Population flow monitoring method, system, device and medium based on residential electricity consumption

Abstract

The application discloses a population flow monitoring method, system, device and medium based on resident electricity consumption, and the method is: acquiring hourly electricity consumption data of a target user in a preset time window, calculating electricity consumption difference values of adjacent time points based on the hourly electricity consumption data to obtain an electricity consumption fluctuation amplitude sequence; fitting the electricity consumption fluctuation amplitude sequence to obtain a fluctuation threshold, and identifying fluctuations with each fluctuation amplitude greater than the fluctuation threshold in the electricity consumption fluctuation amplitude sequence as a to-be-identified fluctuation sequence; extracting a cycle consistency index and a fluctuation duration index of the to-be-identified fluctuation sequence; if the cycle consistency index is less than or equal to a first shape threshold and the fluctuation duration index is greater than a second shape threshold, determining that the fluctuation is an effective human fluctuation; and counting fluctuation frequencies and fluctuation coefficients of the effective human fluctuation to determine that the target user is in a home state, so that the application can accurately identify the real home / away state of a user.

Description

Technical Field

[0001] This invention relates to the field of population monitoring, and more particularly to methods, systems, equipment, and media for monitoring population movement based on residential electricity consumption. Background Technology

[0002] As my country's urbanization process continues to deepen, urban population mobility has significantly increased, with increasingly frequent patterns of population movement such as "separation of residence and household registration," "migratory migration," and "short-term business travel," placing higher demands on the refined management of cities. Accurately grasping the spatiotemporal distribution of population movement is directly related to the rationality of community public service resource allocation, the timeliness of emergency management response, and the precision of livelihood security policies. Traditional methods relying on door-to-door censuses or periodic registrations have long data update cycles and significant lags, failing to reflect real-time dynamic changes in the population. Against this backdrop, residential electricity consumption behavior, as the most closely linked to daily life and possessing the highest spatiotemporal resolution, naturally has the unique advantage of being "unblockable at the household level"—regardless of the communication terminal used by the user, their electricity consumption behavior objectively and truthfully reflects their residential activity status. This makes household-level population movement monitoring based on electricity consumption data an innovative direction that combines technical feasibility with social governance value.

[0003] Current population mobility monitoring methods based on residential electricity consumption data mainly rely on simple electricity consumption statistics and comparison with fixed thresholds. However, the electricity consumption base varies greatly among different users—the standby load of large-sized residences may far exceed the normal activity electricity consumption of small-sized residences, and a uniform threshold simply cannot adapt to the diverse user groups. Summary of the Invention

[0004] This invention provides a method, system, device, and medium for monitoring population movement based on residential electricity consumption. It can distinguish between "electricity fluctuations caused by the automatic operation of electrical equipment" and "electricity fluctuations caused by real human activities" from residential electricity consumption data, thereby accurately identifying the user's actual home / away status.

[0005] This invention provides a method for monitoring population movement based on residential electricity consumption, comprising: Obtain hourly electricity consumption data of the target user within a preset time window, calculate the electricity consumption difference between adjacent times based on the hourly electricity consumption data, and obtain a sequence of electricity consumption fluctuation amplitudes; The fluctuation threshold is obtained by fitting the electricity consumption fluctuation amplitude sequence, and each fluctuation in the electricity consumption fluctuation amplitude sequence whose fluctuation amplitude is greater than the fluctuation threshold is identified as a fluctuation sequence to be identified. Extract the periodic consistency index and fluctuation duration index of the fluctuation sequence to be identified; For each fluctuation in the fluctuation sequence to be identified, the period consistency index is compared with a preset first morphological threshold, and the fluctuation duration index is compared with a preset second morphological threshold. If the period consistency index is less than or equal to the first morphological threshold and the fluctuation duration index is greater than the second morphological threshold, then the fluctuation is determined to be a valid human-induced fluctuation. The frequency and coefficient of the effective human-caused fluctuations are statistically analyzed. If the frequency of fluctuations is greater than a preset frequency threshold, or the coefficient of fluctuations is greater than a preset coefficient threshold, then the target user is determined to be at home.

[0006] This invention, through acquiring hourly electricity consumption data of target users and calculating the difference between adjacent times, transforms the raw electricity readings into a fluctuation amplitude sequence that intuitively reflects the activity level of electricity consumption behavior, providing a basic data carrier for all subsequent discriminant analyses. Secondly, the fluctuation amplitude sequence is subjected to distribution fitting to adaptively determine the fluctuation threshold. This allows for dynamic differentiation between low-amplitude device standby noise and effective electricity consumption events based on the electricity consumption characteristics of different users, avoiding misjudgments caused by using a uniform threshold. Simultaneously, tiny fluctuations below the threshold are directly filtered out, significantly reducing the signal size for subsequent processing. Furthermore, the periodicity index and fluctuation duration index of the fluctuation sequence to be identified, filtered by amplitude, are extracted. These two morphological features characterize the intrinsic attributes of fluctuations from different dimensions—the periodicity index can effectively capture the fluctuations produced by devices such as refrigerator compressors due to their fixed working cycles. The system identifies repetitive fluctuations in electricity consumption, distinguishing between short-term equipment start-ups and shutdowns and continuous changes in electricity consumption caused by human activity. These two indicators are then combined with preset first and second morphological thresholds to construct a dual-screening mechanism based on the fundamental morphological differences between electrical fluctuations and human-induced fluctuations. Fluctuations are only considered valid human-induced fluctuations when their periodic consistency is sufficiently low and their duration is sufficiently long, thus achieving accurate identification and filtering of electrical fluctuations. Finally, by statistically analyzing the frequency and coefficient of all valid human-induced fluctuations within a time window, the system represents the frequency of human intervention from the "number of occurrences" dimension and eliminates the dimensional impact of differences in electricity consumption bases among different users from the "relative fluctuation intensity" dimension. If any indicator exceeds the corresponding threshold, the user is determined to be at home, ensuring comprehensive coverage and stable identification of users with various electricity consumption characteristics. These progressively integrated and interconnected technical methods effectively decouple fluctuations caused by automatic electrical operation from fluctuations caused by real human activity, significantly reducing the misjudgment rate of population flow monitoring and providing reliable technical support for accurate identification of household-level population flow status based on residential electricity consumption data.

