A plan water early warning management method based on data fusion

Abstract

The application discloses a kind of plan water early warning management methods based on data fusion, it is related to water early warning technical field, including the water consumption data and available water source data of each water supply subarea are collected, and are decomposed according to time window, form the water consumption feature set of water supply subarea period, obtain preliminary space-time water consumption characteristics;Based on preliminary space-time water consumption characteristics, the complementarity of adjacent water supply subarea in alternate period is calculated, and the alternative ability set is generated in combination with available water source data;The short-term water gap of each water supply subarea is predicted, and candidate allocation scheme is generated according to complementary relationship, and the output is the allocation instruction set between water supply subarea;When short-term water gap prediction exceeds gap threshold, start alternate buffer window, and the non-key water demand of part of water supply subarea is period translation or cross-zone allocation, and output warning level and corresponding allocation instruction.The application realizes the whole-link closed-loop management from data acquisition, gap prediction to cross-zone allocation.

Description

Technical Field

[0001] This invention relates to the field of water use early warning technology, and in particular to a planned water use early warning management method based on data fusion. Background Technology

[0002] With the development of smart water management and urban water resource scheduling technologies, more and more research is focusing on how to achieve refined scheduling and risk early warning in water supply zone management. Traditional water supply management methods mainly rely on fixed threshold monitoring and experience-based scheduling. While these methods can ensure regional water security to a certain extent, they often suffer from untimely response and unreasonable allocation when facing complex water supply patterns involving multiple zones and water sources. In recent years, the introduction of big data, the Internet of Things, and artificial intelligence has provided new means for water supply data collection, temporal feature extraction, and early warning modeling. However, existing methods mostly focus on leakage monitoring or user water use behavior identification, lacking allocation and early warning mechanisms based on cross-zone spatiotemporal data fusion. This results in shortcomings in short-term gap prediction and cross-zone scheduling optimization.

[0003] CN113657780A discloses a unified assessment and warning system and method for multi-dimensional comprehensive leakage index. Through regional division, flow data collection, index assessment, and early warning display, it achieves comprehensive analysis of leakage in water supply networks. While this method plays a positive role in leakage identification and visualization, it primarily focuses on single-index assessment of network leakage problems and does not involve the fusion and complementary calculation of spatiotemporal features between multi-regional water supply data. Furthermore, the early warning information generated by this scheme emphasizes static state monitoring and lacks dynamic allocation and short-term gap prediction mechanisms, thus making it difficult to support the cross-regional coordinated scheduling needs in complex water supply environments.

[0004] CN119671028A discloses a data-driven method, apparatus, equipment, and storage medium for determining water usage scenarios. It collects user-side water usage data in real time through smart water meters and the Internet of Things (IoT), and combines this with a multimodal model to achieve scenario recognition and personalized services. While this approach demonstrates some innovation in user behavior modeling and scenario-based analysis, its core objectives focus on user water usage habits, abnormal water usage identification, and service optimization. It does not model the overall shortage prediction and inter-regional water allocation for multi-zone water supply systems, nor does it address water source substitutability and buffer regulation. Therefore, although this method improves user-side service capabilities, it cannot effectively solve the problems of regional water shortages and inter-regional water balance.

[0005] In summary, while existing technologies have made some progress in leakage monitoring and user water usage behavior identification, they remain insufficient in addressing short-term shortage prediction, inter-regional allocation, and early warning management in multi-regional water supply environments. This results in delayed responses and inflexible scheduling in the event of sudden water shortages or uneven water supply. This invention establishes a planned water usage early warning and management method for multi-regional water supply systems, achieving closed-loop management across the entire chain from data acquisition and shortage prediction to inter-regional allocation. Summary of the Invention

[0006] The purpose of this section is to outline some aspects of the embodiments of the present invention and to briefly introduce some preferred embodiments. Some simplifications or omissions may be made in this section, as well as in the abstract and title of the present application, to avoid obscuring the purpose of this section, the abstract and title of the invention. Such simplifications or omissions shall not be used to limit the scope of the present invention.

[0007] In view of the aforementioned existing problems, the present invention is proposed.

[0008] Therefore, the technical problem solved by this invention is: how to address the lack of cross-regional spatiotemporal data fusion and dynamic allocation in existing technologies.

[0009] To address the aforementioned technical problems, this invention provides the following technical solution: Water consumption data and available water source data for each water supply zone are collected and decomposed according to time windows to form a set of water consumption characteristics for each water supply zone during different time periods, thus obtaining preliminary spatiotemporal water consumption characteristics; based on these preliminary spatiotemporal water consumption characteristics, the complementarity of adjacent water supply zones during alternating time periods is calculated, and a set of alternative capacity is generated by combining available water source data; short-term water shortages for each water supply zone are predicted, and candidate allocation schemes are generated based on the complementarity relationship, outputting a set of allocation instructions for the water supply zones; when the predicted short-term water shortage exceeds the shortage threshold, an alternating buffer window is activated, and non-critical water demand from some water supply zones is shifted or allocated across time periods, with an early warning level and corresponding allocation instructions output.

