Big data energy management method for community intelligent monitoring

Abstract

The invention discloses a big data energy management method for intelligent monitoring of a community, and relates to the technical field of energy management, and the method is characterized in that through the hierarchical processing of community energy consumption data, the comprehensive analysis of a time dimension and a periodic rule is combined, and the energy use characteristics can be deeply mined. According to the processing mode, data of residential areas, public areas and commercial areas are layered, and detailed statistics is carried out according to categories of electric power, water resources, natural gas and the like, so that extraction of energy consumption characteristics is more targeted and systematic. Through time interval segmentation and clustering analysis, the behavior pattern of energy consumption is identified, and abnormal fluctuation characteristics are found in combination with entropy calculation, so that an abnormal source lower than a threshold value can be accurately positioned, and the efficiency and accuracy of troubleshooting are remarkably improved. Meanwhile, based on the dynamic prediction of the future energy consumption trend, in combination with the correlation analysis of the fluctuation region and the risk equipment, the potential problems in the energy consumption can be effectively pre-judged.

Description

Technical Field

[0001] The present invention relates to the field of energy management technology, and in particular to a big data energy management method for community intelligent monitoring. Background Art

[0002] The field of energy management technology includes energy production, distribution, consumption, and the collection, analysis, and optimization management of related data. The core content of this technical field is to reduce energy loss and improve energy utilization efficiency through scientific management and optimization of scheduling methods, while achieving effective allocation and management of energy resources.

[0003] Among them, the big data energy management method of community intelligent monitoring refers to the optimization management method for energy distribution and consumption through real-time collection, transmission and analysis of energy consumption and related environmental data within the community. The use of various energy sources such as electricity and water resources in the community is obtained through sensor equipment, and the energy consumption trend and distribution are analyzed in combination with data storage and processing; energy consumption modeling technology is used to generate energy use optimization plans; and automated control means are used to achieve real-time dynamic adjustment of community energy distribution and use.

[0004] Existing technologies are difficult to accurately identify regional and categorized energy consumption characteristics in community energy management. They are often based on overall data trends and are unable to analyze energy consumption in different areas such as residential and commercial areas. They are prone to overlooking abnormal high energy consumption problems in certain areas or equipment. For example, equipment that is frequently used in commercial areas may be difficult to detect in a timely manner due to the lack of separate energy consumption monitoring. In addition, in the identification of abnormal energy consumption, existing technologies are difficult to deeply analyze the specific sources of energy consumption fluctuations. They can only judge explicit problems, but cannot capture hidden anomalies such as uneven load or reduced equipment efficiency, which may delay the abnormal investigation process. For example, the energy consumption of lighting equipment in public areas increases abnormally due to aging, but it cannot be identified in advance by existing means. In terms of energy allocation, existing technologies are usually based on static path design, and it is difficult to adjust the energy transmission path according to dynamic changes, which may cause resource waste or local supply and demand imbalances. For example, some areas are overloaded during peak energy consumption, while other areas have idle resources. Summary of the invention

[0005] The present invention provides a big data energy management method for community intelligent monitoring.

[0006] In order to achieve the above object, the present invention adopts the following technical scheme:

[0007] A big data energy management method for community intelligent monitoring includes the following steps:

[0008] S1: Based on all records of community energy consumption, the community energy consumption data is processed in layers, the periodicity of each layer of community energy consumption data is analyzed, and the results of multi-dimensional energy cycle characteristic analysis are obtained;

[0009] S2: segmenting each layered data in the multi-dimensional energy cycle characteristic analysis result into time intervals and performing cluster analysis to obtain cluster analysis results, identifying the behavior pattern within the segment in each layered data according to the cluster analysis results, and generating segmented behavior pattern analysis results;

[0010] S3: Calculate the entropy value of the energy usage of each segmented data in the segmented behavior pattern analysis result, analyze the abnormal fluctuation characteristics of the energy usage in the segment according to the entropy value, extract the abnormal source of the energy usage characteristics below the preset energy usage threshold in the abnormal fluctuation characteristics, and generate the abnormal source analysis result;

[0011] S4: Dynamically predict the future energy consumption trend of the abnormal sources in the abnormal source analysis results, analyze the correlation between the fluctuation area and the risk equipment based on the prediction results, dynamically optimize the energy transmission path corresponding to the abnormal source based on the analysis, and obtain energy allocation and transmission path optimization results.

[0012] The present invention is improved in that the steps of obtaining the multi-dimensional energy cycle characteristic analysis results are specifically as follows:

[0013] S111: Based on all records of community energy consumption, energy usage data in residential areas, public areas and commercial areas are classified according to the time dimension, and the classified data are layered according to energy consumption categories including electricity, water resources and natural gas to generate layered energy consumption data;

[0014] S112: Based on the layered energy consumption data, the energy usage of each layer of data is counted according to multiple time periods, the daily energy consumption is analyzed, the data characteristics of each time dimension and energy consumption category are summarized, and a multi-dimensional energy cycle characteristic analysis result is generated.

[0015] The present invention is improved in that the step of obtaining the cluster analysis result is specifically as follows:

[0016] S211: Based on the multi-dimensional energy cycle characteristic analysis result, the energy usage data in each layered data is segmented according to the time interval, the segmentation point is determined by identifying the change trend of the energy consumption data, and the segmented data is reclassified to generate a time interval segmentation result;

[0017] S212: Based on the time interval segmentation result, cluster analysis is performed on each layer of segmented data using the formula:

[0018]

[0019] Calculate the similarity value d(i,j) between the i-th and j-th segments of data, classify the segmented data according to the similarity value d(i,j), and obtain the cluster analysis result;

[0020] Among them, x ik 、x jk They represent the kth characteristic value in the i-th and j-th segments of data respectively, and n represents the total number of characteristic values.