[0007] Further, the step of fitting the electricity consumption fluctuation amplitude sequence to obtain the fluctuation threshold includes: All fluctuation amplitudes in the electricity consumption fluctuation amplitude sequence are used as sample data and input into a Gaussian mixture model to decouple the overall distribution of the sample data into a first Gaussian distribution component and a second Gaussian distribution component. Determine the first probability density curve corresponding to the first Gaussian distribution component, and determine the second probability density curve corresponding to the second Gaussian distribution component; Calculate the intersection point of the first probability density curve and the second probability density curve, and set the horizontal coordinate value corresponding to the intersection point as the fluctuation threshold.

[0008] This method decouples the fluctuation amplitude sequence into two Gaussian distribution components using a Gaussian mixture model, and takes the intersection of the two probability density curves as the fluctuation threshold. This approach abandons the traditional practice of setting a fixed threshold based on human experience, and can adapt to the different electricity consumption bases and fluctuation characteristics of different users, achieving objective and accurate quantification of the fluctuation threshold. This lays a reliable foundation for the effective separation of electrical fluctuations from human-induced fluctuations.

[0009] Further, the extraction of the periodic consistency index and fluctuation duration index of the fluctuation sequence to be identified includes: The adjacent fluctuation sampling points in the fluctuation sequence to be identified that continuously exceed the fluctuation threshold are merged into a single fluctuation event; Obtain the fluctuation amplitude change sequence of each sampling point within the fluctuation event, calculate the autocorrelation matching degree of the fluctuation amplitude change sequence relative to the preset fixed operating cycle of the equipment, and use the calculated autocorrelation matching degree as the cycle consistency index. The total number of hours that the fluctuation event lasts from the start time to the end time is tracked, and the total number of hours is used as the fluctuation duration indicator.

[0010] By merging consecutively exceeding threshold sampling points into fluctuation events and extracting their periodic consistency index (autocorrelation matching degree) and fluctuation duration index, a morphological feature dimension that distinguishes electrical fluctuations from human-induced fluctuations is constructed. Electrical equipment has fixed, short-term periodic start-stop patterns, while human activities are non-periodic and have a longer duration. By using the joint judgment of these two morphological indicators, real human activities can be identified from the essential differences in fluctuation patterns.

[0011] Further, the statistical analysis of the frequency and coefficient of the effective artificial fluctuations in the fluctuation sequence to be identified includes: Within the preset time window, the fluctuation events that are determined to be valid human-induced fluctuations are counted one by one according to the number of occurrences, and the total number obtained is taken as the fluctuation frequency. Extract the electricity consumption values ​​at each moment corresponding to all fluctuation events that are determined to be valid human-induced fluctuations within the preset time window, and form an effective fluctuation electricity consumption sequence. The dispersion of all values ​​in the effective fluctuating electricity consumption sequence relative to the sequence average is calculated to obtain a dispersion metric. The dispersion metric is then divided by the sequence average, and the ratio obtained by the division is used as the fluctuation coefficient. The sequence average is obtained by averaging the effective fluctuating electricity consumption sequence.

[0012] This approach, based on effective human-induced fluctuations, further statistically analyzes the fluctuation frequency and fluctuation coefficient (dimensionless relative fluctuation intensity), comprehensively determining the home / away status through these two dimensions. The fluctuation coefficient eliminates the influence of differences in the dimensions of electricity consumption among different users, making the judgment standard universally applicable across users and time periods. Combining the fluctuation frequency and fluctuation coefficient with "OR" logic effectively avoids misjudgment problems caused by a single indicator in special electricity consumption scenarios, improving the robustness and accuracy of population flow monitoring.

[0013] Furthermore, the present invention also includes: If the amplitude of each fluctuation in the power consumption fluctuation sequence is less than or equal to the fluctuation threshold, the fluctuation is directly identified as invalid noise fluctuation generated by the appliance standby or the automatic operation of the equipment, and is filtered out.

[0014] Furthermore, the present invention also includes: When the periodic consistency index is greater than the first morphological threshold, or when the fluctuation duration index is less than or equal to the second morphological threshold, the fluctuation is determined to be an invalid fluctuation generated by the automatic operation of the electrical equipment, and this fluctuation is removed from the statistics of valid human-caused fluctuations.

[0015] Furthermore, the present invention also includes: If the fluctuation frequency is less than or equal to a preset frequency threshold, and the fluctuation coefficient is less than or equal to a preset coefficient threshold, then the target user is determined to be away from home.

[0016] Another embodiment of the present invention provides a population flow monitoring system based on residential electricity consumption, comprising: the population flow monitoring system being used to perform the steps of the population flow monitoring method based on residential electricity consumption as described in the present invention.

[0017] Another embodiment of the present invention provides a terminal device, including: a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor. When the processor executes the computer program, it implements the steps of the population flow monitoring method based on residential electricity consumption as described in the present invention.

