Industrial chain electricity consumption early warning method and system based on time scale analysis

By constructing a multi-dimensional space-time cube and space-time coupled prediction model, combined with clustering algorithm, the problem that power consumption early warning system in the existing technology is difficult to capture complex features, and more accurate power consumption forecasting and dynamic early warning are achieved, and early warning is improved. effectiveness and timeliness.

CN119089238APending Publication Date: 2024-12-06YUNNAN POWER GRID CO LTD
View PDF 0 Cites 2 Cited by

Patent Information

Application Number
CN202410982531.9
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-07-22
Publication Date
2024-12-06

AI Technical Summary

Technical Problem

The existing power consumption warning system of the industrial chain focuses on a single time series analysis, ignores the comprehensive influence of spatial dimensions and environmental factors, and is difficult to capture nonlinear, seasonal and complex periodic characteristics. The warning mechanism lacks adaptability, which reduces the warning. effectiveness and timeliness.

Method used

By collecting the electricity consumption, geographical location and climatic conditions of industrial chain enterprises, we build a multi-dimensional space-time cube, extract long-term trend, seasonal and periodic characteristics, analyze the spatial aggregation of electricity consumption distribution in combination with clustering algorithms, and build a space-time coupled prediction. The model and determine the warning level based on the deviation between the actual value of the electricity consumption and the predicted value.

Benefits of technology

It realizes more accurate electricity consumption prediction and dynamic early warning, improves the flexibility and accuracy of early warning, optimizes the energy consumption management and energy conservation and emission reduction in the industrial chain, and provides a scientific basis.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119089238A_ABST
    Figure CN119089238A_ABST
Patent Text Reader

Abstract

The invention discloses an industrial chain electricity consumption early warning method and system based on time scale analysis, and relates to the technical field of industrial chain electricity consumption early warning, and the method comprises the steps: collecting the electricity consumption of an industrial chain enterprise, the geographic position of the enterprise, the climate condition and holidays and festivals, and storing the preprocessed data in a multi-source database; constructing a multi-dimensional space-time cube to obtain space-time cube data; analyzing the electricity consumption of each space unit on different time scales based on the space-time cube data; analyzing the space aggregation of the electricity consumption distribution through a clustering algorithm; combining a clustering result to construct a space-time coupling prediction model, and outputting a predicted estimated value of the electricity consumption; and according to a deviation value between the actual value of the electricity consumption and the predicted estimated value of the electricity consumption, judging an early warning level. According to the method, powerful technical support is provided for promoting intelligence and refinement of energy management in terms of high efficiency of data processing, depth of spatial-temporal feature analysis, accuracy of a prediction model and flexibility of an early warning mechanism.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the technical field of industrial chain electricity consumption early warning, and in particular to an industrial chain electricity consumption early warning method and system based on time scale analysis. Background Art

[0002] In recent years, with the acceleration of global industrialization and informatization, energy consumption management and optimization of the industrial chain have become key areas for improving energy efficiency. Traditional electricity consumption forecasts mostly focus on a single enterprise or region, lack consideration of the overall perspective of the industrial chain, and are difficult to accurately capture electricity consumption patterns under complex interactive relationships. With the rapid development of big data, cloud computing, and artificial intelligence technologies, spatiotemporal data analysis has gradually become a new way to explore energy consumption patterns, providing unprecedented possibilities for energy consumption early warnings for the industrial chain.

[0003] Existing technologies generally face several challenges in the electricity consumption forecasting and early warning systems for the industrial chain. First, most systems focus on single time series analysis, ignoring the combined impact of spatial dimensions and environmental factors, such as geographic location, climate change, and holiday effects, resulting in limited generalization capabilities of the forecasting model. Secondly, in terms of data processing, existing technologies often lack efficient data integration and storage solutions, especially the processing capabilities for multi-source heterogeneous data are insufficient, affecting the comprehensiveness and real-time nature of the data. Furthermore, the forecasting models are mostly concentrated on traditional statistical methods, such as linear regression or simple time series models, which are not accurate enough in capturing nonlinear, seasonal, and complex periodic characteristics, and are difficult to meet the dynamically changing characteristics of industrial electricity consumption. Finally, the early warning mechanism is usually based on a fixed threshold, does not fully consider the dynamic changes in the actual and predicted deviations in electricity consumption, lacks adaptability, and reduces the effectiveness and timeliness of the early warning. Summary of the invention

[0004] In view of the above existing problems, the present invention is proposed.

[0005] Therefore, the present invention provides an industrial chain electricity consumption warning method and system based on time scale analysis to solve the problem that most systems focus on single time series analysis, ignore the combined influence of spatial dimension and environmental factors, are not accurate enough in capturing nonlinear, seasonal and complex periodic characteristics, are difficult to meet the dynamically changing industrial electricity consumption characteristics, and do not fully consider the dynamic changes of actual electricity consumption and predicted deviations.