[0021] The present invention is improved in that the step of obtaining the segmented behavior pattern analysis result is specifically as follows:

[0022] S221: Based on the cluster analysis results, extract the energy usage data in each segment of each layered data, count and classify the energy usage of each time period according to the energy consumption proportion of community equipment and the preset energy consumption threshold, combine the running time and energy consumption characteristics of the equipment, identify the energy consumption behavior exceeding the preset energy consumption threshold and the periodic energy usage characteristics of the equipment, and generate a behavior pattern recognition result;

[0023] S222: Based on the behavior pattern recognition result, the segmented data is classified and counted according to the energy consumption behavior category, and the information is integrated based on the distribution characteristics of each energy category and community equipment type within the time period to generate a segmented behavior pattern analysis result.

[0024] The present invention is improved in that the step of analyzing the abnormal fluctuation characteristics of energy usage in the segment is specifically:

[0025] S311: Based on the energy usage of each segmented data in the segmented behavior pattern analysis result, the formula is used:

[0026]

[0027] Calculate the segment entropy value H;

[0028] Among them, y a represents the energy usage in the ath period within the segment, represents the total energy usage in the segment, b is the time period index for summation, and p a It represents the proportion of energy usage in period a to the total usage in the segment. is the mean of the energy usage in the segmented data set, and m represents the number of data points in the segment.

[0029] S312: Based on the segmented entropy value, the abnormal fluctuation characteristics of energy use are judged in combination with a preset range, and the abnormal fluctuation characteristic analysis results are generated by analyzing the concentration and fluctuation of energy use within the time period.

[0030] The present invention is improved in that the step of obtaining the abnormal source analysis result is specifically as follows:

[0031] S321: Based on the abnormal fluctuation feature analysis results, extract the abnormal energy usage features below the preset energy usage threshold in the segment, and identify the existing equipment efficiency decline, load imbalance and regional energy waste problems by analyzing the energy usage fluctuation residual in each period, and obtain the abnormal feature source information;

[0032] S322: Integrate the abnormal feature source information, combine the proportion data of each abnormal source, and generate an abnormal source analysis result.

[0033] The present invention is improved in that the step of analyzing the correlation between the fluctuation area and the risk equipment is specifically:

[0034] S411: Dynamically predict the future energy consumption trend of the abnormal sources in the abnormal source analysis result, using the formula:

[0035]

[0036] Calculate the probability P of transitioning from state i′ to state j′ at time t+1 i′,j′ (t+1), get the future energy consumption trend prediction result;

[0037] in, represents the accumulation of the intermediate state k′ from 1 to z, P i′,k′ (t) represents the probability of transitioning from state i′ to the intermediate state k′ at time t, T k′,j′ represents the element that transitions from the intermediate state k′ to the target state j′, t represents the current time step, t+1 represents the next time step, i′ is the current state index, j′ is the target state index, and k′ is the intermediate state index;

[0038] S412: Based on the future energy consumption trend prediction result, analyze the change characteristics of the fluctuation area and the risk equipment, analyze the correlation characteristics between the fluctuation area and the risk equipment in combination with the equipment energy consumption change data, and obtain the fluctuation area and equipment correlation analysis result.

[0039] The present invention is improved in that the steps of obtaining the energy allocation and transmission path optimization results are specifically as follows:

[0040] S421: Based on the analysis result of the association between the fluctuation area and the equipment, determine the energy transmission nodes and connection lines between the energy supply point and the energy demand point in the energy transmission path, and generate the energy transmission path;

[0041] S422: Based on the energy transmission path, dynamically optimize the path by adjusting the node distribution and connection lines of the energy transmission path to obtain energy allocation and transmission path optimization results.

[0042] Compared with the prior art, the advantages and positive effects of the present invention are:

[0043] In the present invention, by layering the community energy consumption data and combining the comprehensive analysis of the time dimension and periodic laws, the energy use characteristics can be deeply mined. This processing method makes the extraction of energy consumption characteristics more targeted and systematic by layering the data of residential areas, public areas and commercial areas, and refining the statistics according to categories such as electricity, water resources and natural gas. Through time interval segmentation and cluster analysis, the behavior pattern of energy consumption is identified and the abnormal fluctuation characteristics are found in combination with entropy value calculation, which helps to accurately locate the abnormal source below the threshold and significantly improves the efficiency and accuracy of problem troubleshooting. At the same time, based on the dynamic prediction of future energy consumption trends and the correlation analysis of fluctuation areas and risk equipment, it can effectively predict potential problems in energy use and form risk management capabilities for equipment and regions. Finally, the energy transmission path is optimized, and the allocation efficiency is dynamically optimized by adjusting the transmission nodes and line distribution, further reducing energy waste and improving the rationality of resource use. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for describing the embodiments. The drawings are only used to illustrate the implementation methods and are not to be considered as limitations of the present invention.

[0045] Figure 1 It is a main step flow chart of the present invention;

[0046] Figure 2 A flow chart for obtaining multi-dimensional energy cycle characteristic analysis results in the present invention;

[0047] Figure 3 A flow chart for obtaining cluster analysis results in the present invention;

[0048] Figure 4 A flow chart of obtaining segmented behavior pattern analysis results in the present invention;

[0049] Figure 5 A flow chart of analyzing abnormal fluctuation characteristics of energy usage within a segment in the present invention;

[0050] Figure 6 A flowchart for obtaining abnormal source analysis results in the present invention;

[0051] Figure 7 A flow chart for analyzing the correlation between the fluctuation area and the risk equipment in the present invention;

[0052] Figure 8 This is a flow chart for obtaining energy allocation and transmission path optimization results in the present invention. DETAILED DESCRIPTION

[0053] The technical scheme in the embodiment of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiment of the present invention. Obviously, the described embodiment is only a part of the embodiment of the present invention, not all of the embodiments. Based on the embodiment of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0054] Unless otherwise defined, all technical and scientific terms used in the present invention have the same meaning as commonly understood by technicians in the field of the present invention; the terms used in the present invention are only for the purpose of describing specific embodiments, and are not intended to limit the terms "including" and "having" and any variations thereof in the specification and claims of the present invention and the above-mentioned drawings, and are intended to cover non-exclusive inclusions.

