Prepayment heat supply network metering management method and system

Abstract

The invention discloses a prepayment heat supply network metering management method and system, and relates to the technical field of metering management. Comprising the following steps: performing multi-dimensional division on collected heat supply network metering data to generate a slice matrix containing a state field; different space and equipment combinations under the same time dimension are extracted from the slice matrix, heat difference cross analysis is carried out, and abnormal fragments are screened; constructing an adjacent node set by taking a space node in which the abnormal fragment is located as a center, performing state consistency analysis and measurement correction on the abnormal fragment, and updating a state field of the slice matrix; and if the corrected slice still has the abnormal state, triggering a dormant account verification process, calling historical slice data of a dormant node to carry out trend association analysis, and bringing an effective heat change sequence into final judgment and trend correction of the current node. According to the invention, the stability and the intelligent level of heat supply network metering management are enhanced, and the metering fairness and the operation and maintenance efficiency under a prepayment mechanism can be improved.

Description

Technical Field

[0001] This invention relates to the field of metering management technology, and in particular to a method and system for metering management of prepaid heating networks. Background Technology

[0002] With the continuous expansion of urban centralized heating systems and the gradual promotion of prepaid metering concepts, the heat transmission, metering, and settlement processes involved in heating network systems are becoming increasingly complex. Heating network metering management has become a crucial component of smart energy management. Existing heating network metering management systems commonly employ intelligent flow meters, wireless communication terminals, and centralized monitoring platforms to collect, aggregate, and process user heat consumption data. These systems, relying on IoT, big data, and cloud platform technologies, achieve remote visualization of user heat consumption, online settlement, and anomaly alarms, improving billing efficiency and operational management levels for heating companies. However, existing systems often operate on a "single-point measurement—data upload—timed settlement" model, lacking in-depth analysis and adaptive correction capabilities regarding heat flow states, spatial correlation characteristics, and multi-source anomalies. This makes it difficult to meet the demands for high-precision, closed-loop control and real-time correction in complex heating network environments.

[0003] For example, CN111935217A discloses an internet-based heating network management metering and billing system. By introducing intelligent flow meter devices, a network-wide communication module, and a terminal platform, it achieves functions such as heat monitoring, remote control, and balance reminders for prepaid users. While this solution has certain advantages in equipment integration and platform operation and maintenance, protecting the economic interests of heating providers, its data processing mode mainly remains at the level of static data acquisition and rule-driven automatic control, lacking structured, multi-dimensional, and trend evolution analysis of heat data. Especially when heat anomalies, metering drift, or dormant accounts occur at the user end, it cannot perform in-depth judgment and trend correction based on spatiotemporal topology and historical behavior models. Therefore, in complex heating scenarios, the accuracy of anomaly identification and the closed-loop processing capability are limited, making it difficult to achieve higher-level intelligent heating network metering management.

[0004] CN118093680B discloses a monitoring system and method for heating network metering data. This system improves the accuracy of heating network monitoring by polling and collecting data, caching and transmitting it, parsing it to generate multi-dimensional heat consumption reports, and correcting equipment operating parameters. This method enhances the understanding and monitoring capabilities of different heat-consuming units and improves the level of systematic analysis. However, its monitoring mechanism still relies mainly on data aggregation and analysis and parameter fine-tuning, failing to construct a segmented, state-driven dynamic metering model, and lacking expertise in diagnosing, correlating, and analyzing the evolution trends of abnormal segments. Furthermore, this solution fails to address metering drift and regional heat imbalance caused by dormant accounts or inactive nodes, potentially leading to technical bottlenecks in actual heating network operation and maintenance, such as "no traceability of data anomalies and no correction of distorted trends." Summary of the Invention

[0005] In view of the problems existing in the current heating network metering and management technology, this invention is proposed.

[0006] Therefore, the problem to be solved by this invention is how to construct a multidimensional slice matrix containing a state field, perform differential cross-analysis on heat data under different time and spatial dimensions, and combine historical trends and spatial correlation characteristics to achieve state consistency judgment and anomaly correction. Furthermore, a dormant account historical behavior model is introduced for trend compensation, which significantly improves the accuracy and intelligence level of the metering system.

[0007] To solve the above-mentioned technical problems, the present invention provides the following technical solution: In a first aspect, the present invention provides a prepaid heating network metering management method, which includes: dividing the collected heating network metering data into multiple dimensions to generate a slice matrix containing a status field; extracting different spatial and equipment combinations under the same time dimension from the slice matrix, performing heat difference cross-analysis, and filtering abnormal segments; constructing a set of adjacent nodes centered on the spatial node where the abnormal segment is located, and performing status consistency analysis and metering correction on the abnormal segment based on historical metering trends, and updating the status field of the slice matrix; if the corrected slice still has status abnormalities, triggering a dormant account verification process, calling the historical slice data of the dormant node for trend correlation analysis, and incorporating the effective heat change sequence into the final judgment and trend correction of the current node.