[0006] In order to solve the above technical problems, the present invention provides the following technical solutions:

[0007] In the first aspect, an embodiment of the present invention provides an industrial chain electricity consumption warning method based on time scale analysis, which includes collecting electricity consumption, enterprise geographical location, climate conditions and holidays of industrial chain enterprises, and storing the pre-processed data in a multi-source database; constructing a multidimensional space-time cube with time as the axis and geographical location, climate conditions and holidays as dimensions, converting the original data into an ordered storage structure in the space-time cube to obtain space-time cube data; based on the space-time cube data, stratifying according to the time scale to extract long-term trend characteristics, seasonal characteristics and periodic characteristics, and analyzing the electricity consumption of each spatial unit at different time scales; based on the space-time cube data, analyzing the spatial clustering of electricity consumption distribution through a clustering algorithm; combining the clustering results, constructing a space-time coupling prediction model, and outputting a predicted estimated value of electricity consumption; judging the warning level according to the deviation value between the actual value of electricity consumption and the predicted estimated value of electricity consumption.

[0008] As a preferred solution of the industrial chain electricity consumption early warning method based on time scale analysis described in the present invention, the preprocessing includes the following steps: collecting the electricity consumption of enterprises in the industrial chain, integrating the time axis of holidays and electricity consumption to form a time series; capturing the climate conditions of the target area, integrating the geographical location of the enterprises using GIS technology, and performing standardized processing to match the climate conditions; setting a data receiving port in a multi-source database, performing format checking and standardized conversion on the received raw data, ensuring that all data follow a unified data model, and implementing data cleaning processing.

[0009] As a preferred solution of the industrial chain electricity consumption early warning method based on time scale analysis described in the present invention, the construction of a multidimensional space-time cube includes the following steps: defining the time dimension, taking hours as the basic unit, and constructing multiple time levels of days, weeks, months, quarters, and years upward, converting the timestamp in the original data into the selected time level, and mapping each record to the corresponding hour, day, and week time unit; refining the spatial dimension, using the hexagonal grid method to subdivide the geographic space, and the bottom layer is refined to blocks or more microscopic grid units. Each unit corresponds to the longitude and latitude range under the GIS coordinate system. According to the geographical location of the enterprise, each observation point is mapped to a pre-defined spatial grid to determine its position in the cube; setting additional dimensions, assigning codes to each additional dimension, and designing multi-level classifications for each dimension; using the key as a combination of time dimension-space dimension-additional dimension, and the value as a storage structure of electricity consumption at the corresponding position, the original data is stored in the multidimensional space-time cube to obtain space-time cube data.

[0010] As a preferred solution of the industrial chain electricity consumption warning method based on time scale analysis described in the present invention, the stratification according to the time scale to extract long-term trend characteristics, seasonal characteristics and periodic characteristics includes the following steps: dividing the time dimension of the space-time cube data into long-term, medium-term and short-term, corresponding to long-term trends, seasonal fluctuations and short-term random changes; using the sliding window method to slice the data, the long-term trend is defined as interannual changes, and the time window is set to one year; medium-term fluctuations correspond to seasonal cycles, and the window is set to one quarter; short-term changes focus on daily fluctuations and special events, and the window is set to one week; resampling the original data, converting it into data sequences at different time scales, and aggregating the original daily data by year, season and week respectively to obtain the average electricity consumption of the corresponding scale.

[0011] As a preferred solution of the industrial chain electricity consumption early warning method based on time scale analysis described in the present invention, the spatial clustering of electricity consumption distribution analyzed by clustering algorithm includes the following steps: clustering the geographical locations of enterprises by K-Means algorithm to identify areas with similar electricity consumption behaviors; analyzing the electricity consumption correlation between neighboring enterprises by calculating geographic space distance and using spatial weight matrix; identifying electricity intensive areas by using GIS hot spot analysis; analyzing step by step from micro to macro, observing electricity consumption patterns at different spatial scales by using multi-layer aggregation method; and analyzing the spatial clustering of electricity consumption distribution.

[0012] As a preferred solution of the industrial chain electricity consumption early warning method based on time scale analysis described in the present invention, wherein: the construction of the spatiotemporal coupling prediction model includes the following steps: combining the clustering results, extracting spatiotemporal features, the spatiotemporal features include time series data, geographic location code, seasonal label, holiday indicator and climate feature vector; defining the coding of spatiotemporal features, t is the time series data, indicating the time point; g is the geographic location code, using longitude and latitude pairs to represent geographic coordinates; s is the seasonal label, encoding the season, spring 1, summer 2, autumn 3, winter 4; h is the holiday indicator, assigned according to whether it is a holiday, 0 is a working day, 1 is a holiday; c is a climate feature vector; the fusion of spatiotemporal features is expressed as,

[0013]

[0014] Among them, F is the fusion spatiotemporal feature, λ t is the weight of the time feature, is the normalized time feature of the time series data t, λ g is the weight of the geographic location feature, is the normalized geographic location feature vector of the geographic location code g, λ s is the weight of the seasonal feature, is the normalized seasonal feature of the seasonal label s, λ h is the weight of the holiday feature, is the normalized holiday feature of the holiday indicator h, λ c is the weight of the climate feature, is the normalized value of the climate feature vector c; taking the fused spatiotemporal feature F as the input value, the output power consumption forecast estimate is expressed as,

[0015]

[0016] in, is the predicted estimated value of electricity consumption; Φ(F) represents the prediction function, which is the output prediction model after comprehensively considering the spatiotemporal characteristics and complex information of the attention mechanism, and is used to map from input data to predicted values; STM(F(t)) is the spatiotemporal memory function, which processes the dynamic changes of time series features at different time points; Attention(F,t) is the spatiotemporal attention mechanism function, which is used to dynamically adjust the importance of different features; T is the prediction period.