[0055] In the description of the embodiments of the present invention, the technical terms "first", "second", etc. are only used to distinguish different objects, and cannot be understood as indicating or implying relative importance or implicitly indicating the number, specific order or primary and secondary relationship of the indicated technical features. In the description of the embodiments of the present invention, the meaning of "multiple" is more than two, unless otherwise clearly and specifically defined.

[0056] In the description of the embodiments of the present invention, the term "and / or" is only a description of the association relationship of associated objects, indicating that three relationships may exist. For example, A and / or B may represent: A exists alone, A and B exist at the same time, and B exists alone. In addition, the character " / " in this article generally indicates that the associated objects before and after are in an "or" relationship.

[0057] In the description of the embodiments of the present invention, the term "multiple" refers to more than two (including two). Similarly, "multiple groups" refers to more than two groups (including two groups), and "multiple pieces" refers to more than two pieces (including two pieces).

[0058] In the description of the embodiments of the present invention, the technical terms "center", "longitudinal", "lateral", "length", "width", "thickness", "up", "down", "front", "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inside", "outside", "clockwise", "counterclockwise", "axial", "radial", "circumferential", etc. indicate orientations or positional relationships based on the orientations or positional relationships shown in the accompanying drawings, and are only for the convenience of describing the embodiments of the present invention and simplifying the description, and do not indicate or imply that the referred device or element must have a specific orientation, be constructed and operated in a specific orientation, and therefore should not be understood as limiting the embodiments of the present invention.

[0059] In the description of the embodiments of the present invention, unless otherwise clearly specified and limited, technical terms such as "installed", "connected", "connected", "fixed" and the like should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection or an indirect connection through an intermediate medium, it can be the internal connection of two elements or the interaction relationship between two elements. For ordinary technicians in this field, the specific meanings of the above terms in the embodiments of the present invention can be understood according to the specific circumstances.

[0060] Example

[0061] The embodiment of the present invention provides a big data energy management method for community intelligent monitoring, such as Figure 1 As shown, the following steps are included:

[0062] S1: Based on all records of community energy consumption, the community energy consumption data is processed in layers, the periodicity of each layer of community energy consumption data is analyzed, and the results of multi-dimensional energy cycle characteristic analysis are obtained;

[0063] S2: segmenting each layer of data in the multi-dimensional energy cycle characteristic analysis results into time intervals and performing cluster analysis to obtain cluster analysis results, identifying the behavior patterns within the segments in each layer of data according to the cluster analysis results, and generating segment behavior pattern analysis results;

[0064] S3: Calculate the entropy value of the energy usage of each segmented data in the segmented behavior pattern analysis result, analyze the abnormal fluctuation characteristics of the energy usage in the segment according to the entropy value, extract the abnormal source of the energy usage characteristics below the preset energy usage threshold in the abnormal fluctuation characteristics, and generate the abnormal source analysis result;

[0065] S4: Dynamically predict the future energy consumption trend of the abnormal sources in the abnormal source analysis results, analyze the correlation between the fluctuation area and the risk equipment according to the prediction results, dynamically optimize the energy transmission path corresponding to the abnormal source according to the analysis, and obtain the energy allocation and transmission path optimization results;

[0066] The results of the multi-dimensional energy cycle characteristics analysis include peak energy consumption patterns, valley energy consumption patterns, and seasonal energy consumption patterns. The results of the segmented behavior pattern analysis include energy consumption behaviors that exceed the preset energy consumption threshold, the periodic energy usage characteristics of the equipment, and the energy consumption behavior patterns within the segment. The results of the abnormal source analysis include equipment efficiency decline, load imbalance, and regional energy waste. The results of energy allocation and transmission path optimization include the node optimization distribution of energy supply points, the node optimization distribution of energy demand points, and the transmission path optimization relationship of energy transmission lines.

[0067] like Figure 2As shown in the figure, the specific steps for obtaining the results of multi-dimensional energy cycle characteristic analysis are as follows:

[0068] S111: Based on all records of community energy consumption, energy usage data in residential areas, public areas and commercial areas are classified according to the time dimension, and the classified data are layered according to energy consumption categories including electricity, water resources and natural gas to generate layered energy consumption data;

[0069] Energy usage data in residential areas, public areas and commercial areas are extracted and divided into hourly, daily, weekly and monthly data sets according to the time dimension. For example, the 24-hour electricity consumption records in a residential area are collected hourly from the monitoring equipment to generate an hourly dimension data set containing 24 data points. The water resource consumption in one day is collected and classified as daily record data. The weekly and monthly data are generated by accumulating and summing the daily record data. The collected data are further classified according to the spatial dimension. The electricity consumption records in the residential area are classified into the same spatial category, and the natural gas usage data in the commercial area are classified and processed according to the same category. Finally, the above-mentioned stratified data are further aggregated according to the energy consumption category. For example, the daily electricity consumption, water resource usage and natural gas usage in the residential area are stored and managed separately, and finally a multi-level data set is formed to support subsequent analysis work, complete stratified processing, and obtain stratified energy consumption data.