[0010] As a preferred embodiment of the present invention, the formation of the water use feature set includes: collecting water consumption data and available water source data of each water supply zone, and attaching a uniform time stamp to each collected record to form an original data sequence; performing statistical analysis on the original data sequence to identify outliers caused by sudden events or sensor errors, and generating a corrected sequence through interpolation; dividing the data according to a set time window based on the corrected sequence, and aggregating the water consumption data and available water source data within the same water supply zone to generate a water use feature set; the preliminary spatiotemporal water use characteristics include: extracting the mean, volatility, and time period differences of each time window based on the water use feature set to form a regional time period water use feature set, and outputting it as preliminary spatiotemporal water use characteristics.

[0011] As a preferred embodiment of the present invention, the calculation of the complementarity of adjacent water supply zones during alternating periods includes: aligning the preliminary spatiotemporal water usage characteristics according to the same time window, extracting the peak time and peak amplitude of each water supply zone within each time window to form a peak sequence of the zone period; based on the peak sequence of the zone period, calculating the phase difference and peak amplitude ratio of the peak times of adjacent water supply zones within a short time window to obtain an alternation measurement index; comparing the alternation measurement index with a set index threshold range, and determining that the corresponding time window is an alternating period when the phase difference falls within the preset index range and the peak amplitude ratio is greater than the amplitude threshold, otherwise it is a non-alternating period; and outputting an alternating period identifier sequence.

[0012] In a preferred embodiment of the present invention, the step of generating an alternative capacity set by combining available water source data includes: generating a complementary set according to water supply zones based on the determined alternation period; the complementary set is further divided into complementary strength levels by extracting the alternation period identifier and combining it with the alternation measurement index of each water supply zone; and combining the complementary set with the input available water source data to calculate the available amount for each water supply zone within each alternation period. Difference in gaps with adjacent water supply areas Define substitution capabilities The minimum of the two values ​​is used when there are multiple candidate zones in adjacent water supply zones. The substitution rates are sorted to obtain priorities and recorded as substitution entries for water supply zone pairs. The substitution capacity entries are then summarized in chronological order and by water supply zone pairs to form a substitution capacity set, which is output as a zone substitution capacity sequence.

[0013] As a preferred embodiment of the present invention, the prediction of short-term water shortages for each water supply zone includes: calculating a historical average shortage sequence using a sliding window method within the prediction period for each water supply zone; comparing the historical average shortage sequence as a baseline shortage trend sequence with the actual historical shortage to obtain an error sequence; performing pattern recognition on the error sequence within the sliding window to distinguish between systematic and random errors; wherein, systematic and random errors are corrected separately to form a layered corrected error sequence; and weightedly fusing the baseline shortage trend sequence and the corrected error sequence to generate a residual-corrected short-term shortage prediction sequence, thus obtaining the residual-corrected shortage prediction sequence for each water supply zone during the prediction period.

[0014] In a preferred embodiment of the present invention, the generation of the correction error sequence includes: extracting the mean drift, volatility ratio, and direction of the error sequence within a sliding window; marking the sliding window as a systematic error segment when the mean drift remains unidirectionally increasing or decreasing within a continuous sliding window and the direction factor is stable and consistent; marking the sliding window as a random error segment when the volatility ratio is greater than 1 within a continuous sliding window and the direction changes P times, where P is a constant; and obtaining a long-term correction trend curve by using a weighted average or exponential smoothing on the sliding window sequence marked as a systematic error segment. For sliding window sequences marked as random error segments: based on the ratio of short-term volatility... Calculate the residual amplification factor The correction value is This is used to adjust local short-term corrections; correction values ​​from different sources are concatenated in time sequence to form a hierarchically processed correction error sequence; the short-term fluctuation amplitude is compared to... The calculation is performed by: selecting water consumption or gap data for K consecutive time periods, determining the difference between the maximum and minimum values ​​of water consumption or gap data within a certain time period as the short-term fluctuation range; then using the average or median of the water consumption or gap data within the K consecutive time periods as the benchmark level, and using the ratio of the fluctuation range to the benchmark level as the short-term fluctuation range ratio; where K is a constant.

[0015] As a preferred embodiment of the present invention, the step of generating candidate allocation schemes based on complementarity relationships includes: matching the generated set of alternative capabilities with the predicted residual correction gap values ​​of each water supply zone; if a water supply zone has a positive gap during the prediction period, then finding zones with remaining water volume in the same period from the complementarity set to form a gap-replenishment preliminary pair; the gap-replenishment preliminary pair retains only a one-to-one mapping relationship, and a gap zone corresponds only to the replenishment zone with the highest alternative capability; for each gap-replenishment preliminary pair, the matching degree index is calculated:

[0016]

[0017] in, Indicates water supply zones The gap amount, Indicates water supply zones The available allocation margin is determined; all gap-replenishment pairs are initially sorted according to the matching degree index, and the top M pairs are retained to form a candidate scheme set; where M is a constant; the candidate scheme set is converted into an executable allocation instruction, the allocation instruction content includes: the starting zone, target zone, corresponding time period, and planned allocation volume; for cases involving multiple replenishment zones, allocation is gradually carried out according to the matching degree priority until the gap is satisfied; and a set of allocation instructions by zone is output.