[0017] As a preferred solution of the industrial chain power consumption early warning method based on time scale analysis described in the present invention, the deviation value between the actual power consumption and the predicted estimated power consumption is expressed as:

[0018]

[0019] Among them, Y i is the actual value of the power consumption of the i-th sample; is the predicted estimated value of the electricity consumption of the ith sample; when σ<5%σ, there is a slight deviation, indicating a first-level warning, and the electricity consumption of the industrial chain fluctuates normally; when 5%σ<σ<15%σ, there is a moderate deviation, indicating a second-level warning, and there is a potential impact on the electricity consumption of the industrial chain; when σ>15%σ, there is a severe deviation, indicating a third-level warning, and the electricity consumption of the industrial chain has behaviors that seriously affect the power supply.

[0020] In the second aspect, the present invention provides an industrial chain electricity consumption early warning system based on time scale analysis, including a data collection module, a multidimensional space-time cube construction module, a time series feature extraction module, a spatial feature extraction module, a space-time coupling prediction model construction module and an electricity consumption early warning module, wherein the data collection module is used to collect the electricity consumption, geographical location, climate conditions and holidays of enterprises in the industrial chain, and store the preprocessed data in a multi-source database; the multidimensional space-time cube construction module is used to construct a multidimensional space-time cube with time as the axis and geographical location, climate conditions and holidays as the dimensions, and convert the original data into an ordered storage structure in the space-time cube. The time series feature extraction module is used to extract long-term trend features, seasonal features and periodic features based on the space-time cube data and stratify them according to the time scale, and analyze the power consumption of each spatial unit at different time scales; the spatial feature extraction module is used to analyze the spatial clustering of power consumption distribution through a clustering algorithm based on the space-time cube data; the space-time coupling prediction model construction module is used to construct a space-time coupling prediction model in combination with the clustering results, and output a predicted estimated value of power consumption; the power consumption warning module is used to determine the warning level according to the deviation value between the actual value of power consumption and the predicted estimated value of power consumption.

[0021] In a third aspect, an embodiment of the present invention provides a computer device, including a memory and a processor, wherein the memory stores a computer program, wherein: when the computer program is executed by the processor, it implements any step of the industrial chain electricity consumption warning method based on time scale analysis as described in the first aspect of the present invention.

[0022] In a fourth aspect, an embodiment of the present invention provides a computer-readable storage medium having a computer program stored thereon, wherein: when the computer program is executed by a processor, it implements any step of the industrial chain electricity consumption warning method based on time scale analysis as described in the first aspect of the present invention.

[0023] The beneficial effects of the present invention are as follows: the present invention constructs a multidimensional space-time cube structure by integrating multi-source data such as electricity consumption, geographical location, climate conditions and holidays of enterprises in the industrial chain, and innovatively performs deep feature extraction and analysis at dual scales of time and space. Through the hierarchical processing of the space-time cube, long-term trends, seasonal fluctuations and periodic characteristics are systematically revealed, especially the personalized electricity consumption patterns of spatial units are finely characterized. Combined with the clustering algorithm, the spatial distribution law of electricity consumption is further explored to achieve a more accurate description of spatial characteristics. On this basis, a space-time coupling prediction model is constructed to integrate space-time characteristics, output electricity consumption prediction values, and implement a hierarchical early warning strategy by dynamically analyzing the deviation between the actual and predicted values, thereby improving the flexibility and accuracy of the early warning. The present invention not only optimizes the limitations of the existing technology, but also provides a scientific basis for energy consumption management and energy conservation and emission reduction in the industrial chain, showing significant practical value and theoretical innovation. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without paying creative work.

[0025] Figure 1 This is a flow chart of the industrial chain electricity consumption early warning method based on time scale analysis in Example 1.

[0026] Figure 2 This is a flow chart of the industrial chain electricity consumption early warning system based on time scale analysis in Example 1. DETAILED DESCRIPTION

[0027] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the specific implementation methods of the present invention are described in detail below in conjunction with the accompanying drawings.

[0028] In the following description, many specific details are set forth to facilitate a full understanding of the present invention, but the present invention may also be implemented in other ways different from those described herein, and those skilled in the art may make similar generalizations without violating the connotation of the present invention. Therefore, the present invention is not limited to the specific embodiments disclosed below.