[0070] S112: Based on the layered energy consumption data, the energy usage of each layer of data is counted according to multiple time periods, the daily energy consumption is analyzed, the data characteristics of each time dimension and energy consumption category are summarized, and a multi-dimensional energy cycle characteristic analysis result is generated;

[0071] Analyze the periodicity of energy consumption data of each layer of community, and conduct statistical analysis on the hourly recorded data in the layered energy consumption data by analyzing the electricity consumption, water consumption and natural gas consumption in different time periods. For example, the daily electricity consumption of the residential area within a week is counted, and the daily electricity consumption threshold is preset to 50 kWh. According to the statistical results, it is judged whether the daily electricity consumption exceeds the preset threshold. If there is a situation where the threshold is exceeded, the records of the exceeded time period are further analyzed. For example, the electricity consumption of a residential area from 6:00 to 9:00 in the morning and from 18:00 to 21:00 in the evening reaches 55 kWh respectively. and 60 kWh, and the morning and evening are summarized as the peak hours of electricity consumption. The daily peak electricity consumption data is accumulated and analyzed to obtain the weekly electricity consumption cycle characteristics, and the water resources and natural gas usage data are analyzed in the same way. For example, the natural gas usage in a commercial area increases week by week in winter. Through the analysis of weekly natural gas usage, it is found that the usage in the first week is 150 cubic meters, the second week is 200 cubic meters, and the third week is 250 cubic meters. It is concluded that the natural gas usage in winter shows a linear growth trend. Finally, the multi-dimensional energy cycle characteristic analysis results are obtained based on the statistical and analysis results.

[0072] like Figure 3 As shown in the figure, the steps for obtaining the cluster analysis results are as follows:

[0073] S211: Based on the results of the multi-dimensional energy cycle characteristic analysis, the energy usage data in each layer of data is segmented according to the time interval, the segmentation point is determined by identifying the change trend of the energy consumption data, and the segmented data is reclassified to generate a time interval segmentation result;

[0074] The optimal cutting algorithm is used to segment the stratified data into time intervals. First, the stratified data is preliminarily sorted according to the time dimension. For example, the daily electricity usage records in residential areas are counted as 24 data points by hour. The electricity consumption for each hour is compared hour by hour, and the difference in energy usage between adjacent time periods is calculated. The difference change rate is used as the basis for division. By detecting whether the difference change rate exceeds the preset threshold, the cutting points of adjacent time periods are marked. For example, the threshold for the difference change rate of electricity consumption is set to 20%. When the electricity consumption change rate in a certain period exceeds the threshold, the period is marked as a segmentation point. In this way, the data is divided into sub-data sets of multiple time intervals.

[0075] S212: Based on the time interval segmentation results, cluster analysis is performed on each layer of segmented data using the formula:

[0076]

[0077] Calculate the similarity value d(i,j) between the i-th and j-th segments of data, classify the segmented data according to the similarity value d(i,j), and obtain the cluster analysis result;

[0078] Among them, x ik 、x jk Respectively represent the kth characteristic value in the i-th and j-th segments of data. For example, the kth characteristic value can be an indicator such as the average power consumption, maximum power consumption, or total natural gas consumption during the period. It is calculated or directly obtained through the original data. For example, the average power consumption per hour is calculated by the power consumption data recorded by the monitoring system. If the hourly data is [20, 25, 30], the average power consumption is n represents the total number of characteristic values, usually equal. For example, if the data record contains three categories: electricity, water resources and natural gas, then n = 3. It represents the cumulative sum of characteristic values ​​k from 1 to n, and it represents the accumulation of the squares of the differences of each category (such as electricity, water resources, and natural gas).

[0079] Taking a residential area as an example, if the monitoring records from 6:00-9:00 show: Power usage:

[0080] 30 kWh, 40 kWh, 50 kWh; Water usage:

[0081] 2 cubic meters, 2.5 cubic meters, 3 cubic meters; natural gas usage: 10 cubic meters, 12 cubic meters, 15 cubic meters.

[0082] Through the above records, the average power usage during this period is calculated as kWh, the average water usage is Cubic meters, the average natural gas consumption is Cubic meters. These values ​​correspond to x in the formula ik For data from another time period, the average values ​​of each category can be calculated and used to calculate the Euclidean distance.

[0083] Assume that the two segmented data sets are:

[0084] The data in the i-th section: 40 kWh of electricity, 2.5 cubic meters of water resources, and 12.33 cubic meters of natural gas; the data in the j-th section: 50 kWh of electricity, 3 cubic meters of water resources, and 15 cubic meters of natural gas.

[0085] Calculate the square of the difference between the power categories: (x i1 -x j1 ) 2 =(40-50) 2 =100.

[0086] Calculate the square of the difference between water resource categories: (x i2 -x j2 ) 2 =(2.5-3) 2 =0.25.

[0087] Calculate the square of the difference for the natural gas category: (x i3 -x j3 ) 2 =(12.33-15) 2 =7.11.

[0088] Calculate the Euclidean distance:

[0089] The calculation result shows that the Euclidean distance between the i-th and j-th segments of data is 10.36. The segmented data is classified based on the Euclidean distance of 10.36. First, the classification range of cluster analysis is set, and the Euclidean distance is used as the core judgment basis. Usually, a classification threshold is set, for example, the classification threshold is 15. When the Euclidean distance between two segmented data is less than or equal to 15, it is determined that similar data belongs to the same category. If the Euclidean distance is greater than 15, it is determined to be different categories. For the two segmented data with a current Euclidean distance of 10.36, since the distance is less than the classification threshold of 15, it is determined that they belong to the same type of behavior pattern. The segmented data that meet the classification conditions are marked as a group, for example, marked as "behavior pattern A", and the Euclidean distance between other data segments and the current segment is calculated one by one in the entire segmented data set. The data that meets the classification range continues to be classified into "behavior pattern A". For segmented data whose Euclidean distance exceeds the classification threshold, they are classified into a new category of behavior patterns, for example, labeled as "behavior pattern B". By calculating the Euclidean distance of all segmented data and gradually classifying them, a cluster analysis result is finally generated, in which each category of segmented data has similar behavior pattern characteristics, such as peak energy consumption periods, low energy consumption periods, or time periods with concentrated use of specific equipment.