[0018] Another embodiment of the present invention also provides a computer-readable storage medium item, including: a stored computer program, which, when the computer program is running, controls the device where the computer-readable storage medium is located to perform the steps of the population flow monitoring method based on residential electricity consumption as described in the present invention. Attached Figure Description

[0019] To more clearly illustrate the technical solution of this application, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0020] Figure 1 This is a flowchart illustrating one embodiment of the population flow monitoring method based on residential electricity consumption provided in this application. Figure 2 This is a flowchart illustrating one embodiment of steps S201 to S203 provided in this application; Figure 3 A diagram illustrating the number of users away from home; Figure 4 A diagram illustrating the number of users staying at home; Figure 5 This is a schematic diagram of the inflow user statistics dashboard; Figure 6 This is a diagram illustrating the outflow of users statistics. Figure 7 This is a diagram illustrating the statistics of short-term vacant users. Figure 8 This is a schematic diagram of a dashboard for counting the number of users with long-term vacancy. Detailed Implementation

[0021] To make the objectives, technical solutions, and advantages of this application clearer, the technical solutions of this application will be clearly and completely described below with reference to the accompanying drawings of the embodiments. Obviously, the described embodiments are only some embodiments of this application, not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0022] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application pertains; the terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit the application; the terms “comprising” and “having”, and any variations thereof, in the specification, claims, and foregoing description of the drawings are intended to cover non-exclusive inclusion.

[0023] In the description of the embodiments of this application, technical terms such as "first" and "second" are used only to distinguish different objects and should not be construed as indicating or implying relative importance or implicitly specifying the number, specific order, or primary and secondary relationship of the indicated technical features. In the description of the embodiments of this application, "multiple" means two or more, unless otherwise explicitly defined.

[0024] In this document, the term "embodiment" means that a particular feature, structure, or characteristic described in connection with an embodiment may be included in at least one embodiment of this application. The appearance of this phrase in various places throughout the specification does not necessarily refer to the same embodiment, nor is it a separate or alternative embodiment mutually exclusive with other embodiments. It will be explicitly and implicitly understood by those skilled in the art that the embodiments described herein can be combined with other embodiments.

[0025] In the description of the embodiments in this application, the term "and / or" is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A existing alone, A and B existing simultaneously, and B existing alone. Additionally, the character " / " in this document generally indicates that the preceding and following related objects have an "or" relationship.

[0026] In the description of the embodiments of this application, the term "multiple" refers to two or more (including two), similarly, "multiple sets" refers to two or more (including two sets), and "multiple pieces" refers to two or more (including two pieces).

[0027] In the description of the embodiments of this application, unless otherwise expressly specified and limited, technical terms such as "installation," "connection," "joining," and "fixing" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral part; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; they can refer to the internal communication of two components or the interaction between two components. For those skilled in the art, the specific meaning of the above terms in the embodiments of this application can be understood according to the specific circumstances.

[0028] See Figure 1 To accurately identify a user's actual home / away status, an embodiment of the present invention provides a population flow monitoring method based on residential electricity consumption, including steps S101 to S105: Step S101: Obtain hourly electricity consumption data of the target user within a preset time window, calculate the electricity consumption difference between adjacent times based on the hourly electricity consumption data, and obtain the electricity consumption fluctuation amplitude sequence. In some embodiments, for a target user, hourly measured electricity consumption values ​​are collected within a preset time window (e.g., 24 hours) to form a time-series electricity consumption dataset within that time window. Then, within that time window, the actual electricity consumption of the current hour t is sequentially recorded in chronological order. Compared with the actual electricity consumption in the previous hour t-1 Subtract and take the absolute value, according to the formula The difference in electricity consumption between two adjacent hourly times is calculated; this difference represents the magnitude of electricity consumption variation within that adjacent time period. Then, the differences calculated for all adjacent times within the time window are arranged in chronological order to form a sequence of the target user's electricity consumption fluctuations within that time window. Each element in this sequence... Each corresponds to a specific fluctuation amplitude, which is used for subsequent comparison of fluctuation thresholds and identification of fluctuation events.

[0029] It should be noted that when t is the starting time within the time window, since there is a lack of data for the previous time, this time is not included in the difference calculation. All other times are processed one by one in this way, and the final time series length is equal to the number of hours contained in the time window minus one.

[0030] Step S102: Fit the power consumption fluctuation amplitude sequence to obtain a fluctuation threshold, and identify each fluctuation in the power consumption fluctuation amplitude sequence whose fluctuation amplitude is greater than the fluctuation threshold as a fluctuation sequence to be identified; Please refer to Figure 2 In some embodiments, fitting the electricity consumption fluctuation amplitude sequence to obtain the fluctuation threshold includes steps S201 to S203: Step S201: Take all the fluctuation amplitudes in the electricity consumption fluctuation amplitude sequence as sample data and input them into the Gaussian mixture model to decouple the overall distribution of the sample data into a first Gaussian distribution component and a second Gaussian distribution component. In some embodiments, all fluctuation amplitudes in the electricity consumption fluctuation amplitude sequence are used as sample data and input into a Gaussian mixture model. The core processing logic of this model is that it assumes the overall distribution of the sample data is not a single form, but rather a superposition of two Gaussian distribution components with different properties. The first Gaussian distribution component corresponds to the small electricity consumption fluctuations caused by scenarios such as appliances being in standby mode or equipment operating automatically in an unattended state; these fluctuations are typically small and concentrated. The second Gaussian distribution component corresponds to the significant electricity consumption changes caused by human activity, such as switching appliances on and off or using high-power equipment; these fluctuations are larger and more dispersed. By performing expectation-maximization iterative fitting on the sample data using the Gaussian mixture model, the joint probability distribution of the original fluctuation amplitude sequence can be decoupled into these two independent Gaussian distribution components, and their respective means can be estimated. and Standard deviation and .

[0031] Step S202: Determine the first probability density curve corresponding to the first Gaussian distribution component, and determine the second probability density curve corresponding to the second Gaussian distribution component; In some embodiments, after estimating the parameters of the two Gaussian distribution components, the first probability density curve corresponding to the first Gaussian distribution component and the second probability density curve corresponding to the second Gaussian distribution component are determined respectively. According to the definition of the probability density function of the Gaussian mixture model, the density curve of the first Gaussian distribution component is in the form of… The density curve of the second Gaussian distribution component is in the form of Where x is the independent variable representing the fluctuation amplitude. and These represent the proportions of the two types of fluctuations in the overall sample, and satisfy the following conditions: .