[0029] Secondly, the term "one embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The term "in one embodiment" that appears in different places in this specification does not necessarily refer to the same embodiment, nor does it refer to a separate or selective embodiment that is mutually exclusive with other embodiments.

[0030] Example 1

[0031] Reference Figure 1 and Figure 2 , which is the first embodiment of the present invention, and provides an industrial chain electricity consumption early warning method based on time scale analysis, comprising the following steps:

[0032] S1. Collect the electricity consumption, geographical location, climate conditions and holidays of enterprises in the industrial chain, and store the pre-processed data in a multi-source database.

[0033] Furthermore, the preprocessing includes the following steps:

[0034] Collect the electricity consumption of enterprises in the industrial chain, integrate the time axis of holidays and electricity consumption, and form a time series;

[0035] Capture the climate conditions of the target area, integrate the enterprise's geographical location using GIS technology, and standardize it to match the climate conditions;

[0036] Set up data receiving ports in multi-source databases, perform format checks and standardization conversions on the received raw data, ensure that all data follows a unified data model, and implement data cleaning processing.

[0037] It should be noted that by collecting the electricity consumption, geographical location, climate conditions and holiday information of enterprises in the industrial chain and pre-processing the data, the comprehensive integration and standardization of the data are achieved, ensuring the accuracy and availability of the information. The alignment of time series and holidays enhances the temporal continuity of the data, while the application of GIS technology improves the accuracy of geographical location information, laying a solid foundation for subsequent spatiotemporal analysis. This step ensures the quality of the data and improves the reliability and accuracy of the early warning system.

[0038] S2. With time as the axis and geographic location, climate conditions, and holidays as dimensions, a multidimensional space-time cube is constructed to convert the original data into an ordered storage structure in the space-time cube to obtain space-time cube data.

[0039] Furthermore, constructing a multidimensional space-time cube includes the following steps:

[0040] Define the time dimension, using the hour as the basic unit, and construct multiple time levels of day, week, month, quarter, and year to meet the needs of different forecasting accuracies; convert the timestamp in the original data to the selected time level, and map each record to the corresponding hour, day, and week time unit;

[0041] Refine the spatial dimension and use the hexagonal grid method to subdivide the geographic space. The lowest level is refined to blocks or more microscopic grid units. Each unit corresponds to the longitude and latitude range under the GIS coordinate system. Districts, counties, cities, provinces and even countries are set up as the upper spatial levels to achieve multi-level spatial organization. Based on the geographical location of the enterprise, each observation point is mapped to a pre-defined spatial grid to determine its position in the cube.

[0042] Set additional dimensions, assign codes to each additional dimension, and design multi-level classification for each dimension; the climate condition dimension includes temperature, humidity, and wind speed, and the temperature is divided into low temperature, suitable temperature, and high temperature; the holiday dimension records specific dates;

[0043] The key is a combination of time dimension, space dimension and additional dimension, and the value is a storage structure of the power consumption at the corresponding location. The original data is stored in a multidimensional space-time cube to obtain space-time cube data.

[0044] It should be noted that defining and constructing a multidimensional space-time cube and storing data in an ordered structure not only greatly optimizes the data organization method, but also facilitates efficient query and analysis. By refining the time and space dimensions and introducing additional dimension classification, this step can carefully depict the electricity consumption pattern of each spatial unit at different time scales, providing an intuitive and visual platform for in-depth exploration of space-time correlations and patterns, and greatly enhancing the depth of space-time understanding of electricity consumption behavior.

[0045] S3. Based on the space-time cube data, stratification is performed according to the time scale to extract long-term trend characteristics, seasonal characteristics, and periodic characteristics, and the electricity consumption of each spatial unit at different time scales is analyzed.

[0046] Furthermore, stratifying by time scale to extract long-term trend features, seasonal features, and cyclical features includes the following steps:

[0047] The time dimension of the space-time cube data is divided into long-term, medium-term and short-term, corresponding to long-term trends, seasonal fluctuations and short-term random changes;

[0048] The sliding window method is used to slice the data. The long-term trend is defined as the interannual change, and the time window is set to one year; the medium-term fluctuation corresponds to the seasonal cycle, and the window is set to one quarter; the short-term change focuses on daily fluctuations and special events, and the window is set to one week;

[0049] The original data is resampled and converted into data series at different time scales. The original daily data is aggregated by year, season, and week to obtain the average electricity consumption or total electricity consumption at the corresponding scale.

[0050] It should be noted that by stratifying the space-time cube data by time scale and extracting long-term trends, seasonal characteristics and cyclical characteristics, this process effectively distinguishes the various driving factors of electricity consumption changes. The use of the sliding window method for data slicing and resampling makes the analysis more refined, and can accurately grasp the complex changes in electricity consumption over time, providing key input information for the prediction model and improving the model's sensitivity to abnormal changes and prediction accuracy.

[0051] S4. Based on the space-time cube data, the spatial clustering of electricity consumption distribution is analyzed through clustering algorithm.