[0090] like Figure 4 As shown, the steps for obtaining the segmented behavior pattern analysis results are as follows:

[0091] S221: based on the cluster analysis results, extract the energy usage data in each segment of each layered data, count and classify the energy usage in each time period according to the energy consumption proportion of community equipment and the preset energy consumption threshold, combine the running time and energy consumption characteristics of the equipment, identify the energy consumption behavior exceeding the preset energy consumption threshold and the periodic energy usage characteristics of the equipment, and generate the behavior pattern recognition result;

[0092] For example, the energy usage during the period of 6:00-9:00 every day is extracted, and the electricity consumption for each hour is recorded as 30 kWh, 40 kWh and 50 kWh respectively. The total electricity consumption during this period is 120 kWh. Combined with the equipment energy consumption records, it is found that the air conditioning equipment consumes 60 kWh, accounting for 50% of the total electricity consumption during this period. According to the preset high-energy-consuming equipment identification threshold of 50%, the air conditioner is determined to be a high-energy-consuming equipment, and the behavior pattern during this period is classified as a high-power consumption behavior. At the same time, the use period of the air conditioner is analyzed, and it is found that its electricity consumption during the period of 6:00-9:00 every morning reaches more than 50% of the total electricity consumption, and the air conditioner is determined to have a periodic use characteristic. For water resource usage, the water resource consumption during the period of 8:00-10:00 every day is recorded as 2 cubic meters, 2.5 cubic meters and 3 cubic meters respectively, and the total water resource consumption is 7.5 cubic meters. Since no single equipment consumption exceeds 50% of the total consumption during this period, it is determined that there is no high-water-consuming equipment during this period. Finally, by combining the above process, the segmented behavior pattern is identified as including the high power consumption behavior and periodic energy usage characteristics of the air conditioner, and the behavior pattern recognition result is obtained.

[0093] S222: Based on the behavior pattern recognition results, the segmented data is classified and counted according to the energy consumption behavior category, and the information is integrated by combining the distribution characteristics of each energy category and community device type within the time period to generate the segmented behavior pattern analysis results;

[0094] By presetting the identification threshold of high power consumption behavior to 50% of the total power consumption and the identification threshold of high water consumption behavior to 40% of the total water consumption, the high power consumption behavior and high water consumption behavior in the segment are classified and summarized. For example, the total power consumption in a residential area during the morning peak period of 6:00-9:00 is 120 kWh, and the air conditioning power consumption is 60 kWh, accounting for 50%, which is determined to be a high power consumption behavior. The water consumption in the same period is 7.5 cubic meters, and the water consumption of no single device exceeds 40%, which is determined to be a non-high water consumption behavior. The high power consumption behavior of the air conditioner is classified as the high energy consumption behavior of household equipment. At the same time, for the periodic use characteristics, the use period of the air conditioning equipment is statistically 6:00-9:00 every day. Combined with the data throughout the year, it is further found that its use frequency is higher in summer, and 18:00-21:00 every day is also a high-frequency use period. Its periodic characteristics are classified as high-frequency use characteristics in summer. By integrating and classifying the above data, the segmented behavior pattern analysis results are generated.

[0095] like Figure 5 As shown, the steps for analyzing the abnormal fluctuation characteristics of energy usage within a segment are as follows:

[0096] S311: Based on the energy usage of each segment data in the segment behavior pattern analysis results, the formula is:

[0097]

[0098] Calculate the segment entropy value H;

[0099] Among them, y a Indicates the energy usage in the ath period within the segment, such as hourly electricity consumption, water consumption or natural gas consumption, which is directly obtained through monitoring equipment, such as extracting segment data through daily record data of electricity meters, water meters and natural gas meters. Indicates the total energy usage in the segment, that is, the result obtained by summing up the energy usage of all time periods. It is calculated by accumulating the monitoring record data. For example, the power consumption of each time period within 24 hours is accumulated as the total power consumption of the segment. b is the energy usage in each period, b is the period index in the summation operation, and p a It represents the proportion of energy usage in the ath period to the total usage in the segment, and is used to measure the relative weight of energy usage in each period within the segment. It is calculated that, for example, if the electricity consumption in a certain period is 30 kWh and the total electricity consumption is 120 kWh, then the proportion of this period is is the mean value of energy usage in the segmented data set, which is used to measure the overall level of energy usage in the segment. m represents the number of data points in the segment, that is, the number of time periods contained in the segmented data. is the mean of the squared standardized deviations of energy usage within the segment, which is used to measure the volatility of energy usage. It is calculated by averaging the squared deviations of each data point.

[0100] Taking the daily morning peak period of 6:00-9:00 in a residential area as an example, the recorded power consumption data is [30,40,50] kWh, and the total power consumption is: kilowatt-hour;

[0101] Calculate the proportion p of each period a :

[0102] Substitute the ratio into the entropy value part to calculate: H1 = -

[0103] (p1·ln(p1)+p2·ln(p2)+p3·ln(p3))=-

[0104] (0.25·-1.386+0.33·-1.108+0.42·-0.869)=1.07708;

[0105] Calculate the standardized deviation part:

[0106] First calculate the mean: kilowatt-hour;

[0107] Calculate the mean of the squared deviations:

[0108] =0.04167;

[0109] Add the two parts to get the comprehensive entropy value: H = H1

[0110] +0.04167=1.07708+0.04167=1.11875;

[0111] The piecewise entropy value calculated by the formula is 1.11875.