[0008] As a preferred embodiment of the prepaid heating network metering management method of the present invention, the slice matrix includes time period, node index, equipment identifier, equipment type label, quality level, heat value, prepaid account status field and status code.

[0009] As a preferred embodiment of the prepaid heating network metering management method of the present invention, the heat difference cross-analysis includes: extracting all segments with status codes of 0 in the slice matrix into a subset of slices of the same period according to a unified time period, and selecting comparison combination groups with the same equipment type; in the comparison combination group, constructing a heat reference path set according to node index and quality level, wherein each path represents the trend of heat value difference change under unit spatial offset; evaluating the heat difference of each path in the heat reference path set and calculating the deviation index.

[0010] As a preferred embodiment of the prepaid heating network metering management method of the present invention, the screening of abnormal segments includes: generating a candidate abnormal segment index set based on the dual indicators of deviation value amplitude and continuous occurrence frequency; writing the candidate abnormal segment index set back into the slice matrix, updating the corresponding status code to 1, and marking the source of the abnormal factor.

[0011] As a preferred embodiment of the prepaid heating network metering management method of the present invention, the step of performing state consistency analysis and metering correction on abnormal segments includes: for each segment in the candidate abnormal segment index set, extracting the corresponding spatial node and constructing an adjacent node set based on the spatial topology; for all slices in the adjacent node set, extracting historical metering trends according to the time dimension, constructing a heat change curve cluster of the same period, and marking fluctuation-sensitive nodes; comparing the state consistency of the heat data of the slice corresponding to the abnormal segment with the heat change curve cluster, and if the heat deviation range exceeds the dynamic tolerance, correcting the heat data of the segment according to the median of the trend.

[0012] As a preferred embodiment of the prepaid heating network metering management method of the present invention, the status field of the updated slice matrix includes: for the corrected abnormal segments, the corresponding status code is synchronously updated to 2 in the slice matrix, and the associated trend cluster identifier is recorded.

[0013] As a preferred embodiment of the prepaid heating network metering management method of the present invention, the trend correlation analysis includes: for segments with a corrected status code of 2, calling the user account list of the heating network node corresponding to the current building, and filtering the set of nodes with a status field of dormancy; extracting historical slice data corresponding to the current abnormal segment in the time dimension from the node set, and constructing a comparison matrix with the heat trend of the current node; performing trend normalization processing on the heat change sequence in the comparison matrix, and if there is a change pattern sequence that matches the current node, it is included in the original heat change trend for reconstruction and deduction.

[0014] Secondly, the present invention provides a prepaid heating network metering management system, which includes: The slice generation module divides the collected heating network metering data into multiple dimensions and generates a slice matrix containing status fields. The difference analysis module extracts different spatial and equipment combinations under the same time dimension from the slice matrix, performs cross-analysis of heat difference, and filters out abnormal segments. The status correction module constructs a set of adjacent nodes centered on the spatial node where the abnormal segment is located, and performs status consistency analysis and metering correction on the abnormal segment based on historical metering trends, updating the status fields of the slice matrix. The dormant verification module triggers the dormant account verification process if the corrected slice still has status anomalies, calls the historical slice data of the dormant node for trend correlation analysis, and incorporates the effective heat change sequence into the final judgment and trend correction of the current node.

[0015] Thirdly, the present invention provides a computer device including a memory and a processor, wherein the memory stores a computer program, wherein the computer program instructions, when executed by the processor, implement the steps of the prepaid heating network metering management method as described in the first aspect of the present invention.

[0016] Fourthly, the present invention provides a computer-readable storage medium having a computer program stored thereon, wherein: when the computer program instructions are executed by a processor, they implement the steps of the prepaid heating network metering management method as described in the first aspect of the present invention.