[0052] Furthermore, the spatial clustering of electricity consumption distribution is analyzed by clustering algorithm, which includes the following steps:

[0053] Use the K-Means algorithm to cluster the geographical locations of enterprises and identify areas with similar electricity consumption behaviors;

[0054] By calculating the geographic distance, the spatial weight matrix is ​​used to analyze the correlation of electricity consumption between neighboring enterprises.

[0055] Use GIS hot spot analysis to identify areas with high electricity consumption;

[0056] Analyze from micro to macro level, that is, from block to country level, using multi-layer aggregation method to observe electricity consumption patterns at different spatial scales;

[0057] Analyze the spatial clustering of electricity consumption distribution.

[0058] It should be noted that the application of K-Means algorithm to cluster geographical locations, combined with spatial weight matrix and GIS hot spot analysis, reveals the spatial clustering and correlation of electricity consumption distribution. This not only helps to identify areas with similar electricity consumption behaviors, but also helps to understand the differences in electricity consumption patterns between regions, provides a scientific basis for the formulation of regional management strategies, and enhances the regional pertinence and practicality of the early warning system.

[0059] S5. Combine the clustering results to build a spatiotemporal coupling prediction model and output the predicted estimated value of electricity consumption.

[0060] Furthermore, constructing a spatiotemporal coupling prediction model includes the following steps:

[0061] Combine the clustering results to extract spatiotemporal features, which include time series data, geographic location codes, seasonal labels, holiday indicators, and climate feature vectors;

[0062] Define the coding of spatiotemporal features. t is time series data, indicating a time point; g is the geographic location code, which uses a longitude and latitude pair to indicate geographic coordinates; s is the seasonal label, which encodes the season, indicating spring 1, summer 2, autumn 3, and winter 4; h is the holiday indicator, which is assigned a value based on whether it is a holiday, 0 for a working day and 1 for a holiday; c is the climate feature vector;

[0063] The fused spatiotemporal features are expressed as,

[0064]

[0065] Among them, F is the fusion spatiotemporal feature, λ t is the weight of the time feature, reflecting the relative importance of time series data in prediction. is the normalized time feature of the time series data t, λ g is the weight of the geographic location feature, which is used to measure the importance of geographic location information for predicting the target. is the normalized geographic location feature vector of the geographic location code g, λ s is the weight of the seasonal feature, reflecting the relative importance of season in the prediction, is the normalized seasonal feature of the seasonal label s, λ h is the weight of the holiday feature, is the normalized holiday feature of the holiday indicator h, λ c is the weight of climate characteristics, which directly affects electricity demand, is the normalized value of the climate characteristic vector c;

[0066] Taking the fused spatiotemporal feature F as the input value, the predicted estimated value of the output power consumption is expressed as:

[0067]

[0068] in, is the predicted estimated value of electricity consumption, that is, the predicted result of electricity consumption at a certain moment or within a period of time in the future calculated by the model based on the input features; Φ(F) represents the prediction function, which is the output prediction model after comprehensively considering the spatiotemporal characteristics and complex information of the attention mechanism, and is used to map from input data to predicted values; STM(F(t)) is the spatiotemporal memory function, which processes time series features and uses LSTM to capture long-term dependencies and patterns in the sequence, indicating that this part focuses on the dynamic changes of time series data at different time points; Attention(F,t) is the spatiotemporal attention mechanism function, which is used to dynamically adjust the importance of different features, and gives different weights to different features according to the fused spatiotemporal features F and the current time point t, so that the model focuses more on key information and ignores noise; T is the prediction period, one day, one week, one month or any prediction window.

[0069] It should also be noted that the forecasting model constructed by fusion of spatiotemporal features and advanced machine learning technology (spatial-temporal memory function and spatiotemporal attention mechanism) can comprehensively consider the impact of time series, geographical location, seasonality, holidays and climate factors on electricity consumption. This comprehensive analysis method improves the complex adaptability of the forecasting model, enabling it to more accurately simulate and predict future electricity consumption trends, and provides strong decision-making support for the electricity consumption management of the industrial chain.

[0070] S6. Determine the warning level based on the deviation between the actual power consumption and the predicted estimated power consumption.

[0071] Furthermore, the deviation between the actual value of electricity consumption and the predicted value of electricity consumption is expressed as,

[0072]

[0073] Among them, Y i is the actual value of the power consumption of the i-th sample; is the predicted estimated value of the electricity consumption of the i-th sample;

[0074] When σ<5%σ, there is a slight deviation, indicating a first-level warning, and the power consumption of the industrial chain fluctuates normally;

[0075] When 5%σ<σ<15%σ, a moderate deviation occurs, indicating a secondary warning, and there is a potential impact on the electricity consumption of the industrial chain; analyze the reasons and prepare to adjust the prediction model or take preventive measures;

[0076] When σ>15%σ, a severe deviation occurs, indicating a third-level warning, and the electricity consumption of the industrial chain has serious effects on the power supply. Immediate response may affect the power balance or supply, initiate emergency plans, check whether the model has major omissions, and quickly adjust strategies.