[0032] These two curves respectively depict the probability density of small fluctuations and significant fluctuations in different amplitude ranges. The shape of the curves is determined by their respective means and standard deviations, while the height of the curves is scaled by the weighting coefficients. The two curves exhibit an intersecting or overlapping distribution pattern on the amplitude coordinate axis.

[0033] Step S203: Calculate the intersection of the first probability density curve and the second probability density curve, and set the horizontal coordinate value corresponding to the intersection as the fluctuation threshold.

[0034] In some embodiments, based on the two probability density curves described above, the x-coordinate of their intersection point is calculated, and the value corresponding to this intersection point is set as the fluctuation threshold. This intersection point satisfies... That is, at this amplitude, the probability density of a sample point belonging to the small fluctuation component is exactly equal to the probability density of its belonging to the significant fluctuation component. Therefore, the x value obtained by solving the equation is the critical value for distinguishing between "small fluctuations" and "significant fluctuations". That is, the fluctuation threshold.

[0035] It should be noted that the fluctuation threshold serves as the benchmark for all subsequent fluctuation judgments. Fluctuations with amplitudes lower than or equal to the threshold are classified as normal noise fluctuations in an unmanned state; fluctuations with amplitudes higher than the threshold proceed to the subsequent morphological feature recognition stage.

[0036] This method decouples the fluctuation amplitude sequence into two Gaussian distribution components using a Gaussian mixture model, and takes the intersection of the two probability density curves as the fluctuation threshold. This approach abandons the traditional practice of setting a fixed threshold based on human experience, and can adapt to the different electricity consumption bases and fluctuation characteristics of different users, achieving objective and accurate quantification of the fluctuation threshold. This lays a reliable foundation for the effective separation of electrical fluctuations from human-induced fluctuations.

[0037] In some embodiments, fluctuations in the electricity consumption fluctuation amplitude sequence whose amplitudes are greater than the fluctuation threshold are identified as fluctuation sequences to be identified. Specifically, for the calculated electricity consumption fluctuation amplitude sequence, each fluctuation amplitude in the sequence is... The fluctuation threshold determined in advance by fitting a Gaussian mixture model Comparison is only possible when > Only then is the fluctuation determined to be likely caused by human activity or other valid electricity consumption events, and thus identified and retained for inclusion in the subsequent fluctuation sequence to be identified; Furthermore, the present invention also includes: if the amplitude of each fluctuation in the power consumption fluctuation sequence is less than or equal to the fluctuation threshold, then the fluctuation is directly determined to be invalid noise fluctuation generated by the appliance in standby mode or the equipment in automatic operation, and is filtered out. Specifically, if ≤ If the fluctuation is too low, it will be directly filtered out and not included in any subsequent analysis. The purpose of this screening layer is to use the fluctuation threshold to first eliminate the standby fluctuations of electrical appliances (such as routers, chargers, smart home standby, etc.) with extremely low amplitudes that cannot be caused by human activity. Only fluctuations with amplitudes exceeding the threshold are regarded as candidate fluctuation events that need to be further morphologically identified, thereby providing input data for the second layer of fine identification based on periodic consistency and fluctuation duration.

[0038] Step S103: Extract the periodic consistency index and fluctuation duration index of the fluctuation sequence to be identified; In some embodiments, step S103 includes: merging adjacent fluctuation sampling points in the fluctuation sequence to be identified that consecutively exceed the fluctuation threshold into a single fluctuation event; obtaining the fluctuation amplitude change sequence of each sampling point within the fluctuation event; calculating the autocorrelation matching degree of the fluctuation amplitude change sequence relative to a preset fixed operating cycle of the device; using the calculated autocorrelation matching degree as the cycle consistency index; and tracking the total number of hours the fluctuation event lasts from the start time to the end time, using the total number of hours as the fluctuation duration index. Specifically, in the fluctuation sequence to be identified, multiple consecutive sampling points exceeding the fluctuation threshold are considered... The scenarios are combined into a single fluctuation event, specifically according to the formula. To track, among which For indicator functions, The sampling time interval (i.e., 1 hour). The number of sampling points that continuously exceed the threshold is determined by tracking the total number of hours the fluctuation event lasts from its start to its end, and this total number is used as an indicator of the fluctuation duration. Simultaneously, the fluctuation amplitude change sequence of each sampling point within the fluctuation event is calculated using the formula... Calculate its autocorrelation matching degree relative to the preset fixed operating cycle p of the equipment, where p is the candidate cycle (such as the working cycle of a refrigerator compressor). The autocorrelation matching degree is the mean value of the change in electricity consumption within the time window, and it serves as the periodic consistency index.

[0039] It should be noted that the automatic operation of electrical appliances (such as the periodic switching of a refrigerator compressor) will generate regular autocorrelation peaks, resulting in higher PC values ​​and shorter PD values ​​(usually a fixed short working cycle, such as 10-30 minutes). In contrast, fluctuations caused by human activities have lower PC values ​​and longer PD values. Therefore, sampling points that continuously exceed the threshold are first merged into a complete fluctuation event, and then both morphological indicators are extracted simultaneously within this event. This provides a quantitative basis for distinguishing between the automatic operation of electrical appliances and human activities.

[0040] By merging consecutively exceeding threshold sampling points into fluctuation events and extracting their periodic consistency index (autocorrelation matching degree) and fluctuation duration index, a morphological feature dimension that distinguishes electrical fluctuations from human-induced fluctuations is constructed. Electrical equipment has fixed, short-term periodic start-stop patterns, while human activities are non-periodic and have a longer duration. By using the joint judgment of these two morphological indicators, real human activities can be identified from the essential differences in fluctuation patterns.