[0017] The beneficial effects of this invention are as follows: By constructing a multi-dimensional slice matrix and combining spatial topological relationships and historical trend characteristics, this invention significantly improves the accuracy and efficiency of identifying and locating metering anomalies, effectively avoiding the accumulation of metering errors caused by single-point failures or data distortion. Furthermore, by introducing a state consistency comparison mechanism, this invention ensures that the correction process not only relies on numerical judgment but also integrates spatial flow logic and trend rationality, guaranteeing high reliability of the correction results. When facing complex segments, this invention further utilizes historical data from dormant accounts to achieve trend reconstruction, overcoming the limitation of traditional methods due to a lack of comparative samples. In summary, this invention significantly enhances the stability and intelligence level of heating network metering management, improving metering fairness, user satisfaction, and operational efficiency under the prepaid mechanism. Attached Figure Description

[0018] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0019] Figure 1 A flowchart for the prepaid heating network metering management method; Figure 2 This is a structural diagram of the prepaid heating network metering management system. Detailed Implementation

[0020] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

[0021] Many specific details are set forth in the following description in order to provide a full understanding of the invention. However, the invention may also be practiced in other ways different from those described herein, and those skilled in the art can make similar extensions without departing from the spirit of the invention. Therefore, the invention is not limited to the specific embodiments disclosed below.

[0022] Secondly, the term "one embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The phrase "in one embodiment" appearing in different places in this specification does not necessarily refer to the same embodiment, nor is it a single or selective embodiment that is mutually exclusive with other embodiments.

[0023] Figure 1 This is a flowchart of a prepaid heating network metering management method according to an embodiment of the present invention. Figure 1 As shown, the prepaid heating network metering management method includes: S1: Divide the collected heating network metering data into multiple dimensions to generate a slice matrix containing status fields.

[0024] In this embodiment of the invention, generating a slice matrix containing a state field includes the following steps: First, the continuously collected raw metering data of the heating network is sliced ​​according to a unified time period to generate each... A collection of heat fragments for a single slice at any given time. Each set of heat fragments It includes the source timestamp, the data acquisition node number, and the device type identifier.

[0025] Secondly, to enhance the operability of the sliced ​​data in terms of spatial distribution, each heat fragment is set up. Reclassify by spatial dimension, and construct a spatial index set based on building unit number and adjacent node information; each record in the spatial index set retains The original equipment number and the associated unit number.

[0026] Secondly, based on the spatial dimension partitioning, in order to improve the ability of the slice matrix to distinguish device behavior characteristics, for each device number in the spatial index set, a subset is divided according to the device function type (valve control device, flow meter, etc.). and for each subset Establish equipment type labels and collect quality level parameters, supplementing the original data. Field structure. Among them, the collection quality level is calculated based on a comprehensive weighted average of dimensions such as data upload frequency, anomaly rate, and packet loss rate, and is divided into three levels: high, medium, and low.

[0027] Finally, the data will be segmented according to the dimensions of time, space, and device type. The slice set is uniformly structured and encapsulated into a multi-dimensional slice matrix, containing time periods. The system includes a node index, device identifier, device type label, quality level, heat value, prepaid account status field, and status code. The prepaid account status field specifically includes: real-time recorded user account balance level, current node's heat network control strategy, and user-defined or system-set balance warning line.

[0028] Specifically, in this embodiment of the invention, the status code, as the most critical discriminative field in the sliced ​​data, has the following value setting logic: If the data is complete, of high quality, and has no historical anomalies, the status code is set to 0, representing a normal state. If there are minor missing data or fluctuations in upload frequency, the status code is set to 1. If the heat value deviates significantly from the historical trend or the upload failure rate is high, the status code is set to 2. In addition, when a node is in a dormant state but is actively woken up for trend reference, the status code is set to 3.

[0029] This status field not only provides a health indicator for the data within the slice matrix, but also serves as an entry condition field for cross-validation of time and space segments in the subsequent step S2, possessing high discriminability and process-driven capabilities.

[0030] S2: Extract different spatial and equipment combinations under the same time dimension from the slice matrix, perform heat difference cross-analysis, and filter out abnormal segments.

[0031] S2.1: For all segments in the slice matrix with a status code of 0, sort them according to a uniform time period. Extract subsets of slices that constitute the same period and filter out comparison groups with the same equipment type.

[0032] In this context, segments with a status code of 0 represent normal measurements in the original data that were not marked as abnormal. These segments are extracted according to a uniform time period to form a subset of slices with the same period. This involves aggregating and processing multiple data points with consistent time fields to ensure that the difference analysis operation is performed under a uniform time reference plane.

[0033] Within the aforementioned slice subsets of the same period, comparable equipment combinations are selected based on equipment type tags to ensure consistency in functional attributes and technical parameters between the compared objects. For example, only horizontal heat difference comparisons are performed between valve control devices and between flow meters to avoid invalid biases caused by differences in equipment functions.

[0034] During the selection process of equipment combination groups, the quality level parameter is also used as a sorting factor, with high-level data as the benchmark source and low-level data as the reference item, thereby constructing a difference channel with a reliable reference direction.