[0077] It should also be noted that by comparing the deviation between the actual power consumption and the predicted value, intelligent judgment of the warning level is achieved, which not only simplifies the warning decision-making process, but also makes the response measures more timely and accurate. This dynamically adjusted warning mechanism can quickly identify the imbalance between power supply and demand, thereby effectively preventing power shortages, ensuring the stable operation of the industrial chain, and ultimately achieving the beneficial effect of optimizing resource allocation and improving energy utilization efficiency.

[0078] The present embodiment also provides an industrial chain power consumption warning system based on time scale analysis, including a data collection module, a multidimensional space-time cube construction module, a time series feature extraction module, a spatial feature extraction module, a space-time coupling prediction model construction module and a power consumption warning module, wherein the data collection module is used to collect the power consumption, enterprise geographical location, climate conditions and holidays of industrial chain enterprises, and store the pre-processed data in a multi-source database; the multidimensional space-time cube construction module is used to construct a multidimensional space-time cube with time as the axis and geographical location, climate conditions and holidays as the dimensions, and convert the original data into an ordered storage structure in the space-time cube to obtain space-time cube data; the time series feature extraction module is used to extract long-term trend characteristics, seasonal characteristics and periodic characteristics based on the space-time cube data and stratify them according to the time scale, and analyze the power consumption of each spatial unit at different time scales; the spatial feature extraction module is used to analyze the spatial aggregation of power consumption distribution through a clustering algorithm based on the space-time cube data; the space-time coupling prediction model construction module is used to construct a space-time coupling prediction model in combination with the clustering results, and output a predicted estimated value of power consumption; the power consumption warning module is used to judge the warning level according to the deviation value between the actual value of power consumption and the predicted estimated value of power consumption.

[0079] This embodiment also provides a computer device, which is suitable for the case of an industrial chain electricity consumption warning method based on time scale analysis, including: a memory and a processor; the memory is used to store computer executable instructions, and the processor is used to execute computer executable instructions to implement the file encryption method proposed in the above embodiment.

[0080] The computer device may be a terminal, and the computer device includes a processor, a memory, a communication interface, a display screen and an input device connected via a system bus. The processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The communication interface of the computer device is used to communicate with an external terminal in a wired or wireless manner, and the wireless manner can be achieved through WIFI, an operator network, NFC (near field communication) or other technologies. The display screen of the computer device may be a liquid crystal display screen or an electronic ink display screen, and the input device of the computer device may be a touch layer covering the display screen, or a key, trackball or touchpad provided on the housing of the computer device, or an external keyboard, touchpad or mouse, etc.

[0081] This embodiment also provides a storage medium on which a computer program is stored. When the program is executed by a processor, the method for early warning of power consumption in an industrial chain based on time scale analysis proposed in the above embodiment is implemented; the storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (Static Random Access Memory, referred to as SRAM), electrically erasable programmable read-only memory (Electrically Erasable Programmable Read-Only Memory, referred to as EEPROM), erasable programmable read-only memory (Erasable Programmable Read Only Memory, referred to as EPROM), programmable read-only memory (Programmable Red-Only Memory, referred to as PROM), read-only memory (Read-Only Memory, referred to as ROM), magnetic storage, flash memory, disk or optical disk.

[0082] In summary, the present invention gradually realizes the whole process from data collection and integration to complex spatiotemporal structured processing, deep feature extraction, and advanced prediction and early warning mechanism. Each step reflects the in-depth understanding and application of data processing, and finally forms a highly integrated and forward-looking power consumption early warning system, which is significantly better than traditional methods and improves the intelligent level of power management and energy utilization efficiency of the industrial chain.

[0083] Example 2

[0084] Referring to Table 1, which is the second embodiment of the present invention, in order to further verify the advancement of the present invention, experimental simulation data of an industrial chain electricity consumption early warning method based on time scale analysis is provided.

[0085] This example selects 100 representative industrial chain enterprises in a certain region as research objects, covering multiple industries such as manufacturing and service industries, and a two-year data set. The implementation process includes data collection, preprocessing, space-time cube construction, feature extraction, cluster analysis, model construction and early warning judgment. The specific steps are as follows:

[0086] First, we collected daily electricity consumption data of these enterprises over two years, and simultaneously obtained climate conditions (temperature, humidity) and holiday information on the corresponding dates through public APIs. The data was aligned to form electricity consumption data containing time series, and the geographical location information of the enterprises was standardized through GIS technology to ensure accurate matching with climate data.

[0087] Furthermore, the time dimension is subdivided into five levels from hours to years, the space dimension is subdivided to the block level using the hexagonal grid method, and additional dimensions include holidays and climate types. Through encoding and mapping, the original data is converted into a space-time cube structure, and each record has a clear time-space-additional attribute identifier.

[0088] Using the sliding window method, long-term trends, seasonal fluctuations, and short-term change features were extracted from the space-time cube. For example, it was found that electricity consumption in winter has a significant increase trend, and electricity consumption has a small increase in the week before the holiday. Periodic changes.

