficient; Determining a power consumption prediction influence tendency for the next preset monitoring period based on the power change characterization value; When the strong influence tendency of the power consumption prediction is determined for the next preset monitoring period, the power consumption is corrected through the elastic force influence weight coefficient, and whether the power consumption prediction is qualified or not is determined based on the energy change characterization value; when the prediction abnormality for the electricity consumption is determined, correcting the prediction parameters for determining the electricity consumption in the next preset monitoring period based on the regional industry change quantized value, wherein the prediction parameters comprise identifying a main analysis data set, and determining an addition model training set or a separation prediction model based on the period quantized value of the main analysis data; When the next preset monitoring period is determined to be the weak influence trend of the electricity consumption prediction, the electricity consumption in the next preset monitoring period is determined based on the electricity consumption data set; A process for determining a power change characterization value based on a historical power spring rate, comprising: drawing an electric power elastic coefficient time domain curve based on each electric power elastic coefficient in the historical data; Solving the absolute value of the ratio of the absolute value of the slope of the power elastic coefficient time domain curve at the current time node to the preset stable slope to obtain a power change characterization value; the process for correcting the power consumption by the elastic force influence weight coefficient comprises the following steps: Determining an elastic coefficient selection interval based on the current electric power elastic coefficient; determining each preset monitoring period of the electric power elastic coefficient in the elastic coefficient selection interval in the historical data as a reference period; obtaining the influence multiplying power of each reference period; determining an elasticity influence weight coefficient based on each influence multiplying power; a process for determining an energy source change characterization value, comprising: Acquiring the ratio of the natural gas consumption in the latest preset monitoring period to the average natural gas consumption in each preset monitoring period in the historical data to obtain the natural gas change rate; Acquiring the ratio of the coal consumption in the latest preset monitoring period to the average coal consumption in each preset monitoring period in the historical data to obtain the coal change rate; Acquiring the ratio of the petroleum consumption in the latest preset monitoring period to the average petroleum consumption in each preset monitoring period in the historical data to obtain the petroleum change rate; acquiring the ratio of the power consumption in the latest preset monitoring period to the average power consumption in each preset monitoring period in the historical data to obtain the power consumption change rate; calculating the average value of the natural gas change rate, the coal change rate and the petroleum change rate to obtain the change quantity of the alternative energy sources; solving the ratio of the absolute value of the difference value between the variable quantity of the alternative energy and the electricity utilization change rate to obtain an energy change characterization value; a process for determining a regional industry change quantization value, comprising: acquiring industrial electricity consumption and service electricity consumption in each history preset monitoring period in the area to acquire an industrial electricity consumption data set and a service electricity consumption data set respectively; acquiring the ratio of the industrial power consumption of the latest preset monitoring period to the average industrial power consumption of each historical preset monitoring period to obtain a regional industrial change quantized value; acquiring the ratio of the service electricity consumption of the latest preset monitoring period to the average service electricity consumption of each historical preset monitoring period to obtain a regional service industry change quantized value; A process for correcting a prediction parameter for determining an amount of electricity used in a next preset monitoring period based on a regional industry change quantized value, comprising: When the regional industry change quantized value is smaller than or equal to the regional service industry change quantized value, determining the service electricity data set as a main analysis data set; When the regional industry change quantized value is greater than the regional service industry change quantized value, determining the industrial electricity data set as a main analysis data set; A process for determining an augmented model training set or a separate predictive model based on periodic quantized values of primary analytical data, comprising: drawing a main power consumption time domain curve based on the power consumption of each preset monitoring period in the acquired main analysis data set; converting signals of a time domain curve of main electricity consumption into signals of a frequency domain through Fourier transformation to obtain a spectrogram; Acquiring each peak value in the spectrogram; when the existing peak value is larger than the preset peak value, determining to increase the model training set; when each peak value is smaller than or equal to a preset peak value, separating the prediction model; A process for adding a training set of models, comprising: the amount of data of the training set used to train the first predictive model is adjusted to a corresponding value based on the maximum peak, wherein, Obtaining each peak value in the spectrogram to obtain a maximum peak value; the increasing amplitude of the data volume is inversely related to the maximum peak value; A process for separating predictive models, comprising: determining the predicted electricity consumption output by the model trained by the industrial electricity data set as the industrial predicted electricity consumption; determining predicted electricity consumption output by a model trained by the service electricity consumption data set as service predicted electricity consumption; And determining the sum of the industrial predicted power consumption and the service predicted power consumption as the predicted power consumption of the next preset monitoring period.