[0041] Step S104: For each fluctuation in the fluctuation sequence to be identified, the period consistency index is compared with a preset first form threshold, and the fluctuation duration index is compared with a preset second form threshold. If the period consistency index is less than or equal to the first form threshold and the fluctuation duration index is greater than the second form threshold, then the fluctuation is determined to be a valid human-induced fluctuation. In some embodiments, for each fluctuation in the fluctuation sequence to be identified, the average value of the periodic consistency index of that fluctuation is used. Compared with the preset first form threshold Compare the two, and simultaneously measure the duration of the fluctuation. With the preset second form threshold Comparison is only possible when and Only when both conditions are met is the fluctuation considered a valid artificial fluctuation and retained. It should be noted that the essence of this judgment logic is as follows: the automatic operation of electrical appliances (such as refrigerator compressors) has a fixed short working cycle (usually 10-30 minutes) and exhibits regular periodic autocorrelation peaks, therefore its PC value is high and PD value is short; while the power consumption fluctuations caused by human activities often have a low PC value (non-periodic) and a long PD value (longer duration and more complex changes), hence the use of... and As a joint discrimination boundary.

[0042] It should be noted that, The value is the lower quantile (e.g., the 10th percentile) of the periodic consistency of users with the user label being idle, meaning that 90% of the periodic consistency of appliance fluctuations is higher than this value. The value is the percentile (e.g., the 90th percentile) of the duration of fluctuations for users whose user tags are vacant, meaning that 90% of the appliance fluctuation durations are less than this value.

[0043] In some embodiments, the present invention further includes: when the periodic consistency index is greater than the first morphological threshold, or when the fluctuation duration index is less than or equal to the second morphological threshold, then determining that the fluctuation is an invalid fluctuation generated by the automatic operation of the electrical equipment, and removing this fluctuation from the statistics of valid human-caused fluctuations. Specifically, the automatic operation of electrical equipment (such as a refrigerator compressor) usually has a fixed short-term working cycle (e.g., 10-30 minutes), exhibiting a strong periodicity and short duration in its fluctuation pattern, i.e., a high PC value and a short PD value; while the electricity consumption fluctuations caused by real human activities do not have this fixed periodicity, and the duration is often longer and more complex. Therefore, when the periodic consistency index of a certain fluctuation is greater than the first morphological threshold, or when the fluctuation duration index is less than or equal to the second morphological threshold, then determining that the fluctuation is an invalid fluctuation generated by the automatic operation of the electrical equipment, and removing this fluctuation from the statistics of valid human-caused fluctuations. Greater than the preset first-form threshold At that time, or the duration of the fluctuation. Less than or equal to the preset second-form threshold If the fluctuation is found to be invalid due to the automatic operation of electrical equipment, it will be completely removed from the statistics of subsequent valid human-caused fluctuations.

[0044] Step S105: Calculate the frequency and coefficient of the effective human-induced fluctuations. If the frequency is greater than a preset frequency threshold or the coefficient is greater than a preset coefficient threshold, then determine that the target user is at home.

[0045] In some embodiments, the step of statistically analyzing the frequency and coefficient of effective human-induced fluctuations in the fluctuation sequence to be identified includes: within a preset time window, progressively counting the fluctuation events identified as effective human-induced fluctuations by the number of occurrences, and using the total number of occurrences as the fluctuation frequency; extracting the electricity consumption values ​​at each moment corresponding to all fluctuation events identified as effective human-induced fluctuations within the preset time window to form an effective fluctuating electricity consumption sequence; calculating the dispersion of all values ​​in the effective fluctuating electricity consumption sequence relative to the sequence average to obtain a dispersion metric, and dividing the dispersion metric by the sequence average, using the ratio obtained by the division as the fluctuation coefficient, wherein the sequence average is obtained by averaging the effective fluctuating electricity consumption sequence. Specifically, within a preset time window, progressively counting the fluctuation events identified as effective human-induced fluctuations by the number of occurrences, and using the total number of occurrences as the fluctuation frequency. First, the number of effective human-induced fluctuations within a time window; second, the actual electricity consumption at each moment corresponding to all fluctuation events identified as effective human-induced fluctuations within that time window. These electricity consumption values ​​are combined into an effective fluctuating electricity consumption sequence; then, the average value of all electricity consumption values ​​in the sequence is calculated relative to the sequence average. The degree of dispersion is determined to obtain a discrete measure value, and then the standard deviation is obtained by taking the square root of the discrete measure value. Finally, according to the formula Standard deviation Dividing by the series mean, the resulting ratio is the volatility coefficient. .

[0046] In some embodiments, sequence average The calculation formula is: .

[0047] In some embodiments, standard deviation The calculation formula is: In the formula, The number of hours included in a sliding time window (e.g., a 24-hour time window). ).

[0048] It should be noted that the volatility coefficient It is a dimensionless indicator, which eliminates the influence of the absolute value of electricity consumption between different users and different time periods, so that the fluctuation characteristics can be effectively compared in cross-household and cross-time period scenarios. The larger the coefficient, the higher the relative fluctuation.

[0049] It should be noted that the constraint condition is: when When the volatility coefficient is meaningless, it is denoted as . .

[0050] This approach, based on effective human-induced fluctuations, further statistically analyzes the fluctuation frequency and fluctuation coefficient (dimensionless relative fluctuation intensity), comprehensively determining the home / away status through these two dimensions. The fluctuation coefficient eliminates the influence of differences in the dimensions of electricity consumption among different users, making the judgment standard universally applicable across users and time periods. Combining the fluctuation frequency and fluctuation coefficient with "OR" logic effectively avoids misjudgment problems caused by a single indicator in special electricity consumption scenarios, improving the robustness and accuracy of population flow monitoring.