[0035] In addition, the formation of the combination group also takes into account the adjacency relationship in the node topology, and prioritizes the combination of equipment pairs located in the same or adjacent spatial units to ensure spatial comparability and logical continuity of the path.

[0036] S2.2: In the comparison combination group, a heat reference path set is constructed according to the node index and quality level, and each path represents the trend of calorific value difference under unit spatial offset.

[0037] In this embodiment of the invention, each heat reference path is defined as a trend function of the change between heat values ​​recorded by different devices under a unit spatial offset (such as floor, adjacent room, thermal zone). The trend is represented in vector form and records key features such as positive and negative offset of heat value, unit rate of change, and difference fluctuation period.

[0038] During the path construction process, spatial order modeling is performed based on node index information, source-target device pairs are established according to device number, and quality level is used as a path weight factor to form a structured reference path map.

[0039] S2.3: Evaluate the heat difference for each path in the heat reference path set, calculate the deviation index, and generate a candidate abnormal segment index set based on the deviation magnitude and the frequency of consecutive occurrences. This includes the following steps: S2.3.1: For any path in the heat reference path set, extract the heat acquisition data within the current time window of the path in the continuously sampled slice sequence, and establish a snapshot of the interference factors corresponding to the time point based on the equipment information and load level of the spatial node, including but not limited to: whether there is a sudden increase in equipment load at the spatial node corresponding to the path, whether there is a switching behavior of the acquisition equipment source, and whether the path contains level classification boundary nodes.

[0040] S2.3.2: After completing the snapshot analysis of interference factors, and combining the heat change trend within the time period, establish cross-judgment rules, including: If a path exhibits a continuous offset across multiple adjacent sampling points, and this offset is rare in historical periods (e.g., less than 5% probability of occurrence in similar paths), it is marked as a trend-type deviation segment. If the difference in heat data before and after the current time period for a certain path increases significantly (e.g., more than twice the average deviation of similar paths in history), and there is no equipment replacement record for that node, it is marked as a sudden deviation segment. If a path has a record of equipment level change or adjacent path switching before the start of an abnormal segment, and the change matches the heat increase trend, it is marked as a level disturbance type segment. If the heat value of a certain path is consistently lower than expected and the associated account balance is in an overdraft state, it will be marked as a prepaid restriction type anomaly. This type of anomaly needs to be distinguished from equipment failure or spatial disturbance anomalies to avoid incorrect correction.

[0041] Number the path segments that satisfy any of the above rules and generate a path-level anomaly index set; the data given above is sample data, and the actual operation needs to be set according to the requirements. S2.3.3: Map the numbers of all abnormal segments screened from the reference paths to the row and column indices in the slice matrix to form a preliminary candidate abnormal index set.

[0042] The index segments are then merged based on the following two-level decision logic: If the horizontal (at the same time) or vertical (at the same path) distance between any two abnormal index segments in the slice matrix is ​​less than or equal to 1 unit, and the abnormal source markers are consistent, then they are considered to be the same segment and merged. If the merged abnormal segments span multiple sampling periods in time (e.g., ≥3), the start and end boundaries are recorded and marked as structural abnormal segments.

[0043] The final output is a unified set of candidate anomaly fragment indexes.

[0044] It can be seen that by constructing an aggregation mechanism based on proximity distance and anomaly factor consistency, this invention improves the structure and expressive power of anomaly detection, and can avoid state fragmentation caused by discrete judgment.

[0045] S2.4: Write the candidate anomaly fragment index set back into the slice matrix, update the corresponding status code to 1, and mark the source of the anomaly factor (spatial / device difference, level disturbance, continuity).

[0046] For the output set of candidate anomaly fragment indices, locate the corresponding index position in the slice matrix and perform the following status field write-back operation: The status field corresponding to the index position is uniformly updated to the value "1", indicating that the segment is in an abnormal state. The status code update does not depend on manual setting, but is automatically generated by the previous abnormal rules, and is only valid for the current period to avoid interference from historical states.

[0047] For each slice index with a status code marked as "1", its source factor tag is extracted from the original anomaly screening and written into the anomaly factor field. The tags are represented in a structured label manner. The tag sources include, but are not limited to: spatial disturbances (such as path changes, adjacent node interventions); device behavior (such as sudden changes in acquisition devices, interface switching); hierarchical structure (such as cross-deployment of device types, load level reconstruction); and continuous behavior (such as multi-period trend shifts, unidirectional fluctuation accumulation).

[0048] If the same segment satisfies multiple factors simultaneously, it is recorded in a primary cause + secondary cause structure, and the priority of the primary cause source is automatically determined according to the order of deviation triggering.