[0018] As a preferred embodiment of the present invention, the step of activating the alternating buffer window to shift or allocate non-critical water demand in some water supply zones across time periods includes: when the shortfall in a certain water supply zone exceeds the shortfall threshold during the predicted time period, it is determined that a buffer mechanism needs to be activated, and a buffer zone needs to be established; an alternating buffer window is established for the buffer zone, the length of which is equal to H times the predicted shortfall time period; the design logic of the alternating buffer window includes: during the shortfall prediction period, non-critical water demand is allowed to be postponed to the latter half of the alternating buffer window or transferred to an adjacent water supply zone; within the alternating buffer window, non-critical demand in the water supply zone is first identified and its proportion is calculated; by sorting water demand by priority, it is shifted step by step to the latter half of the alternating buffer window, and a time period adjustment table is formed after the shift, recording the original time period and the adjusted time period of demand for each water supply zone; if the shortfall in the adjusted time period still exceeds the shortfall threshold, the allocation margin is searched in complementary zones to form cross-regional supplementation; wherein, the allocation rule is based on the matching degree calculation, but the matching degree priority is limited to the part corresponding to non-critical demand; and a cross-regional allocation instruction is output.

[0019] The beneficial effects of this invention are as follows: This invention can accurately extract time-specific water use characteristics in a multi-zone environment, identify peak differences and complementary relationships between zones, and combine available water source data to quantify substitution capabilities, thereby forming candidate allocation schemes before the predicted gap occurs; and by introducing alternating buffer windows, it can not only shift non-critical needs across time periods and coordinate across zones, but also achieve flexible adjustment when the gap exceeds the threshold, avoiding water supply risks caused by sudden fluctuations in water use in a single zone.

[0020] Overall, this invention effectively improves the accuracy of short-term water shortage prediction and the rationality of cross-regional scheduling, enabling the water supply system to have stronger risk resistance and adaptability. It has significant application value in ensuring water supply security, improving water resource utilization efficiency, and realizing smart water management. Attached Figure Description

[0021] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the description of the embodiments will be briefly introduced below. Obviously, the 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. Wherein:

[0022] Figure 1 This is a flowchart of the planned water use early warning management method based on data fusion as shown in this invention. Detailed Implementation

[0023] 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. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments.

[0024] Based on the embodiments of this invention, all other embodiments obtained by those skilled in the art without inventive effort should fall within the scope of protection of this invention.

[0025] 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.

[0026] According to an embodiment of the present invention, in combination Figure 1 The flowchart shown illustrates a data fusion-based method for planned water use early warning management, comprising:

[0027] S1: Collect water consumption data and available water source data for each water supply zone, and decompose them according to time windows to form a set of water consumption characteristics for each water supply zone during different time periods, thus obtaining preliminary spatiotemporal water consumption characteristics.

[0028] S1.1: Collect water consumption data and available water source data for each water supply zone, and attach a uniform time stamp to each collected record to form the original data sequence.

[0029] Specifically, during the raw data acquisition phase, the values ​​obtained by each water supply zone through water meters, flow meters, and reservoir water level sensors must be recorded according to a unified standard, including the acquisition time, data source number, and measurement value, in order to form a raw data sequence. This ensures that each data point has a clear time and source identifier, which can support subsequent cross-zone analysis and data fusion, and avoid analysis errors caused by different data sources or time misalignments.

[0030] After the data recording is completed, the collected water consumption data and available water source data are arranged in chronological order to form the original data sequence. Each record includes a timestamp, partition number, water consumption value, and available water source value.

[0031] S1.2: Perform statistical analysis on the original data sequence to identify outliers caused by sudden events or sensor errors, and generate a corrected sequence through interpolation.

[0032] When processing the raw data sequence, statistical analysis is first performed on each data point, including calculating the mean, variance, and historical fluctuation range within the time window, in order to identify outlier data points. Outliers may be caused by leaks in the water supply network, equipment failures, or sudden peak water usage, and these anomalies can significantly bias the overall data analysis.

[0033] After identifying outliers, interpolation methods are used to correct the outlier data. The missing or outlier data is processed by linear interpolation, spline interpolation, or weighted average with the adjacent normal data within the time window, thereby generating a continuous and smooth correction sequence.

[0034] S1.3: Based on the corrected sequence, divide the data according to the set time window, and aggregate the water consumption data and available water source data in the same water supply zone to generate a water consumption feature set.

[0035] The data within each time window are aggregated to obtain statistical indicators, such as the total water consumption, maximum value, minimum value, and average value within the time period, as well as the total available water source, forming a set of water consumption characteristics for each region.

[0036] During the aggregation process, all data points in the same water supply zone are included in the calculation. Through strict time alignment, water consumption data and available water source data are matched within the same time window to ensure the accuracy of the aggregation results.

[0037] S1.4: Based on the water use feature set, extract the mean, volatility and time period differences of each time window to form a water use feature set for each time period in the partition, and output the preliminary spatiotemporal water use features.

[0038] S2: Based on the preliminary spatiotemporal water use characteristics, calculate the complementarity of adjacent water supply zones during alternating periods, and generate a set of alternative capabilities by combining available water source data.

[0039] S2.1: Align the preliminary spatiotemporal water usage characteristics with the same time window, extract the peak time and peak amplitude of each water supply zone within each time window, and form a peak sequence of the zone time period.

[0040] For each time window, iterate through the water consumption data sequence, identify the maximum water consumption and its corresponding time point, and record this value as the peak amplitude and peak time of that window.

[0041] By repeating the above operation for all time windows, a complete peak sequence for each time period can be obtained. Each sequence includes a time window identifier, peak time, and peak amplitude, providing standardized input for subsequent alternation measurements.