[0051] In some embodiments, if the fluctuation frequency is greater than a preset frequency threshold, or the fluctuation coefficient is greater than a preset coefficient threshold, then the target user is determined to be at home. Specifically, within a preset time window, after filtering out electrical appliance fluctuations, the frequency of valid human activity fluctuations is counted. And effective human activity fluctuation coefficient When either of these two indicators exceeds the corresponding threshold, i.e. ,in, Time window Frequency of human activity fluctuations following fluctuations in the internal power eliminator; Fluctuation frequency threshold; Time window The fluctuation coefficient of human activity after the internal electrical fluctuation; The fluctuation coefficient threshold determines whether the target user is at home.

[0052] The judgment rule adopts "or" logic, that is, as long as either the frequency or the intensity of human activity reaches the threshold level, the user is considered to be at home, thus avoiding the failure to judge the true situation of being at home due to insufficient single indicators.

[0053] In some embodiments, the present invention further includes: if the fluctuation frequency is less than or equal to a preset frequency threshold and the fluctuation coefficient is less than or equal to a preset coefficient threshold, then the target user is determined to be in an away-from-home state. Specifically, when both the effective human activity fluctuation frequency and the effective human activity fluctuation coefficient are less than or equal to the coefficient threshold within a time window, i.e. If so, the target user is determined to be away from home.

[0054] The determination rule adopts the "AND" logic, that is, the user is considered to be away from home only when the frequency and fluctuation intensity of effective human activities are both below the threshold, so as to ensure the rigor of the status judgment and avoid misjudgment due to abnormal fluctuations of a single indicator.

[0055] It should be noted that the preset frequency threshold Coefficient threshold The threshold is not arbitrarily set or fixed based on experience, but rather derived through cluster analysis of all historical electricity users in the region. It is based on the statistical analysis of electricity consumption fluctuations of a specific user group with clear physical meaning within the cluster results—the "vacant user" group. The reason for using clustering to calibrate the threshold is that different user groups exhibit significantly different electricity consumption patterns: a long-term unoccupied vacant house has almost constant electricity consumption, with only very low-amplitude standby noise from equipment; while a normally occupied household may experience some electricity consumption fluctuations even when the household is away, due to the periodic automatic operation of devices such as refrigerators, routers, and smart home appliances. If a uniform, experience-based fixed threshold is applied to all users, standby load fluctuations in large apartments may be misjudged as human activity, while genuine human activity fluctuations in small apartments may be missed due to their smaller amplitude. Therefore, it is necessary to adaptively determine the discrimination boundary applicable to the entire user group in the region, taking into account the electricity consumption base and fluctuation characteristics of different users.

[0056] The specific implementation process of cluster analysis is as follows. First, acquire the daily electricity consumption data of all users in the region within a preset statistical period (e.g., the past month or quarter), and construct a daily electricity consumption feature vector for each user. Using the user ID plus the natural day as the unique primary key, the electricity consumption data of the same user at 24 specific times on the same day are used as 24 features, corresponding to the electricity consumption in each time period from 00:00 to 23:00, forming a standard daily electricity consumption time series dataset. Then, using these feature vectors as input samples, a hierarchical clustering algorithm is used to perform unsupervised clustering of all users in the region. Hierarchical clustering calculates the Euclidean distance between samples or clusters, gradually merging samples with the highest similarity from bottom to top, and finally constructing a complete clustering hierarchy. The advantage of this algorithm is that it does not require pre-specifying the number of clusters, and the number of cluster categories can be flexibly determined according to the actual data distribution characteristics. Moreover, the clustering results have a clear hierarchical structure, which facilitates business personnel to interpret and label the physical meaning of different categories in conjunction with common sense about electricity consumption. After completing the first hierarchical clustering, the percentage of tags in each user's first cluster within each period is calculated based on the results, and statistical data is generated. The statistical data fields include user ID, percentage of tag 1, percentage of tag 2, ..., percentage of tag n, and the user's average daily electricity consumption. Here, n is the number of tags in the first hierarchical clustering. Then, a spatial topology graph is constructed, and the spatialized PageRank weights are calculated. The statistical data and PageRank weights are then used for a second hierarchical clustering, and Ward's method is used to calculate inter-class distances to construct a cluster tree. Pruning is performed using the optimal number of clusters, dividing users into stable categories corresponding to different user profiles, achieving accurate user category segmentation. Based on the results of the second clustering, manual or semi-automatic category labeling was performed. The criteria for classifying "vacant users" were: throughout the statistical period, the daily base load of these users remained consistently low with minimal daily fluctuations; the daily electricity consumption variance approached zero; the effective fluctuation frequency of daily electricity consumption exceeding the base load was almost zero; and the electricity consumption curves of these users exhibited a long-term stable, unremarkable pattern, clearly distinguishable from the daily electricity consumption curves of normal residential households. Generally, this category corresponds to houses that have been unoccupied for extended periods, vacant commercial properties that have not yet been occupied, or residences of users who are frequently away on business trips. Their electricity consumption comes entirely from standby appliances or automatic equipment operation, without any actual human-induced electricity consumption.

[0057] After clustering and labeling the "vacant users" category, all samples labeled as vacant users are extracted. Within a preset time window, the effective human activity fluctuation frequency and effective human activity fluctuation coefficient for each vacant user are calculated (no appliance fluctuations are filtered out during the calculation process, as all fluctuations for vacant users come from the automatic operation of appliances). Then, the average values ​​of these two indicators for all vacant users are taken as frequency thresholds. Sum of coefficient thresholds The value of is determined by the fact that the vacant user group represents the baseline of electricity consumption in the real state of "no one is home". For the vast majority of the time, there is no fluctuation in electricity consumption caused by any human activity. Although there may be slight fluctuations in a few periods due to the automatic operation of equipment such as refrigerator compressors, the frequency of such events is extremely low and the duration is fixed. By taking the average value, the numerical interference caused by these very few occasional fluctuations can be effectively smoothed out, making the calibrated threshold closer to the baseline level in the real state of being away from home.