[0049] S3: Using the spatial node where the abnormal segment is located as the center, construct its adjacent node set, and based on its historical measurement trend, perform state consistency analysis and measurement correction on the abnormal segment, and update the state field of the original slice.

[0050] Since the data of the heating network is affected not only by the trend of time change, but also by the coupling relationship of spatial structure, after completing the initial judgment and correction of anomalies based on the time dimension, a structural grid model is further introduced to model the spatial structure of the heating network system. The structural grid model uses thermal units as the basic topological units to construct the adjacency relationship between nodes to support spatial adjacency inference and trend comparison.

[0051] In this embodiment of the invention, performing state consistency analysis and measurement correction on abnormal segments includes the following steps: First, for each segment in the candidate abnormal segment index set, the corresponding spatial node is extracted, and a set of adjacent nodes is constructed based on the spatial topology.

[0052] In this embodiment of the invention, when processing the candidate abnormal fragment index set, the position coordinate information of each fragment in the spatial distribution is first indexed according to the slice number corresponding to each fragment, and the mapping node in the structured mesh model is extracted accordingly. The mapping node is defined as the master node of the abnormal fragment.

[0053] For each master node, based on a pre-constructed multi-dimensional spatial topology network model, the set of adjacent nodes is constructed according to the following three types of constraints: Prioritize nodes that have a direct geometric boundary connection with the master node, such as those with consecutive slice numbers or shared boundaries in the X, Y, or Z directions; For nodes within the same thermal unit, even if they are spatially distant, they should be included in the set of adjacent nodes due to their functional coupling characteristics. Based on the thermal coupling map established by the heat reference path in the historical period, nodes with stable thermal response linkage relationship are weighted and judged. If the heat conduction correlation coefficient is higher than the set threshold, it is also regarded as an adjacent node.

[0054] The final set of adjacent nodes is the effective heat reference set of the master node under the three-dimensional cross-filtering.

[0055] Secondly, for all slices within the adjacent node set, historical measurement trends are extracted according to the time dimension to construct a cluster of heat change curves with the same period, and nodes sensitive to fluctuations are marked.

[0056] For each slice in the set of adjacent nodes determined in the previous step, extract the heat change curves within multiple standard periods (such as 24-hour, 48-hour, or 7-day rolling windows) based on the heat measurement data in the time dimension, and perform the following processing steps: Periodic standardization is performed on all time series data to unify the time base to the same starting point and eliminate sensor sampling bias. The second-order difference value of each heat change curve is calculated using a sliding window to obtain the transient fluctuation intensity of heat change; at the same time, the periodic fluctuation factor is extracted by combining Fourier transform to achieve decoupling of periodic and non-periodic components. Based on fluctuation similarity (such as Dynamic Time Warping (DTW) distance), K-means clustering or density clustering is performed on all heat change curves to divide them into several trend clusters, each cluster representing a dominant heat evolution pattern. Nodes belonging to high volatility clusters (e.g., those with a mean standard deviation exceeding 1.5 times the mean of neighboring nodes, this is just an example) are identified as volatility-sensitive nodes because they have a stronger responsiveness and are more likely to reveal the spread of potential abnormal states.

[0057] Finally, the heat data of the slice corresponding to the abnormal segment is compared with the heat change curve cluster for state consistency. If the heat deviation range exceeds the dynamic tolerance, the heat data of the segment is corrected according to the median trend.

[0058] In this embodiment of the invention, for each candidate abnormal segment, a state consistency comparison is performed between the heat measurement data sequence in the time dimension and the heat change curve cluster to which the corresponding adjacent node belongs. This comparison process includes: First, calculate the trend deviation between the heat curve of the abnormal segment and the central curve of each curve cluster (e.g., by using the Pearson correlation coefficient r or slope matching index). If the maximum correlation is lower than the preset lower limit (e.g., r < 0.4), it is judged as trend heterogeneity. The dynamic tolerance adopts an adaptive strategy based on the data distribution type: if the fluctuation sequence meets the condition of an approximate normal distribution, the mean of the fluctuation range of the corresponding curve cluster at the same time point ± 1.25 times the standard deviation is taken as the judgment threshold; otherwise, based on the interquartile range, the tolerance range is extended downward and upward by 1.5 times the interquartile range to make a consistency judgment, so as to avoid misjudgment due to the shift in distribution characteristics. If the heat value of an abnormal segment exceeds the tolerance range for more than a threshold time window at any given time, it is marked as a heat deviation. For the aforementioned heat deviation segments, the central trend line of the curve cluster closest to its trend is retrieved, and the median trend curve of the corresponding time period is used as a reference for performing linear stretching or compression correction; The correction should maintain the original curve's direction of fluctuation, only adjusting the magnitude of the deviation to ensure data continuity; If, after performing the correction operation, the similarity between the segment and the corrected trend curve is still below a certain threshold (e.g., r < 0.3), it is recorded as a low-confidence correction, and it is recommended to retain the original value for manual review to avoid incorrect correction.