2. The intelligent prediction method for regional power usage based on multi-source heterogeneous data according to claim 1, wherein the process of determining the predicted influence tendency of the power usage for the next preset monitoring period based on the power change characterization value comprises: when the power change characterization value is smaller than or equal to the preset power change characterization value, determining the next preset monitoring period as a power consumption prediction weak influence tendency, and determining the power consumption in the next preset monitoring period based on a power consumption data set; When the power change characterization value is larger than the preset power change characterization value, and the next preset monitoring period is determined to be the power consumption prediction strong influence tendency, the power consumption is determined to be corrected through the elastic force influence weight coefficient.
3. The method for intelligently predicting regional power usage based on multi-source heterogeneous data according to claim 2, wherein determining whether the prediction for power usage is acceptable based on the energy change characterization value comprises: when the energy change characterization value is smaller than or equal to the preset energy change characterization value, determining that the prediction for the power consumption is qualified, and determining the product of the elasticity influence weight coefficient and the power consumption predicted by the first prediction model as the predicted power consumption of the next preset monitoring period; when the energy change characterization value is larger than the preset energy change characterization value, determining prediction abnormality for the power consumption, and correcting a prediction parameter for determining the power consumption in the next preset monitoring period based on the regional industry change quantization value.

Description

Regional power consumption intelligent prediction method based on multi-source heterogeneous data Technical Field The invention relates to the technical field of electricity consumption prediction, in particular to an intelligent regional electricity consumption prediction method based on multi-source heterogeneous data. Background In the current society, electric power is a vital energy source, and stable supply and reasonable distribution play a key role in economic development, social stability and guarantee of life quality of people. The accurate prediction of regional power consumption is not only beneficial to the reasonable arrangement of power generation plans and the optimization of power grid dispatching of power enterprises, reduces the power generation cost and the energy waste, but also provides scientific basis for government to make energy policies and planning energy infrastructure construction. At present, the regional power consumption prediction methods are numerous, but with the rapid development of economy, continuous adjustment of energy structures and increasingly diversified power consumption behaviors of users, the traditional prediction methods often have difficulty in fully considering the influence of multi-source heterogeneous data, so that the accuracy and reliability of prediction results are limited to a certain extent. The multi-source heterogeneous data has wide sources and various formats, and contains rich information related to the power consumption. In cities, with the improvement of living standards of residents and the continuous perfection of urban functions, the power consumption of residents and the power consumption demand of businesses show different change trends, and in industrial parks, the upgrading and adjustment of industrial structures can directly influence the change of industrial power consumption. The economic activities are complex and various, the energy consumption structure is continuously changed, the industrial development is also in the process of dynamic adjustment, and great challenges are caused to the electricity consumption prediction. The Chinese patent publication No. CN116108979A discloses a cross-period multi-source heterogeneous power data processing system, which comprises a data acquisition module, a data analysis module, a data modeling module, a multi-source heterogeneous information module, a prediction module, a switching module and a data pushing module, wherein the data acquisition module, the data analysis module, the data modeling module, the multi-source heterogeneous information module, the prediction module, the switching module and the data pushing module are application software installed in a PC, and the application method of the cross-period multi-source heterogeneous power data processing system comprises seven steps. The power utilization load data model of the power utilization area can be obtained, the power utilization prediction can be carried out according to the obtained model, and the power utilization load data model can be adjusted in advance before the power utilization power of the area changes. Disclosure of Invention Therefore, the invention provides an intelligent prediction method for regional power consumption based on multi-source heterogeneous data, which is used for solving the problem that the prediction accuracy of the regional power consumption is affected by carrying out targeted analysis prediction on the power consumption according to the specific situation of the power elasticity coefficient of the region to be predicted in the prior art. In order to achieve the above object, the present invention provides an intelligent prediction method for regional power consumption based on multi-source heterogeneous data, comprising: collecting electricity consumption in each history preset monitoring period in the area to construct an electricity consumption data set; Determining a power change characterization value based on the historical power spring force coefficient; Determining a power consumption prediction influence tendency for the next preset monitoring period based on the power change characterization value; When the strong influence tendency of the power consumption prediction is determined for the next preset monitoring period, the power consumption is corrected through the elastic force influence weight coefficient, and whether the power consumption prediction is qualified or not is determined based on the energy change characterization value; when the prediction abnormality for the electricity consumption is determined, correcting the prediction parameters for determining the electricity consumption in the next preset monitoring period based on the regional industry change quantized value, wherein the prediction parameters comprise identifying a main analysis data set, and determining an addition model training set or a separation prediction model based on the period quantized value of t