[0058] It should be noted that the inflow status is that the user is at home at the current moment and was not at home at the previous moment; the outflow status is that the user is away from home at the current moment and was not away from home at the previous moment; the short-term vacancy status is that the number of consecutive days away from home D satisfies 7≤D≤30; and the long-term vacancy status is that the number of consecutive days away from home D>30. These criteria are used to generate a visual dashboard that displays the number of users in each status.

[0059] It should be noted that visual dashboards, such as Figures 3-8 As shown, where, Figure 3 This is a schematic diagram of the number of users who have left home, showing the number of users who have been identified as being away from home within a preset time window. It can be used to monitor the population outflow in a region. Figure 4 This is a schematic diagram of the number of users at home, showing the number of users identified as being at home within a time window, reflecting the size of the region's permanent and active population. Figure 5 This is a schematic diagram of the inflow user statistics dashboard, showing the number of users who are at home at the current time and were not at home at the previous time, representing the population of newly returning or newly entering the area. Figure 6 This is a schematic diagram of the outflow of users, showing the number of users who are away from home at the current time and who were not away from home at the previous time, representing the population that has just left the area. Figure 7 This is a schematic diagram of a short-term vacancy count dashboard, showing the number of users who have been away from home for 7 to 30 consecutive days, used to identify houses that are temporarily unoccupied. Figure 8 This is a schematic diagram of a long-term vacancy count dashboard, showing the number of users who have been away from home for more than 30 consecutive days, used to identify vacant houses that have been unoccupied for a long time.

[0060] This invention, through acquiring hourly electricity consumption data of target users and calculating the difference between adjacent times, transforms the raw electricity readings into a fluctuation amplitude sequence that intuitively reflects the activity level of electricity consumption behavior, providing a basic data carrier for all subsequent discriminant analyses. Secondly, the fluctuation amplitude sequence is subjected to distribution fitting to adaptively determine the fluctuation threshold. This allows for dynamic differentiation between low-amplitude device standby noise and effective electricity consumption events based on the electricity consumption characteristics of different users, avoiding misjudgments caused by using a uniform threshold. Simultaneously, tiny fluctuations below the threshold are directly filtered out, significantly reducing the signal size for subsequent processing. Furthermore, the periodicity index and fluctuation duration index of the fluctuation sequence to be identified, filtered by amplitude, are extracted. These two morphological features characterize the intrinsic attributes of fluctuations from different dimensions—the periodicity index can effectively capture the fluctuations produced by devices such as refrigerator compressors due to their fixed working cycles. The system identifies repetitive fluctuations in electricity consumption, distinguishing between short-term equipment start-ups and shutdowns and continuous changes in electricity consumption caused by human activity. These two indicators are then combined with preset first and second morphological thresholds to construct a dual-screening mechanism based on the fundamental morphological differences between electrical fluctuations and human-induced fluctuations. Fluctuations are only considered valid human-induced fluctuations when their periodic consistency is sufficiently low and their duration is sufficiently long, thus achieving accurate identification and filtering of electrical fluctuations. Finally, by statistically analyzing the frequency and coefficient of all valid human-induced fluctuations within a time window, the system represents the frequency of human intervention from the "number of occurrences" dimension and eliminates the dimensional impact of differences in electricity consumption bases among different users from the "relative fluctuation intensity" dimension. If any indicator exceeds the corresponding threshold, the user is determined to be at home, ensuring comprehensive coverage and stable identification of users with various electricity consumption characteristics. These progressively integrated and interconnected technical methods effectively decouple fluctuations caused by automatic electrical operation from fluctuations caused by real human activity, significantly reducing the misjudgment rate of population flow monitoring and providing reliable technical support for accurate identification of household-level population flow status based on residential electricity consumption data.

[0061] Based on the above method embodiments, corresponding apparatus embodiments are provided; An embodiment of the present invention provides a population flow monitoring system based on residential electricity consumption, comprising: the population flow monitoring system being used to perform the steps of the population flow monitoring method based on residential electricity consumption as described in the present invention.

[0062] It is understood that the above-described device embodiments correspond to the method embodiments of the present invention, and can implement the population flow monitoring method based on residential electricity consumption provided by any of the above-described method embodiments of the present invention.

[0063] It should be noted that the device embodiments described above are merely illustrative, and some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Furthermore, in the accompanying drawings of the device embodiments provided by this invention, the connection relationships between modules indicate that they have communication connections, which can specifically be implemented as one or more communication buses or signal lines. Those skilled in the art can understand and implement this without any creative effort.

[0064] Based on the above embodiments of the population flow monitoring method based on residential electricity consumption, another embodiment of the present invention provides a terminal device, which includes a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor. When the processor executes the computer program, it implements the population flow monitoring method based on residential electricity consumption of any embodiment of the present invention.

[0065] For example, in this embodiment, the computer program can be divided into one or more modules, which are stored in the memory and executed by the processor to complete the present invention. The one or more modules may be a series of computer program instruction segments capable of performing a specific function, which describe the execution process of the computer program in the terminal device.

[0066] The terminal device may be a desktop computer, laptop, handheld computer, or cloud server, etc. The terminal device may include, but is not limited to, a processor and a memory.

[0067] The processor can be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor can be a microprocessor or any conventional processor. The processor is the control center of the terminal device, connecting all parts of the terminal device via various interfaces and lines.

[0068] Based on the above-described method embodiments, another embodiment of the present invention provides a computer-readable storage medium including a stored computer program, wherein, when the computer program is executed, it controls the device where the computer-readable storage medium is located to execute the population flow monitoring method based on residential electricity consumption described in any of the above-described method embodiments of the present invention.

[0069] The modules / units integrated in the device / terminal equipment, if implemented as software functional units and sold or used as independent products, can be stored in a computer-readable storage medium. Based on this understanding, all or part of the processes in the above embodiments of the present invention can also be implemented by a computer program instructing related hardware. The computer program can be stored in a computer-readable storage medium, and when executed by a processor, it can implement the steps of the various method embodiments described above. The computer program includes computer program code, which can be in the form of source code, object code, executable files, or certain intermediate forms. The computer-readable medium can include: any entity or device capable of carrying the computer program code, a recording medium, a USB flash drive, a portable hard drive, a magnetic disk, an optical disk, a computer memory, a read-only memory (ROM), a random access memory (RAM), an electrical carrier signal, a telecommunication signal, and a software distribution medium, etc.