[0059] Through the above consistency comparison and correction process, dynamic judgment and target-oriented correction of abnormal segment heat data can be achieved, thereby improving the heat rationality of the overall visual inspection data and the fault tolerance of local anomalies.

[0060] In addition, if the account balance associated with the abnormal segment is in an overdraft state, the prepaid strategy correction will be implemented first (such as manual verification of payment required for power supply restoration) rather than direct heat value correction. If the account balance is restored, heat value backfilling correction will be automatically triggered (such as compensating for heat loss during the flow restriction period).

[0061] Furthermore, in this embodiment of the invention, updating the state field of the original slice includes: For corrected abnormal segments, the corresponding status code is synchronously updated to 2 in the slice matrix, and the associated trend cluster identifier is recorded to support subsequent tracing and abnormal propagation control.

[0062] Specifically, after the correction operation is completed, the status code field corresponding to the corrected abnormal segment in the slice matrix is ​​updated to 2, indicating that the segment has completed the status consistency check and data correction. At the same time, a related trend cluster identifier field is added to the slice matrix to record the ID of the heat change curve cluster referenced in this correction, which facilitates subsequent anomaly path tracking and effectiveness evaluation of the correction mechanism.

[0063] It should be noted that after the heat data correction is completed, a correction status flag field is attached to the corrected data and written to the original data channel through an asynchronous caching mechanism. This allows subsequent analysis to select whether to use the corrected data based on the validity flag. If the current operating environment is in offline batch processing mode, batch data application is performed during low-load periods each day to avoid calculation conflicts during peak hours, while ensuring data consistency and correction accuracy.

[0064] S4: If the corrected slice still has an abnormal state, the dormant account verification process is triggered, the historical slice data of the dormant node is called to perform trend correlation analysis, and the effective heat change sequence is included in the final judgment and trend correction of the current node.

[0065] S4.1: For the fragment with status code 2 after correction, call the user account list of the corresponding heating network node of the current building, and filter the set of nodes with status field of dormancy.

[0066] Specifically, for candidate segments that have completed heat data correction and whose status field is marked as 2 in the S3 stage, the user account index table configured in the corresponding heating network model of the building to which the segment belongs will be called according to the building identifier.

[0067] The index table records the account status, historical behavior cycle, and average daily heat consumption characteristics of each heat network node and its bound user, forming a user-side heat behavior database.

[0068] In this account list, nodes with a status field indicating a dormant state are extracted based on the following filtering criteria: no stable heat consumption behavior within a specified period, and no manual adjustment records or abnormal flow; daily average heat consumption is less than 10% of the building's historical average, excluding extreme abnormal fluctuations; the control status field fed back by the smart terminal shows that the heating mode is not turned on, verifying that the dormant state is a system-level behavior; if a user's account is in an overdraft state for N consecutive days and has not paid the outstanding fees, it is forcibly marked as a prepaid dormant node. Prepaid dormant nodes do not participate in heat trend reference (to avoid unpaid nodes interfering with normal data); the node can only be activated and power supply restored after manual confirmation of payment.

[0069] The final extracted set of dormant nodes constitutes the non-behavioral disturbance thermal response background node group for the current abnormal segment reference, which is used to determine whether the thermal change of the current segment may be caused by system mislabeling.

[0070] S4.2: Extract the time dimension of the current abnormal segment from the node set. The corresponding historical slice data is used to construct a comparison matrix with the current node's heat trend.

[0071] Specifically, for the set of dormant nodes that have been extracted above, heat slice data that completely corresponds to the current abnormal segment time point is extracted from historical data.

[0072] To ensure the accuracy of time alignment, the following steps are required: First, historical slice data corresponding to the current abnormal segment in the time dimension is extracted from the node set, and window synchronization is performed based on the unified time slice alignment rules defined in step S1. The time slice alignment rules include sliding window boundary alignment and step size consistency requirements to ensure that the historical data window and the abnormal segment maintain an equal matching relationship in the time domain. Subsequently, the heat change curve of the current segment is matched one-to-one with the heat change sequence in the set of dormant nodes to form a two-dimensional heat trend comparison matrix. Each row in the comparison matrix represents the heat trend of a dormant node, and each column represents a time sampling point.