[0112] S312: judging the abnormal fluctuation characteristics of energy use based on the segmented entropy value and in combination with the preset range, and generating abnormal fluctuation characteristics analysis results by analyzing the concentration and fluctuation of energy use within the time period;

[0113] Based on the calculated segment entropy value result H=1.11875, compare it with the preset entropy value threshold range, for example, set the entropy value threshold range to [1.0, 1.2], and judge the abnormal fluctuation characteristics of energy usage in the segment by analyzing the entropy value result. If the entropy value is lower than the lower limit of 1.0, it means that the energy usage concentration in the segment is high, and energy consumption is concentrated in certain time periods or equipment. If the entropy value is higher than the upper limit of 1.2, it means that the energy usage in the segment is relatively dispersed and the fluctuation range is large. At this time, it is necessary to further analyze the energy usage of each time period in the segment. For example, by comparing the monitoring data, if it is found that the energy usage y1=30 kWh in the 6:00 time period is significantly lower than the average energy usage in other time periods kWh, it can be inferred that there may be a problem of equipment operating efficiency decline during this period. In addition, combined with periodic residual detection, feature extraction is further performed on the abnormal fluctuation interval. For example, for the peak period of abnormal fluctuation (such as the power consumption at 7:00 is y2 = 50 kWh, which is higher than the mean 25%), it can be inferred that it may be caused by excessive load during peak hours. After comprehensive analysis using the above steps, the abnormal fluctuation characteristics of energy usage within the segment are finally summarized, including concentrated fluctuations in time periods, reduced equipment efficiency, and load imbalance.

[0114] like Figure 6 As shown in the figure, the specific steps for obtaining the abnormal source analysis results are:

[0115] S321: Based on the abnormal fluctuation feature analysis results, extract the abnormal features of energy usage below the preset energy usage threshold in the segment, and identify the existing equipment efficiency decline, load imbalance and regional energy waste problems by analyzing the energy usage fluctuation residual of each period, and obtain the abnormal feature source information;

[0116] The abnormal fluctuation characteristics of energy consumption below the preset energy consumption threshold in the segment are extracted through the period residual detection method. The preset energy consumption threshold range is set to [15 kWh, 50 kWh], and the energy consumption fluctuation residual of each period in each segment is calculated. If the energy consumption of a certain period is significantly lower than the lower limit, for example, the energy consumption of a certain device in the segment from 10:00 to 11:00 is 12 kWh, which is lower than the lower limit of the threshold, it is preliminarily judged that the efficiency of the device in this period may decrease. By further analyzing the operating status and load conditions of the equipment, such as checking the abnormal current and voltage records in the equipment operation data, and combining the energy consumption fluctuations in other periods, the energy consumption of the equipment in the segment is reduced. The fluctuation pattern can be verified by the situation. If the energy usage in a certain period is higher than the upper limit, for example, the energy usage of the equipment in the segment from 14:00 to 15:00 is 52 kWh, which exceeds the upper threshold, the load distribution during the peak hours can be checked to confirm that the anomaly may be caused by load imbalance. In addition, by comparing the mean energy usage and fluctuation range of each period in the segment, the regional energy waste characteristics can be further identified. For example, specific equipment in the area still maintains high energy consumption during idle time periods, thereby extracting the characteristics of the abnormal source, and finally summarizing the specific abnormal sources such as decreased equipment efficiency, load imbalance and regional energy waste.

[0117] S322: Integrate the abnormal feature source information, combine the proportion data of each abnormal source, and generate an abnormal source analysis result;

[0118] By summarizing the distribution characteristics of each abnormal source, for example, recording the equipment number and specific time period where the equipment efficiency decreases as a data set, and recording the proportion of each abnormal source in the region, such as equipment efficiency decrease accounting for 40%, load imbalance accounting for 35%, and regional energy waste accounting for 25%, these results are further integrated to generate abnormal source analysis results.

[0119] like Figure 7 As shown in the figure, the steps for analyzing the correlation between the fluctuation area and the risk equipment are as follows:

[0120] S411: Dynamically predict the future energy consumption trend of the abnormal sources in the abnormal source analysis results, using the formula:

[0121]

[0122] Calculate the probability P of transitioning from state i′ to state j′ at time t+1 i′,j′ (t+1), get the future energy consumption trend prediction result;

[0123] in, It means to accumulate all possible states of the intermediate state k′ from 1 to z. By analyzing the number of all state categories in the community energy usage data, z, that is, the total number of possible states in the system, is determined. i′,k′ (t) represents the probability of transitioning from state i′ to the intermediate state k′ at time t. It is obtained by statistically analyzing the frequency of state transitions based on historical data. The probability of the device transitioning from the high energy consumption state i′ to the shutdown state k is analyzed through the device operation log. k′,j′ It represents the element in the transition probability matrix from the intermediate state k′ to the target state j′. The probability of a device transferring from the intermediate state k′ (short shutdown state) to the target state j′ (low energy consumption state) is calculated by counting the number of shutdown recovery times. t represents the current time step, which belongs to a certain moment in the discrete time series and is obtained from the time record in the community energy usage monitoring system. For example, energy consumption data is collected at a frequency of hours, days, months, etc. t+1 represents the next time step, which is derived from the time step t and represents the prediction moment. i′ is the current state index, which represents a specific initial state. j′ is the target state index, which represents the final state to be calculated. k′ is the intermediate state index, which represents the possible transition state. z represents the total number of states, which counts the number of possible states of the device in community energy management.