[0042] After the peak sequence is formed, the sequence needs to be smoothed. This invention can use moving average or local weighted regression methods to eliminate the interference of short-term spikes and ensure the stability and repeatability of peak time and amplitude in alternating measurement calculation.

[0043] The processed peak sequence is stored as a matrix structure, with rows representing time windows and columns representing partition numbers. Each matrix element contains information on peak amplitude and peak time.

[0044] S2.2: Based on the peak sequence of the partitioned time period, calculate the phase difference and peak amplitude ratio of the peak times of adjacent water supply partitions within a short time window to obtain the alternation measurement index.

[0045] Specifically, the first step is to determine the length of the short time window, which is generally a series of consecutive time windows, to ensure that it can reflect the peak synchronization or staggered peak characteristics between partitions.

[0046] For each pair of adjacent water supply zones, the peak time difference, i.e. the phase difference, is calculated for each time window. The positive vector representing the peak misalignment degree is obtained by taking the absolute value. At the same time, the peak amplitude ratio is calculated by dividing the peak water consumption by the peak water consumption of the adjacent zone to obtain the amplitude ratio.

[0047] By combining phase difference and amplitude ratio to form an alternation measurement index matrix, each matrix record includes time window identifier, water supply zone pair, phase difference value and amplitude ratio value, providing a quantifiable basis for subsequent alternation period determination.

[0048] To ensure computational stability, the phase difference and amplitude ratio are weighted averaged or weighted sliding within the short time window to reduce the impact of short-term abnormal peak fluctuations. The maximum, minimum, and average values ​​of each water supply zone within the short time window are recorded to provide a reference for subsequent threshold determination during alternating periods.

[0049] S2.3: Compare the alternation measurement index with the set index threshold range. When the phase difference falls into the preset index range and the peak amplitude ratio is greater than the amplitude threshold, the corresponding time window is determined to be an alternation period. Otherwise, it is a non-alternating period. Output the alternation period identifier sequence to provide a time reference for the construction of complementary sets.

[0050] When determining the alternation period, for each time window and each pair of adjacent water supply zones, the phase difference calculated for the short time window is compared with the set threshold interval.

[0051] The threshold range can be set based on historical water use data and empirical rules. For example, a phase difference between 10 and 30 minutes indicates reasonable peak shifting. When the phase difference meets the threshold and the peak amplitude ratio is greater than the preset amplitude threshold, the corresponding time window is determined to be an alternating period.

[0052] Each determination records the time window identifier, water supply zone pair, and determination result, forming an alternating time period identifier sequence, which serves as a time reference for constructing a complementary set.

[0053] During the determination process, consecutive alternating windows can be merged, treating adjacent consecutive alternating time windows as the same alternating period to reduce the impact of fragmented data on subsequent complementarity calculations. Non-alternating time windows are also recorded in chronological order to facilitate subsequent analysis of the substitution potential and allocation needs of partitions within non-alternating periods.

[0054] S2.4: Based on the determined alternation period, generate a complementarity set according to the water supply zone. The complementarity set is divided into complementarity intensity levels by extracting the alternation period identifier and combining the alternation measurement index of each water supply zone, which serves as a complementarity description between zones.

[0055] A superior approach is to traverse each alternation period identifier during the generation of the complementarity set, analyze the alternation measurement indicators of adjacent water supply areas, and classify the complementarity intensity level based on the combination results of phase difference and amplitude ratio, such as high complementarity, general complementarity, and low complementarity.

[0056] The complementarity level can be numerically scored. For example, when the phase difference is close to the median threshold and the amplitude ratio is close to 1.0, it is recorded as high complementarity, ensuring that the classification method is quantifiable and repeatable.

[0057] Each complementary entry includes the start and end times of the alternation period, the water supply zone pair number, and the complementarity level, forming a standardized complementary set.

[0058] During the generation process, for consecutive alternating windows spanning different time periods, the complementarity levels can be weighted and averaged to reflect the overall complementarity characteristics of the entire alternation period. Simultaneously, the complementarity set is structured and stored in matrix or table form, facilitating index queries by time window and water supply zone, providing a clear and operable basis for the automated generation of allocation plans.

[0059] S2.5: Combine the complementary set with the input available water source data to calculate the available water supply for each water supply zone during each alternation period. Difference in gaps with adjacent water supply areas Define substitution capabilities The minimum of the two values ​​is used to ensure that the allocation does not cause a secondary shortage in the source area. When there are multiple candidate zones in adjacent water supply zones, the following method is adopted. The substitution ratio is used to sort the data, obtain priorities, and record them as substitution entries for water supply zone pairs.

[0060] When generating each replacement entry, the time window, water supply zone pair number, available quantity, and priority information are recorded to ensure that the replacement capacity entry is traceable in both time and space. By processing each water supply zone pair throughout the entire alternation period, a complete set of replacement capacity entries can be formed.

[0061] S2.6: Summarize the alternative capacity items according to time sequence and water supply zone to form an alternative capacity set, and output it as a zone alternative capacity sequence, which is used as input for gap prediction and allocation scheme generation.