[0073] S4.3: Perform trend normalization on the heat change sequence in the comparison matrix. If there is a change pattern sequence that matches the current node, it is incorporated into the original heat change trend for reconstruction and deduction, and the trend median reference is updated.

[0074] After constructing the comparison matrix, to eliminate the interference of absolute numerical differences on trend similarity analysis, trend normalization processing needs to be performed on each heat series, including the following key steps: Using the local maxima and minima of each sequence as benchmarks, the range of heat variation is normalized to ensure balanced fluctuations. The normalized sequence is converted into a first-order difference sequence to extract the direction and slope of the trend. The heat sequence of the current node is matched with each difference sequence in the dormant node using DTW dynamic time warping, and a matching tolerance threshold is set (e.g., error integral < 20%) to screen out candidate sequences with matching trend patterns.

[0075] If the trend sequences of one or more dormant nodes are highly consistent with the heat sequence change pattern of the current node (e.g., the matching score exceeds 0.8), these trend sequences are included in the trend reference set, and the current node's heat trend is reconstructed and deduced based on the aggregated feature curve (e.g., weighted median).

[0076] During the simulation, the high-deviation segments in the current node trend segment are replaced with the median trend of the trend reference set to achieve a more stable and reasonable trend median update, providing a target trend for subsequent secondary corrections.

[0077] S4.4: Based on the reconstructed heat trend, the heat data of the target slice is corrected a second time, and the status code is updated to 3 at the same time, indicating that the final state after the dormant trend reference judgment has been completed.

[0078] In this embodiment of the invention, based on the reconstructed heat trend curve generated in step S4.3, a deviation correction operation is performed again on the heat data of the current target slice, as follows: Calculate the point-to-point residual between the current heat data and the reconstructed trend. If the proportion of samples with residuals outside the tolerance band exceeds D, for example, D=25%, then a correction is triggered. Local regression is used to fit the target segment of heat data, and the trend reconstruction result is used as the regression center point to achieve curve approximation and correction of the data; Ensure that the corrected curve maintains a continuous slope with the original curve at the start and end points to avoid physical inconsistencies caused by sudden trend changes. After the second correction is completed, the status field of the target segment is updated from "2" to "3", indicating that the segment has completed the thermal rationality determination of the final state based on the dormancy reference trend. In addition, the correction trajectory and correction confidence fields are written simultaneously during the status update.

[0079] Furthermore, this embodiment also provides a prepaid heating network metering management system, including, The slice generation module 100 divides the collected heating network metering data into multiple dimensions and generates a slice matrix containing status fields. The difference analysis module 200 extracts different spatial and equipment combinations under the same time dimension from the slice matrix, performs cross-analysis of heat difference, and filters out abnormal segments. The state correction module 300 constructs a set of adjacent nodes centered on the spatial node where the abnormal segment is located, and performs state consistency analysis and measurement correction on the abnormal segment based on historical measurement trends, and updates the state field of the slice matrix. If the corrected slice still has an abnormal state, the hibernation verification module 400 will trigger the hibernation account verification process, call the historical slice data of the hibernation node to perform trend correlation analysis, and incorporate the effective heat change sequence into the final judgment and trend correction of the current node.

[0080] This embodiment also provides a computer device applicable to the prepaid heating network metering management method, including a memory and a processor; the memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions to implement the prepaid heating network metering management method proposed in the above embodiment.

[0081] The computer device can be a terminal, comprising a processor, memory, communication interface, display screen, and input devices connected via a system bus. The processor provides computing and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs stored in the non-volatile storage media. The communication interface is used for wired or wireless communication with external terminals; wireless communication can be achieved through Wi-Fi, carrier networks, NFC (Near Field Communication), or other technologies. The display screen can be an LCD screen or an e-ink screen. The input devices can be a touch layer covering the display screen, buttons, a trackball, or a touchpad on the computer device's casing, or an external keyboard, touchpad, or mouse.

[0082] This embodiment also provides a storage medium storing a computer program that, when executed by a processor, implements the prepaid heating network metering management method proposed in the above embodiments.

[0083] In summary, by constructing a multi-dimensional slice matrix and combining spatial topological relationships and historical trend characteristics, this invention significantly improves the accuracy and efficiency of identifying and locating metering anomalies, effectively avoiding the accumulation of metering errors caused by single-point failures or data distortion. Furthermore, by introducing a state consistency comparison mechanism, this invention ensures that the correction process not only relies on numerical judgment but also integrates spatial flow logic and trend rationality, guaranteeing high reliability of the correction results. When facing complex segments, this invention further utilizes historical data from dormant accounts to achieve trend reconstruction, overcoming the limitation of traditional methods due to a lack of comparative samples. In conclusion, this invention significantly enhances the stability and intelligence of heating network metering management, improving metering fairness, user satisfaction, and operational efficiency under the prepaid mechanism.