[0089] The K-Means algorithm was used to cluster enterprises into five areas with similar electricity consumption behaviors according to their geographical locations. The spatial weight matrix revealed the positive correlation of electricity consumption between adjacent areas, and GIS hot spot analysis confirmed that industrial areas were areas with high electricity consumption.

[0090] Among them, time series data, geographic location codes, season labels, holiday indicators and climate characteristics are considered when building the model. The spatiotemporal memory function and attention mechanism are used to process the features. The model predicts the electricity consumption in the next week. The predicted value is expressed as

[0091] By calculating the actual power consumption Y and the predicted value The threshold is set to ±5% for slight deviation, ±10% for moderate deviation, and ±20% for severe deviation. The details are shown in Table 1:

[0092] Table 1 Experimental data table

[0093]

[0094]

[0095] Through the comparative analysis of the above data, it can be seen that the beneficial effects of the present invention are remarkable;

[0096] Enterprises E001 and E002 showed slight deviations and activated the first-level warning, which showed that the model made accurate predictions within the normal fluctuation range and could effectively distinguish normal fluctuations from abnormal changes. Enterprise E010 showed moderate deviations and activated the second-level warning, which successfully captured the unexpected increase in electricity consumption, which may be due to the unforeseen increase in production activities, reflecting the model's sensitivity to potential problems.

[0097] Most critically, enterprise E025 experienced severe deviations and activated the third-level warning, with actual electricity consumption far exceeding the forecast. This immediately aroused concern about the pressure on electricity supply and prompted the need to activate emergency plans to prevent the risk of power outages.

[0098] Compared with traditional methods, this invention improves the accuracy of identifying patterns of electricity consumption behavior through space-time cubes and multi-scale analysis, especially in dealing with seasonality, regionality and the impact of emergencies. The space-time coupling prediction model reduces the prediction error rate by 15% and increases the warning accuracy rate to 85% by comprehensively considering multiple factors, effectively supporting the decision-making of power management departments, ensuring the stable operation of the industrial chain, and showing significant innovation and advantages.

[0099] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

Claims

1. An industrial chain electricity consumption early warning method based on time scale analysis, characterized in that: include: Collect the electricity consumption, geographical location, climate conditions and holidays of enterprises in the industrial chain, and store the pre-processed data in a multi-source database; With time as the axis and geographic location, climate conditions, and holidays as the dimensions, a multidimensional space-time cube is constructed, and the original data is converted into an ordered storage structure in the space-time cube to obtain space-time cube data; Based on the space-time cube data, the long-term trend characteristics, seasonal characteristics and periodic characteristics are extracted by stratification according to the time scale, and the electricity consumption of each spatial unit at different time scales is analyzed; Based on the space-time cube data, the spatial clustering of electricity consumption distribution is analyzed through clustering algorithm; Combined with the clustering results, a spatiotemporal coupling prediction model is constructed to output the predicted estimated value of electricity consumption; The warning level is determined based on the deviation between the actual power consumption and the predicted estimated power consumption.

2. The method for early warning of power consumption in an industrial chain based on time scale analysis according to claim 1, characterized in that: The pre-processing comprises the following steps: Collect the electricity consumption of enterprises in the industrial chain, integrate the time axis of holidays and electricity consumption, and form a time series; Capture the climate conditions of the target area, integrate the enterprise's geographical location using GIS technology, and standardize it to match the climate conditions; Set up data receiving ports in multi-source databases, perform format checks and standardization conversions on the received raw data, ensure that all data follows a unified data model, and implement data cleaning processing.

3. The method for early warning of power consumption in an industrial chain based on time scale analysis according to claim 2 is characterized in that: The construction of a multidimensional space-time cube comprises the following steps: Define the time dimension, using hours as the basic unit, and build multiple time levels of day, week, month, quarter, and year upwards. Convert the timestamp in the original data to the selected time level, and map each record to the corresponding hour, day, or week time unit. Refine the spatial dimension and use the hexagonal grid method to subdivide the geographic space. The lowest level is refined to blocks or more microscopic grid units. Each unit corresponds to the longitude and latitude range in the GIS coordinate system. Based on the enterprise's geographical location, each observation point is mapped to a pre-defined spatial grid to determine its position in the cube. Set additional dimensions, assign codes to each additional dimension, and design multi-level classification for each dimension; The key is a combination of time dimension, space dimension and additional dimension, and the value is a storage structure of the power consumption at the corresponding location. The original data is stored in a multidimensional space-time cube to obtain space-time cube data.

4. The method for early warning of power consumption in an industrial chain based on time scale analysis according to claim 3 is characterized in that: The step of extracting long-term trend features, seasonal features and periodic features by stratification according to time scale includes the following steps: The time dimension of the space-time cube data is divided into long-term, medium-term and short-term, corresponding to long-term trends, seasonal fluctuations and short-term random changes; The sliding window method is used to slice the data. The long-term trend is defined as the interannual change, and the time window is set to one year; the medium-term fluctuation corresponds to the seasonal cycle, and the window is set to one quarter; the short-term change focuses on daily fluctuations and special events, and the window is set to one week; The original data is resampled and converted into data series at different time scales. The original daily data is aggregated by year, season, and week to obtain the average electricity consumption at the corresponding scale.