[0062] All generated alternative capability entries are arranged in chronological order and categorized by water supply zone numbering to form a complete set of alternative capabilities. Each record in the set includes a time window identifier, water supply zone pair, complementary strength level, available capacity, and priority, ensuring a consistent data structure that can be directly used for subsequent calculations.

[0063] During the output process, the set of alternative capabilities can be stored as a matrix or table, with rows representing time windows and columns representing water supply zone pairs. Each unit records the alternative capability, priority, and complementarity strength. This ensures data continuity and indexability, provides a clear data structure for automated processing, and directly supports subsequent gap allocation plan generation and early warning management operations.

[0064] S3: Predict the short-term water shortage in each water supply zone, generate candidate allocation schemes based on the complementarity relationship, and output a set of allocation instructions for the water supply zone.

[0065] First, within the prediction period for each water supply zone, a sliding window method (with a window length equal to the most recent k time periods) is used to calculate the historical gap average sequence. This sequence reflects the typical gap change patterns of the water supply zones in the short term, providing preliminary quantitative basis for gap prediction. The sliding window length k is set according to the actual water supply scheduling frequency.

[0066] The historical gap average sequence is used as the basic gap trend sequence and compared with the actual historical gap to obtain the error sequence. Pattern recognition is performed on the error sequence within a sliding window to distinguish between systematic errors and random errors. The systematic errors and random errors are corrected separately to form a layered corrected error sequence.

[0067] It should be noted that traditional forecasting methods often directly extrapolate trends using raw historical water gap values. This approach is prone to deviations from long-term trends and lacks stability when cyclical fluctuations or short-term anomalies exist. This invention, however, first calculates a historical gap average sequence based on a sliding window within a preset forecast period, using this as the base gap trend sequence. This smooths out some short-term outliers, making the trend more stable and representative. After obtaining the base gap trend sequence, it is compared point-by-point with the actual historical gap, and the difference between the two is calculated. The set of differences constitutes the error sequence. This method effectively separates the trend component from the deviation component, laying the foundation for subsequent error identification and correction. This invention differs from conventional single-difference extraction by introducing an average sequence as a benchmark, which can suppress the interference of occasional fluctuations and improve the reliability of the information reflected in the error sequence.

[0068] Specifically, the generation of the corrected error sequence includes:

[0069] Within a sliding window, the mean shift, volatility ratio, and direction of the error sequence are extracted. The mean shift reflects the overall deviation of the error within the window, revealing whether there is a persistent trend. The volatility ratio reflects the dispersion of the error within the window, revealing whether there are short-term sharp fluctuations. The direction factor records the positive and negative changes of the error, revealing whether the error exhibits trend consistency or frequent reversals within the window. These three types of feature parameters are independent yet interrelated, ensuring that both systematic deviations and short-term fluctuations are captured while avoiding judgment biases caused by a single parameter.

[0070] When the mean drift remains unidirectionally increasing or decreasing within a continuous sliding window, and the direction factor remains stable and consistent, the sliding window is marked as a systematic error segment. When the fluctuation ratio is greater than 1 within a continuous sliding window and the direction changes P times (which can be set as needed, for example, 3 times), it indicates that the error does not have a fixed trend direction, but rather exhibits frequent reversals, and the sliding window is marked as a random error segment.

[0071] For sliding window sequences marked as systematic error segments, this invention no longer performs point-by-point correction, but instead uses weighted averaging or exponential smoothing to obtain a long-term correction trend curve. For sliding window sequences marked as random error segments: based on the ratio of short-term volatility... Calculate the residual amplification factor The correction value is It is used to adjust local short-term corrections.

[0072] It should be noted that for the random error range, conventional smoothing corrections can excessively weaken the prediction's response to short-term fluctuations. Therefore, this invention calculates the ratio of short-term fluctuation amplitudes to reflect the intensity of local fluctuations, and then determines the residual amplification factor based on this ratio. The residual amplification factor can dynamically amplify or reduce the local correction value, thereby enhancing the prediction results' response to short-term fluctuations.

[0073] This approach not only maintains the stability of the overall trend in the prediction results, but also has the ability to sensitively capture local anomalies, making it more consistent with the dynamic changes in actual water shortages.

[0074] Correction values ​​from different sources are concatenated in time sequence to form a hierarchical correction error sequence.

[0075] Among them, the short-term volatility is higher The calculation is performed by: selecting water consumption or gap data for K consecutive time periods, determining the difference between the maximum and minimum values ​​of water consumption or gap data within a certain time period as the short-term fluctuation range; then using the average or median of the water consumption or gap data within the K consecutive time periods as the benchmark level, and using the ratio of the fluctuation range to the benchmark level as the short-term fluctuation range ratio; where K is a constant.

[0076] By weighted and fused the basic gap trend and the correction error sequence, a residual-corrected short-term gap prediction sequence is generated, resulting in a residual-corrected gap prediction sequence for each water supply zone during the prediction period. This sequence contains gap information for each water supply zone within each time window, enabling direct matching of complementarity and substitution capacity data. The residual correction method improves prediction accuracy and operability by capturing fluctuations in recent deviation trends, while avoiding prediction lag caused by solely relying on historical averages.

[0077] Optionally, the residual correction gap prediction sequence can be compared with the previous round of actual gap data to calculate the prediction accuracy index (such as mean absolute error, MAE) to determine the reasonableness of the prediction. If the prediction accuracy index exceeds the set threshold, the sliding window length or error weight can be further adjusted to ensure that the prediction sequence has credibility in each water supply interval.