[0124] For example, a community energy device has three states (high energy consumption state i′=1, low energy consumption state j′=2, shutdown state k′=3), and its transition probability matrix is:

[0125] At time t=1, the probability of state i′=1 is: P 1,1 (t=1)=0.4,P 1,2 (t=1)=0.4,P 1,3 (t=1)=0.2;

[0126] Calculate the probability that the state transitions from i′=1 to j′=2 at time t+1=2:

[0127] Bring in data: P 1,2 (t=2)=P 1,1 (t=1)·T 1,2 +P 1,2 (t=1)·T 2,2 +P 1,3 (t=1)·T 3,2 ;

[0128] P 1,2 (t=2)

[0129] =0.4 0.3 + 0.4 0.7 + 0.2 0.2 = 0.12 + 0.28 + 0.04 = 0.44;

[0130] Calculation result P 1,2 (t=2)=0.44 indicates that at time t=2, the probability that the device transfers from the high-energy state i′=1 to the low-energy state j′=2 is 44%.

[0131] S412: Based on the future energy consumption trend prediction results, analyze the change characteristics of the fluctuation area and the risk equipment, analyze the correlation characteristics between the fluctuation area and the risk equipment in combination with the equipment energy consumption change data, and obtain the fluctuation area and equipment correlation analysis results;

[0132] According to the calculated prediction results, the probability of the device transferring from the high energy consumption state i′=1 to the low energy consumption state j′=2 is P 1,2 (t=2)=0.44. Combined with the abnormal source distribution data of the fluctuation area and the time series change records of the equipment status, the change characteristics of the fluctuation area in the prediction are analyzed. By extracting the energy consumption data characteristics of the fluctuation area, including the total energy consumption value, the fluctuation amplitude range and the fluctuation frequency, according to the preset high energy consumption fluctuation range setting (for example, the single fluctuation amplitude is greater than 50%, and the frequency is higher than 5 times / hour, the data is regarded as the high energy consumption fluctuation range), the fluctuation area that meets the conditions is screened out, and the list of risk equipment corresponding to each fluctuation area is obtained. The energy consumption change time series of each equipment is extracted, and the contribution rate between the equipment energy consumption and the total energy consumption of the area is compared. By analyzing the contribution rate higher than the preset threshold (such as the equipment contribution rate exceeds 30% of the total energy consumption of the area), such equipment is further marked as high-risk equipment. The distribution frequency of high-risk equipment and its position overlap with the fluctuation area are counted, and a one-to-one correspondence is established between the fluctuation area and the equipment with a position overlap higher than 80%. The correlation characteristics between the fluctuation area and the high-risk equipment are identified, and the correlation analysis results between the fluctuation area and the risk equipment are generated.

[0133] like Figure 8 As shown in the figure, the steps for obtaining the energy allocation and transmission path optimization results are as follows:

[0134] S421: Based on the analysis results of the correlation between the fluctuation area and the equipment, determine the energy transmission nodes and connection lines between the energy supply point and the energy demand point in the energy transmission path, and generate the energy transmission path;

[0135] Firstly, according to the results of abnormal source analysis, the key nodes in the energy transmission path are determined, including high-energy consumption nodes and energy bottleneck nodes. The energy distribution and transmission efficiency of each node are analyzed using historical energy transmission data to obtain the energy demand and supply ratio of each node. According to the preset energy transmission efficiency range (for example, within the range of 85%-95%), the node combination of the energy transmission path is optimized through the neighborhood search algorithm, including adjusting node distribution, removing inefficient nodes, adding auxiliary nodes and other operations. Then, the connection status of the transmission line is analyzed, the status of high-load lines and idle lines is recorded, and the line connection direction and transmission priority are adjusted according to the optimized node position. Through the optimization operation of nodes and lines, the dynamically optimized energy transmission path is obtained.

[0136] S422: Based on the energy transmission path, dynamically optimize the path by adjusting the node distribution and connection lines of the energy transmission path, and obtain energy allocation and transmission path optimization results;

[0137] Combined with the optimized node and line data, the transmission efficiency and load balancing status of each line are calculated one by one, and the preset range of transmission efficiency (such as between 90% and 95%) and load balancing ratio (such as not exceeding 10% of the node capacity) of each line are set. If the efficiency of a line is lower than the range or the load exceeds the ratio, the line is reallocated or a backup line is added. By simulating the actual energy transmission scenario, it is analyzed whether the adjusted line transmission results meet the energy supply and demand balance goals. For example, in a public area, after removing the lines with a transmission efficiency lower than 85%, a new line is added to increase the energy transmission speed. According to the adjustment results, it is calculated whether the overall energy allocation time is reduced, and finally a transmission path optimization result that meets the energy supply and demand optimization goals is generated.

[0138] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, not to limit them; although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the above embodiments, or replace some or all of the technical features therein by equivalents; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the scope of the technical solutions of the embodiments of the present invention, and they should all be included in the scope of the claims and specification of the present invention. In particular, as long as there is no structural conflict, the various technical features mentioned in each embodiment can be combined in any way. The present invention is not limited to the specific embodiments disclosed herein, but includes all technical solutions that fall within the scope of the claims.

Claims

1. A big data energy management method for community intelligent monitoring, characterized in that: The following steps are involved: S1: Based on all records of community energy consumption, the community energy consumption data is processed in layers, the periodicity of each layer of community energy consumption data is analyzed, and the results of multi-dimensional energy cycle characteristic analysis are obtained; S2: segmenting each layered data in the multi-dimensional energy cycle characteristic analysis result into time intervals and performing cluster analysis to obtain cluster analysis results, identifying the behavior pattern within the segment in each layered data according to the cluster analysis results, and generating segmented behavior pattern analysis results; S3: Calculate the entropy value of the energy usage of each segmented data in the segmented behavior pattern analysis result, analyze the abnormal fluctuation characteristics of the energy usage in the segment according to the entropy value, extract the abnormal source of the energy usage characteristics below the preset energy usage threshold in the abnormal fluctuation characteristics, and generate the abnormal source analysis result; S4: Dynamically predict the future energy consumption trend of the abnormal sources in the abnormal source analysis results, analyze the correlation between the fluctuation area and the risk equipment based on the prediction results, dynamically optimize the energy transmission path corresponding to the abnormal source based on the analysis, and obtain energy allocation and transmission path optimization results.