[0070] The above description represents the preferred embodiments of the present invention. It should be noted that those skilled in the art can make various improvements and modifications without departing from the principles of the present invention, and these improvements and modifications are also considered to be within the scope of protection of the present invention.

Claims

1. A method for monitoring population movement based on residential electricity consumption, characterized in that, include: Obtain hourly electricity consumption data of the target user within a preset time window, calculate the electricity consumption difference between adjacent times based on the hourly electricity consumption data, and obtain a sequence of electricity consumption fluctuation amplitudes; The fluctuation threshold is obtained by fitting the electricity consumption fluctuation amplitude sequence, and each fluctuation in the electricity consumption fluctuation amplitude sequence whose fluctuation amplitude is greater than the fluctuation threshold is identified as a fluctuation sequence to be identified. Extract the periodic consistency index and fluctuation duration index of the fluctuation sequence to be identified; For each fluctuation in the fluctuation sequence to be identified, the period consistency index is compared with a preset first morphological threshold, and the fluctuation duration index is compared with a preset second morphological threshold. If the period consistency index is less than or equal to the first morphological threshold and the fluctuation duration index is greater than the second morphological threshold, then the fluctuation is determined to be a valid human-induced fluctuation. The frequency and coefficient of the effective human-caused fluctuations are statistically analyzed. If the frequency of fluctuations is greater than a preset frequency threshold, or the coefficient of fluctuations is greater than a preset coefficient threshold, then the target user is determined to be at home.

2. The population flow monitoring method based on residential electricity consumption according to claim 1, characterized in that, The process of fitting the electricity consumption fluctuation amplitude sequence to obtain the fluctuation threshold includes: All fluctuation amplitudes in the electricity consumption fluctuation amplitude sequence are used as sample data and input into a Gaussian mixture model to decouple the overall distribution of the sample data into a first Gaussian distribution component and a second Gaussian distribution component. Determine the first probability density curve corresponding to the first Gaussian distribution component, and determine the second probability density curve corresponding to the second Gaussian distribution component; Calculate the intersection point of the first probability density curve and the second probability density curve, and set the horizontal coordinate value corresponding to the intersection point as the fluctuation threshold.

3. The population flow monitoring method based on residential electricity consumption according to claim 1, characterized in that, The extraction of the periodic consistency index and fluctuation duration index of the fluctuation sequence to be identified includes: The adjacent fluctuation sampling points in the fluctuation sequence to be identified that continuously exceed the fluctuation threshold are merged into a single fluctuation event; Obtain the fluctuation amplitude change sequence of each sampling point within the fluctuation event, calculate the autocorrelation matching degree of the fluctuation amplitude change sequence relative to the preset fixed operating cycle of the equipment, and use the calculated autocorrelation matching degree as the cycle consistency index. The total number of hours that the fluctuation event lasts from the start time to the end time is tracked, and the total number of hours is used as the fluctuation duration indicator.

4. The population flow monitoring method based on residential electricity consumption according to claim 1, characterized in that, The statistical analysis of the effective human-induced fluctuations and their coefficients in the fluctuation sequence to be identified includes: Within the preset time window, the fluctuation events that are determined to be valid human-induced fluctuations are counted one by one according to the number of occurrences, and the total number obtained is taken as the fluctuation frequency. Extract the electricity consumption values ​​at each moment corresponding to all fluctuation events that are determined to be valid human-induced fluctuations within the preset time window, and form an effective fluctuation electricity consumption sequence. The dispersion of all values ​​in the effective fluctuating electricity consumption sequence relative to the sequence average is calculated to obtain a dispersion metric. The dispersion metric is then divided by the sequence average, and the ratio obtained by the division is used as the fluctuation coefficient. The sequence average is obtained by averaging the effective fluctuating electricity consumption sequence.

5. The population flow monitoring method based on residential electricity consumption according to any one of claims 1-4, characterized in that, Also includes: If the amplitude of each fluctuation in the power consumption fluctuation sequence is less than or equal to the fluctuation threshold, the fluctuation is directly identified as invalid noise fluctuation generated by the appliance standby or the automatic operation of the equipment and is filtered out.

6. The population flow monitoring method based on residential electricity consumption according to any one of claims 1-4, characterized in that, Also includes: When the periodic consistency index is greater than the first morphological threshold, or when the fluctuation duration index is less than or equal to the second morphological threshold, the fluctuation is determined to be an invalid fluctuation generated by the automatic operation of the electrical equipment, and this fluctuation is removed from the statistics of valid human-caused fluctuations.

7. The population flow monitoring method based on residential electricity consumption according to any one of claims 1-4, characterized in that, Also includes: If the fluctuation frequency is less than or equal to a preset frequency threshold, and the fluctuation coefficient is less than or equal to a preset coefficient threshold, then the target user is determined to be away from home.

8. A population flow monitoring system based on residential electricity consumption, characterized in that, include: The population flow monitoring system is used to perform the population flow monitoring method based on residential electricity consumption as described in any one of claims 1-7.

9. A terminal device, characterized in that, include: One or more processors; A memory, coupled to the processor, for storing one or more programs; When the one or more programs are executed by the one or more processors, the one or more processors implement the steps of the population flow monitoring method based on residential electricity consumption as described in any one of claims 1-7.

10. A computer-readable storage medium, characterized in that, include: A stored computer program, wherein, when the computer program is executed, it controls the device containing the computer-readable storage medium to perform the steps of the population flow monitoring method based on residential electricity consumption as described in any one of claims 1-7.

Images