[0078] The generated set of alternative capabilities is matched with the predicted residual correction gap values ​​for each water supply zone. If a water supply zone has a positive gap (supply less than demand) during the prediction period, then zones with surplus water volume during the same period are selected from the complementarity set to form preliminary gap-replenishment pairs. Each preliminary pair includes the gap zone number, replenishment zone number, prediction period, and available water volume information.

[0079] To avoid redundancy, the gap-supply initial pair only retains a one-to-one mapping relationship, with a gap partition corresponding only to the supply partition with the highest substitution capacity.

[0080] For each gap-supply preliminary pair, calculate the matching index:

[0081]

[0082] in, Indicates water supply zones The gap amount, Indicates water supply zones Adjustable margin;

[0083] All gap-replenishment pairs are initially sorted based on the matching degree index, and the top M pairs are retained to form a candidate solution set; where M is a constant to ensure that the number of solutions is controllable and covers the main gap needs. The sorting rule is based on the matching degree index, while also taking into account the complementarity strength level and the adjustable water volume, to ensure that the generated candidate solutions are efficient and reasonable.

[0084] The candidate scheme set is converted into an executable allocation instruction, which includes: the starting zone, target zone, corresponding time period, and planned allocation volume of water.

[0085] For situations involving multiple supply zones, allocation is carried out step by step according to the matching priority until the gap is met, ensuring that the allocation process is continuous and conflict-free, and recording the allocation amount and remaining gap amount information at each step, and outputting a set of allocation instructions for each zone.

[0086] S4: When the predicted short-term water shortage exceeds the shortage threshold, the alternating buffer window is activated to shift or allocate non-critical water demand in some water supply zones across time periods, and outputs the warning level and corresponding allocation instructions.

[0087] In the short-term gap prediction results, the gap amount for each water supply zone during the prediction period is first determined, and the gap amount is compared with a set gap threshold. When the gap of a water supply zone exceeds the gap threshold during the prediction period, it is determined that a buffer mechanism needs to be activated, and a buffer zone is established.

[0088] Once the buffer zones are established, the predicted shortfall, time window, and relevant historical water usage data for each zone will be recorded. This ensures accurate calculations during water transfers or inter-zone allocations and provides quantifiable data for setting early warning levels. By clearly defining buffer zones, the system can make orderly scheduling decisions when facing sudden or periodic water shortages.

[0089] Establish an alternating buffer window for the buffered partition. The length of the alternating buffer window is equal to H times the gap prediction period length (e.g., 0.5~0.7), which is used to shift non-critical requirements.

[0090] The design logic of the alternating buffer window includes: during the gap prediction period, non-critical water use (such as landscaping, flushing, etc.) is allowed to be postponed to the second half of the alternating buffer window or transferred to an adjacent water supply zone.

[0091] Specifically, the alternating buffer window design logic explicitly stipulates that non-critical water use during the gap prediction period can be postponed to the latter half of the window or transferred across adjacent water supply zones. The window length and location are dynamically adjusted through historical load analysis and real-time gap prediction to ensure that shifting or cross-zone allocation will not trigger new peaks or system conflicts. The buffer window parameters for each buffer zone include the time period range and the upper limit of the shiftable demand. The parameter records are used for feasibility calculations of subsequent shift operations, forming a traceable data link to provide a complete basis for allocation execution.

[0092] The output is the buffer window parameters (time range, maximum amount of shiftable demand) for each buffered partition.

[0093] Within the alternating buffer window, non-critical needs within the water supply zones are first identified and their proportions calculated. These non-critical needs are then prioritized based on water usage, including but not limited to landscaping, public facility flushing, and off-peak domestic water use. For example, industrial cooling water has a lower priority than public landscaping, and public landscaping has a lower priority than off-peak domestic water use. According to priority, non-critical needs are progressively shifted to the latter half of the alternating buffer window, resulting in a time-period adjustment table that records the original and adjusted time periods for each water supply zone's needs. This adjustment table provides foundational data for subsequent calculations of remaining demand gaps and inter-zone allocation, while ensuring a balanced distribution of water supply system load across time periods and preventing the creation of new gaps.

[0094] If the gap in the adjusted time period still exceeds the gap threshold, the allocation surplus is found in the complementary zones to form cross-regional supplementation; the allocation rules are based on the matching degree calculation, but the matching degree priority is limited to the part corresponding to non-critical demand; the cross-regional allocation instruction is output, including the water supply starting point, target zone, allocation water volume and corresponding time period.

[0095] It can be seen that by limiting the matching range, key needs are not affected, while the available capacity of complementary partitions is used to alleviate peak gaps, thus achieving coordinated allocation across multiple partitions.

[0096] Finally, based on the proportion of the remaining gap to the threshold, the warning levels are set as follows: Level I: Gap ≤ 1.2 times the gap threshold (mild tension, can be resolved with buffer); Level II: Gap ≤ 1.5 times the gap threshold (moderate tension, requires cross-regional allocation); Level III: Gap > 1.5 times the gap threshold (severe tension, requires multi-regional coordination) (the above values ​​can be set according to actual conditions); the final warning level + allocation instruction set is output as the basis for the execution of the water supply dispatching system.