[0084] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A method for metering and managing prepaid heating networks, characterized in that: include: The collected heating network metering data is divided into multiple dimensions to generate a slice matrix containing status fields; Different spatial and equipment combinations under the same time dimension are extracted from the slice matrix, and cross-analysis of heat difference is performed to screen out abnormal segments; Centered on the spatial node where the abnormal segment is located, a set of adjacent nodes is constructed, and based on historical measurement trends, state consistency analysis and measurement correction are performed on the abnormal segment, and the state field of the slice matrix is ​​updated. If the corrected slice still has an abnormal state, the dormant account verification process is triggered, historical slice data of the dormant node is called to perform trend correlation analysis, and the effective heat change sequence is included in the final judgment and trend correction of the current node.

2. The prepaid heating network metering management method as described in claim 1, characterized in that: The slice matrix includes time period, node index, device identifier, device type label, quality level, calorie value, prepaid account status field, and status code.

3. The prepaid heating network metering management method as described in claim 1, characterized in that: The heat difference cross-analysis includes: For all segments with a status code of 0 in the slice matrix, extract them according to the same time period to form a slice subset of the same period, and filter out the comparison combination group with the same equipment type. In the comparison combination group, a set of heat reference paths is constructed according to node index and quality level, where each path represents the trend of calorific value difference under unit spatial offset; The heat difference of each path in the heat reference path set is evaluated, and the deviation index is calculated.

4. The prepaid heating network metering management method as described in claim 3, characterized in that: The screening of abnormal segments includes: generating a candidate abnormal segment index set based on two indicators: deviation magnitude and frequency of consecutive occurrence; writing the candidate abnormal segment index set back into the slice matrix, updating the corresponding status code to 1, and marking the source of the abnormal factor.

5. The prepaid heating network metering management method as described in claim 1, characterized in that: The state consistency analysis and measurement correction for abnormal segments include: For each segment in the candidate abnormal segment index set, extract the corresponding spatial node and construct an adjacent node set based on the spatial topology. For all slices within the adjacent node set, extract historical measurement trends along the time dimension, construct a cluster of heat change curves for the same period, and mark nodes sensitive to fluctuations. The heat data of the slice corresponding to the abnormal segment is compared with the heat change curve cluster. If the heat deviation range exceeds the dynamic tolerance, the heat data of the segment is corrected according to the median trend.

6. The prepaid heating network metering management method as described in claim 5, characterized in that: The status fields of the updated slice matrix include: for corrected abnormal segments, the corresponding status code is synchronously updated to 2 in the slice matrix, and the associated trend cluster identifier is recorded.

7. The prepaid heating network metering management method as described in claim 1, characterized in that: The trend correlation analysis includes: For the segment with a status code of 2 after correction, retrieve the list of user accounts of the corresponding heating network node for the current building, and filter out the set of nodes with a status field of dormancy. Extract historical slice data corresponding to the current abnormal segment in the time dimension from the node set, and construct a comparison matrix with the heat trend of the current node; The heat change sequence in the comparison matrix is ​​normalized for trend. If there is a change pattern sequence that matches the current node, it is incorporated into the original heat change trend for reconstruction and deduction.

8. A prepaid heating network metering management system, based on the prepaid heating network metering management method according to any one of claims 1 to 7, characterized in that: Also includes: The slice generation module divides the collected heating network metering data into multiple dimensions and generates a slice matrix containing status fields. The difference analysis module extracts different spatial and equipment combinations under the same time dimension from the slice matrix, performs cross-analysis of heat difference, and filters out abnormal segments; The state correction module constructs a set of adjacent nodes centered on the spatial node where the abnormal segment is located, and performs state consistency analysis and measurement correction on the abnormal segment based on historical measurement trends, updating the state field of the slice matrix. If the corrected slice still has an abnormal status, the dormant account verification module will trigger the dormant account verification process, call the historical slice data of the dormant node for trend correlation analysis, and incorporate the effective heat change sequence into the final judgment and trend correction of the current node.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that: When the processor executes the computer program, it implements the steps of the prepaid heating network metering management method according to any one of claims 1 to 7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by the processor, it implements the steps of the prepaid heating network metering management method according to any one of claims 1 to 7.

Images