2. The big data energy management method for community intelligent monitoring according to claim 1 is characterized in that: The steps for obtaining the multi-dimensional energy cycle characteristic analysis results are specifically as follows: S111: Based on all records of community energy consumption, energy usage data in residential areas, public areas and commercial areas are classified according to the time dimension, and the classified data are layered according to energy consumption categories including electricity, water resources and natural gas to generate layered energy consumption data; S112: Based on the layered energy consumption data, the energy usage of each layer of data is counted according to multiple time periods, the daily energy consumption is analyzed, the data characteristics of each time dimension and energy consumption category are summarized, and a multi-dimensional energy cycle characteristic analysis result is generated.

3. The big data energy management method for community intelligent monitoring according to claim 1 is characterized in that: The steps for obtaining the cluster analysis results are specifically as follows: S211: Based on the multi-dimensional energy cycle characteristic analysis result, the energy usage data in each layered data is segmented according to the time interval, the segmentation point is determined by identifying the change trend of the energy consumption data, and the segmented data is reclassified to generate a time interval segmentation result; S212: Based on the time interval segmentation result, cluster analysis is performed on each layer of segmented data using the formula: Calculate the similarity value d(i,j) between the i-th and j-th segments of data, classify the segmented data according to the similarity value d(i,j), and obtain the cluster analysis result; Among them, x ik 、x jk They represent the kth characteristic value in the i-th and j-th segments of data respectively, and n represents the total number of characteristic values.

4. The big data energy management method for community intelligent monitoring according to claim 1 is characterized in that: The steps for obtaining the segmented behavior pattern analysis results are specifically as follows: S221: Based on the cluster analysis results, extract the energy usage data in each segment of each layered data, count and classify the energy usage of each time period according to the energy consumption proportion of community equipment and the preset energy consumption threshold, combine the running time and energy consumption characteristics of the equipment, identify the energy consumption behavior exceeding the preset energy consumption threshold and the periodic energy usage characteristics of the equipment, and generate a behavior pattern recognition result; S222: Based on the behavior pattern recognition result, the segmented data is classified and counted according to the energy consumption behavior category, and the information is integrated based on the distribution characteristics of each energy category and community equipment type within the time period to generate a segmented behavior pattern analysis result.

5. The big data energy management method for community intelligent monitoring according to claim 1 is characterized in that: The step of analyzing the abnormal fluctuation characteristics of energy usage in the segment is specifically as follows: S311: Based on the energy usage of each segmented data in the segmented behavior pattern analysis result, the formula is used: Calculate the segment entropy value H; Among them, y a represents the energy usage in the ath period within the segment, represents the total energy usage in the segment, b is the time period index for summation, and p a It represents the proportion of energy usage in period a to the total usage in the segment. is the mean of the energy usage in the segmented data set, and m represents the number of data points in the segment. S312: Based on the segmented entropy value, the abnormal fluctuation characteristics of energy use are judged in combination with a preset range, and the abnormal fluctuation characteristic analysis results are generated by analyzing the concentration and fluctuation of energy use within the time period.

6. The big data energy management method for community intelligent monitoring according to claim 1 is characterized in that: The steps for obtaining the abnormal source analysis result are specifically as follows: S321: Based on the abnormal fluctuation feature analysis results, extract the abnormal energy usage features below the preset energy usage threshold in the segment, and identify the existing equipment efficiency decline, load imbalance and regional energy waste problems by analyzing the energy usage fluctuation residual in each period, and obtain the abnormal feature source information; S322: Integrate the abnormal feature source information, combine the proportion data of each abnormal source, and generate an abnormal source analysis result.

7. The big data energy management method for community intelligent monitoring according to claim 1 is characterized in that: The step of analyzing the correlation between the fluctuation area and the risk equipment is specifically as follows: S411: Dynamically predict the future energy consumption trend of the abnormal sources in the abnormal source analysis result, using the formula: Calculate the probability P of transitioning from state i′ to state j′ at time t+1 i′,j′ (t+1), get the future energy consumption trend prediction result; in, represents the accumulation of the intermediate state k′ from 1 to z, P i′,k′ (t) represents the probability of transitioning from state i′ to the intermediate state k′ at time t, T k′,j′ represents the element that transitions from the intermediate state k′ to the target state j′, t represents the current time step, t+1 represents the next time step, i′ is the current state index, j′ is the target state index, and k′ is the intermediate state index; S412: Based on the future energy consumption trend prediction result, analyze the change characteristics of the fluctuation area and the risk equipment, analyze the correlation characteristics between the fluctuation area and the risk equipment in combination with the equipment energy consumption change data, and obtain the fluctuation area and equipment correlation analysis result.

8. The big data energy management method for community intelligent monitoring according to claim 1 is characterized in that: The steps for obtaining the energy allocation and transmission path optimization results are specifically as follows: S421: Based on the analysis result of the association between the fluctuation area and the equipment, determine the energy transmission nodes and connection lines between the energy supply point and the energy demand point in the energy transmission path, and generate the energy transmission path; S422: Based on the energy transmission path, dynamically optimize the path by adjusting the node distribution and connection lines of the energy transmission path to obtain energy allocation and transmission path optimization results.

Images