[0097] Through the above operations, the combination of alternating buffer windows and cross-regional allocation ensures both the scientific validity and feasibility of transferring non-critical needs, and enables timely mitigation of gaps in cross-regional allocation. Simultaneously, it outputs standardized early warning levels and allocation instruction sets, providing highly operable, scientifically sound, and quantifiable data support for planned water management, enabling dynamic response and optimized scheduling of short-term water shortages. Compared to existing technologies, this approach can automatically adjust non-critical needs when the gap exceeds a threshold, rationally utilize complementary regional allocation reserves, and provide decision-making basis through quantified levels, thereby improving the precision and reliability of planned water management.

[0098] The present invention also includes one or more processors and a memory.

[0099] The memory is used to store operable instructions that, when executed by the one or more processors, cause the one or more processors to perform operations, including the flow of the data fusion-based planned water use early warning management method of the foregoing embodiments, especially... Figure 1 The flowchart of the method is shown.

[0100] Other aspects disclosed in the embodiments of the present invention also propose a computer-readable medium for storing software including instructions executable by one or more computers, which, upon execution, cause the one or more computers to perform operations including the flow of the data fusion-based planned water use early warning management method of the foregoing embodiments, particularly... Figure 1 The flowchart of the method is shown.

[0101] It should be recognized that embodiments of the present invention may be implemented or carried out by computer hardware, a combination of hardware and software, or by computer instructions stored in a non-transitory computer-readable storage medium.

[0102] The method can be implemented using standard programming techniques, including a non-transitory computer-readable storage medium configured with a computer program in the computer program, wherein the storage medium is configured such that the computer operates in a specific and predefined manner.

[0103] Each program can be implemented in a high-level procedural or object-oriented programming language to communicate with the computer system; however, if required, the program can be implemented in assembly or machine language.

[0104] In any case, the language can be either compiled or interpreted.

[0105] Furthermore, for this purpose, the program can run on programmed application-specific integrated circuits.

[0106] The processes described herein (or variations and / or combinations thereof) can be executed under the control of one or more computer systems configured with executable instructions, and can be implemented by hardware or a combination thereof as code (e.g., executable instructions, one or more computer programs, or one or more applications) that commonly executes on one or more processors. The computer program includes a plurality of instructions executable by one or more processors.

[0107] Furthermore, the method can be implemented in any suitable computing platform, including but not limited to personal computers, minicomputers, mainframes, workstations, networked or distributed computing environments, standalone or integrated computer platforms, or in communication with charged particle tools or other imaging devices.

[0108] Various aspects of the present invention can be implemented in machine-readable code stored on a non-transitory storage medium or device, whether portable or integrated into a computing platform, such as a hard disk, optical read and / or write storage medium, RAM, ROM, etc., such that it can be read by a programmable computer, and when the storage medium or device is read by the computer, it can be used to configure and operate the computer to perform the processes described herein.

[0109] Furthermore, machine-readable code, or parts thereof, can be transmitted via wired or wireless networks.

[0110] When such media includes instructions or programs that combine with a microprocessor or other data processor to implement the steps described above, the invention described herein includes these and other different types of non-transitory computer-readable storage media.

[0111] 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 planned water use early warning management method based on data fusion, characterized in that, include: Water consumption data and available water source data for each water supply zone are collected and decomposed according to time windows to form a set of water consumption characteristics for each water supply zone during different time periods, thus obtaining preliminary spatiotemporal water consumption characteristics. Based on preliminary spatiotemporal water use characteristics, the complementarity of adjacent water supply zones during alternating periods is calculated, and a set of alternative capabilities is generated by combining available water source data. Predict the short-term water shortage in each water supply zone, generate candidate allocation schemes based on complementarity, and output a set of allocation instructions for the water supply zone. When the predicted short-term water shortage exceeds the shortage threshold, an alternating buffer window is activated to shift or allocate non-critical water demand from some water supply zones across time periods, and to output the warning level and corresponding allocation instructions. The calculation of the complementarity of adjacent water supply zones during alternating periods includes: aligning preliminary spatiotemporal water usage characteristics according to the same time window; extracting the peak time and peak amplitude of each water supply zone within each time window to form a peak sequence for each zone's time period; based on the peak sequence of the zone's time period, calculating the phase difference and peak amplitude ratio of the peak times of adjacent water supply zones within a short time window to obtain an alternation measurement index; comparing the alternation measurement index with a set index threshold range; when the phase difference falls within the preset index range and the peak amplitude ratio is greater than the amplitude threshold, the corresponding time window is determined to be an alternating period; otherwise, it is a non-alternating period; and outputting an alternating period identifier sequence. The process of generating an alternative capacity set by combining available water source data includes: generating a complementary set based on the determined alternation period, according to water supply zones; classifying the complementary set by extracting the alternation period identifier and combining it with the alternation measurement index of each water supply zone; and combining the complementary set with the input available water source data to calculate the available capacity of each water supply zone within each alternation period. Difference in gaps with adjacent water supply areas Define substitution capabilities The minimum of the two values ​​is used when there are multiple candidate zones in adjacent water supply zones. The substitution ratios are sorted to obtain priorities and recorded as substitution entries for water supply zone pairs. The substitution entries are then summarized in chronological order and by water supply zone pairs to form a set of substitution capacities, which is output as a sequence of zone substitution capacities.