5. The method for early warning of power consumption in an industrial chain based on time scale analysis according to claim 4 is characterized in that: The method of analyzing the spatial aggregation of power consumption distribution by clustering algorithm comprises the following steps: Use the K-Means algorithm to cluster the geographical locations of enterprises and identify areas with similar electricity consumption behaviors; By calculating the geographic distance, the spatial weight matrix is ​​used to analyze the correlation of electricity consumption between neighboring enterprises. Use GIS hot spot analysis to identify areas with high electricity consumption; From micro to macro level, we analyze step by step, using multi-layer aggregation method to observe electricity consumption patterns at different spatial scales; and analyze the spatial aggregation of electricity consumption distribution.

6. The method for early warning of power consumption in an industrial chain based on time scale analysis according to claim 5, characterized in that: The constructing of the spatiotemporal coupling prediction model comprises: Combine the clustering results to extract spatiotemporal features, which include time series data, geographic location codes, seasonal labels, holiday indicators, and climate feature vectors; Define the coding of spatiotemporal features. t is time series data, indicating a time point; g is the geographic location code, which uses a longitude and latitude pair to indicate geographic coordinates; s is the seasonal label, which encodes the season, spring 1, summer 2, autumn 3, and winter 4; h is the holiday indicator, which is assigned a value based on whether it is a holiday, 0 for a working day and 1 for a holiday; c is the climate feature vector; The fused spatiotemporal features are expressed as, Among them, F is the fusion spatiotemporal feature, λ t is the weight of the time feature, is the normalized time feature of the time series data t, λ g is the weight of the geographic location feature, is the normalized geographic location feature vector of the geographic location code g, λ s is the weight of the seasonal feature, is the normalized seasonal feature of the seasonal label s, λ h is the weight of the holiday feature, is the normalized holiday feature of the holiday indicator h, λ c is the weight of the climate feature, is the normalized value of the climate characteristic vector c; Taking the fused spatiotemporal feature F as the input value, the predicted estimated value of the output power consumption is expressed as: in, is the predicted estimated value of electricity consumption; Φ(F) represents the prediction function, which is the output prediction model after comprehensively considering the spatiotemporal characteristics and complex information of the attention mechanism, and is used to map from input data to predicted values; STM(F(t)) is the spatiotemporal memory function, which processes the dynamic changes of time series features at different time points; Attention(F,t) is the spatiotemporal attention mechanism function, which is used to dynamically adjust the importance of different features; T is the prediction period.

7. The method for early warning of power consumption in an industrial chain based on time scale analysis according to claim 6, characterized in that: The deviation between the actual value of the power consumption and the predicted estimated value of the power consumption is expressed as: Among them, Y i is the actual value of the power consumption of the i-th sample; is the predicted estimated value of the electricity consumption of the i-th sample; When σ<5%σ, there is a slight deviation, indicating a first-level warning, and the power consumption of the industrial chain fluctuates normally; When 5%σ<σ<15%σ, a moderate deviation occurs, indicating a secondary warning, and there is a potential impact on the electricity consumption of the industrial chain; When σ>15%σ, a severe deviation occurs, indicating a third-level warning, and the electricity consumption of the industrial chain has serious effects on the power supply.

8. An industrial chain electricity consumption early warning system based on time scale analysis according to any one of claims 1 to 7, characterized in that: It includes data collection module, multi-dimensional space-time cube construction module, time series feature extraction module, spatial feature extraction module, space-time coupling prediction model construction module and power consumption warning module. The data collection module is used to collect the electricity consumption, geographical location, climate conditions and holidays of enterprises in the industrial chain, and store the pre-processed data in a multi-source database; The multidimensional space-time cube construction module is used to construct a multidimensional space-time cube with time as the axis and geographical location, climate conditions, and holidays as the dimensions, and convert the original data into an ordered storage structure in the space-time cube to obtain space-time cube data; The time series feature extraction module is used to extract long-term trend features, seasonal features and periodic features based on the space-time cube data and stratify them according to the time scale, and analyze the power consumption of each spatial unit at different time scales; The spatial feature extraction module is used to analyze the spatial aggregation of power consumption distribution through a clustering algorithm based on the space-time cube data; The spatiotemporal coupling prediction model building module is used to combine the clustering results, build a spatiotemporal coupling prediction model, and output a predicted estimated value of electricity consumption; The power consumption warning module is used to determine the warning level according to the deviation value between the actual power consumption value and the predicted estimated power consumption value.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the industrial chain electricity consumption early warning method based on time scale analysis described in any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the industrial chain electricity consumption early warning method based on time scale analysis described in any one of claims 1 to 7 are implemented.

Citation Information

Cited By

  • Smart factory energy efficiency optimization method and system, electronic equipment and storage medium

    CN119990413A

  • Economic database construction method for algae industry

    CN121935414A