2. The planned water use early warning management method based on data fusion as described in claim 1, characterized in that: The formation of the water use characteristic set includes: Collect water consumption data and available water source data for each water supply zone, and add a uniform time stamp to each collected record to form the original data sequence; Statistical analysis is performed on the original data sequence to identify outliers caused by sudden events or sensor errors, and a corrected sequence is generated by interpolation. Based on the corrected sequence, the water consumption data and available water source data within the same water supply zone are divided according to the set time window to generate a water consumption feature set. The preliminary spatiotemporal water use characteristics include: based on the water use characteristic set, extracting the mean, volatility and time period differences of each time window to form a regional time period water use characteristic set, and outputting the preliminary spatiotemporal water use characteristics.

3. The planned water use early warning management method based on data fusion as described in claim 2, characterized in that: The prediction of short-term water shortages in each water supply zone includes: Within the forecast period for each water supply zone, a sliding window method is used to calculate the historical gap average sequence; The historical gap average sequence is used as the basic gap trend sequence, and compared with the actual historical gap to obtain the error sequence; Pattern recognition is performed on the error sequence within a sliding window to distinguish between systematic and random errors; systematic and random errors are then corrected separately to form a hierarchically processed corrected error sequence. The basic gap trend sequence and the correction error sequence are weighted and fused to generate a short-term gap prediction sequence after residual correction, thus obtaining the residual corrected gap prediction sequence for each water supply zone during the prediction period.

4. The planned water use early warning management method based on data fusion as described in claim 3, characterized in that: The generation of the correction error sequence includes: Within the sliding window, extract the mean shift, volatility ratio, and direction of the error sequence; When the mean shift remains unidirectionally increasing or decreasing within a continuous sliding window, and the direction factor remains stable and consistent, the sliding window is marked as a systematic error segment. When the fluctuation ratio is greater than 1 within a continuous sliding window and the direction changes P times, the sliding window is marked as a random error segment; where P is a constant. For the sliding window sequence marked as a systematic error segment: weighted averaging or exponential smoothing is used to obtain the long-term correction trend curve. ; For sliding window sequences marked as random error segments: based on the ratio of short-term volatility... Calculate the residual magnification factor The correction value is It is used to adjust local short-term corrections; Correction values ​​from different sources are concatenated in time sequence to form a hierarchical correction error sequence; The short-term volatility ratio The calculation is performed by: selecting water consumption or gap data for K consecutive time periods, determining the difference between the maximum and minimum values ​​of water consumption or gap data within a certain time period as the short-term fluctuation range; then using the average or median of the water consumption or gap data within the K consecutive time periods as the benchmark level, and using the ratio of the fluctuation range to the benchmark level as the short-term fluctuation range ratio; where K is a constant.

5. The planned water use early warning management method based on data fusion as described in claim 4, characterized in that: The generation of candidate allocation schemes based on complementarity includes: The generated set of alternative capabilities is matched with the predicted residual correction gap values ​​of each water supply zone: if a water supply zone has a positive gap during the prediction period, then the zones with remaining water volume during the same period are found from the complementarity set to form a preliminary gap-replenishment pair. The initial gap-supply pair retains only a one-to-one mapping relationship, with a gap partition corresponding only to the supply partition with the highest substitution capacity; For each gap-supply preliminary pair, calculate the matching index: in, Indicates water supply zones The gap amount, Indicates water supply zones Adjustable margin; All gap-supply pairs are initially sorted according to the matching degree index, and the top M pairs are retained to form a candidate solution set; where M is a constant. The candidate scheme set is converted into an executable allocation instruction, the allocation instruction including: the starting zone, the target zone, the corresponding time period, and the planned allocation volume; In cases involving multiple supply zones, allocation is made step by step according to the matching priority until the gap is met; Output a set of interval allocation instructions.

6. The planned water use early warning management method based on data fusion as described in claim 5, characterized in that: The activation of the alternating buffer window, which involves shifting or allocating non-critical water demand from some water supply zones across time periods, includes: When the water supply zone's shortfall exceeds the shortfall threshold during the forecast period, it is determined that a buffer mechanism needs to be activated and a buffer zone needs to be established. An alternating buffer window is established for the buffered partition, and the length of the alternating buffer window is equal to H times the length of the gap prediction period; The design logic of the alternating buffer window includes: during the gap prediction period, non-critical water use is allowed to be postponed to the second half of the alternating buffer window or transferred to an adjacent water supply zone; Within the alternating buffer window, non-critical needs within the water supply zones are first identified and their proportions are calculated. By prioritizing water usage, the needs are gradually shifted to the later part of the alternating buffer window. After shifting, a time period adjustment table is formed, recording the original time period and the adjusted time period for the needs of each water supply zone. If the gap in the adjusted time period still exceeds the gap threshold, then the allocation surplus is searched in the complementary partition to form cross-regional supplementation; the allocation rules are based on the matching degree calculation, but the matching degree priority is limited to the part corresponding to non-critical needs; and the cross-regional allocation instruction is output.

7. 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 data fusion-based planned water use early warning management method according to any one of claims 1 to 6.

8. A computer-readable medium for storing software, characterized in that: The software includes instructions executable by one or more computers, which, upon execution, cause the one or more computers to perform operations including the flow of the data fusion-based planned water use early warning management method as described in any one of claims 1 to 6.

Images