Cloud platform-based multi-site water quality monitoring data management system

Abstract

The application relates to the technical field of data processing, and discloses a multi-station water quality monitoring data management system based on a cloud platform, which comprises a data acquisition module, an abnormal process analysis module, an event correlation analysis module, a reverse correction module, a correlation state updating module and an event data management module. The data acquisition module generates a site abnormal data segment; the abnormal process analysis module generates an abnormal process label; the event correlation analysis module generates an event correlation time interval and a correlation conservation state, and generates an event correlation inconsistent state according to a candidate abnormal data segment; the reverse correction module generates an event correlation confirmation state; the correlation state updating module updates the correlation conservation state; and the event data management module generates a cross-site abnormal event data group. The application can distinguish cross-site propagation abnormalities from local independent abnormalities by constructing a site abnormal data segment, an abnormal process label and a correlation conservation state, and performing corresponding relationship matching of the candidate abnormal data segment and the correlation conservation state.

Description

Technical Field

[0001] This invention relates to the field of data processing technology, and in particular to a multi-site water quality monitoring data management system based on a cloud platform. Background Technology

[0002] With the development of water environment monitoring technology, Internet of Things (IoT) technology, and cloud computing technology, multi-site water quality monitoring systems have been widely applied in water environment monitoring scenarios such as rivers, lakes, reservoirs, and watersheds. Multiple water quality monitoring stations upload the collected water quality monitoring data to a cloud platform via a communication network. The cloud platform then performs unified storage, processing, and management of the data from each monitoring station, and combines multi-site monitoring data to conduct anomaly analysis, achieving centralized management and comprehensive analysis of cross-site water quality monitoring data.

[0003] Existing multi-site water quality monitoring data management methods typically analyze the correlation between abnormal data from different monitoring stations based on the water quality monitoring data uploaded by each monitoring station, combined with the location relationship between monitoring stations, water flow status data, or preset time windows. They also establish cross-site abnormal event correlations based on the time of abnormal occurrence, changes in abnormal indicators, or spatial location relationships to form cross-site abnormal event data management results.

[0004] However, in special circumstances such as the superposition of continuous pollution events, the convergence and propagation of abnormal events in tributaries and main streams, and the simultaneous occurrence of local water anomalies at local monitoring stations, abnormal events from different sources overlap in both time and space dimensions. Establishing cross-site abnormal event correlations solely based on the time of occurrence, spatial location, or fixed time window makes it difficult to effectively distinguish between cross-site propagated anomalies and local independent anomalies, which can easily lead to erroneous correlations between different abnormal events, resulting in inaccurate cross-site abnormal event correlation results. Summary of the Invention

[0005] To overcome the above shortcomings, this invention provides a cloud-based multi-site water quality monitoring data management system, which aims to improve the existing technology's difficulty in effectively distinguishing between cross-site propagated anomalies and local independent anomalies, which easily leads to erroneous correlations between different anomaly events, resulting in inaccurate cross-site anomaly event correlation results.

[0006] The present invention provides the following technical solution: a multi-site water quality monitoring data management system based on a cloud platform, comprising: multiple water quality monitoring stations and a cloud platform that is communicatively connected to the multiple water quality monitoring stations;

[0007] The cloud platform includes a data acquisition module, an abnormal process analysis module, an event correlation analysis module, a reverse correction module, a correlation status update module, and an event data management module.

[0008] The data acquisition module generates abnormal data segments for the stations based on water quality monitoring data and station operation status data.

[0009] The abnormal process analysis module determines the order of occurrence, recovery order, direction of change, and stage transition relationship of each water quality indicator based on the abnormal data segment of the station, and generates an abnormal process marker.

[0010] The event correlation analysis module determines the event correlation time interval and filters candidate abnormal data segments based on the abnormal data segments and water flow status data at the site. It constructs a correlation conservation state based on the abnormal process marker, maps the candidate abnormal data segments to the correlation conservation state, performs corresponding relationship matching, and generates an event correlation inconsistency state.

[0011] The reverse correction module performs classification correction on the event association time interval and candidate abnormal data segment according to the event association inconsistency status, and generates an event association confirmation status based on the correction result;

[0012] The association status update module updates the association conservation status according to the event association confirmation status, and uses the updated association conservation status as the association constraint for subsequent event association analysis at monitoring stations.

[0013] The event data management module generates cross-site abnormal event data groups based on the associated confirmation status of multiple events.

[0014] Preferably, the data acquisition module generates a site anomaly data segment based on the site operation status data and the water quality monitoring data, specifically including:

[0015] Valid water quality monitoring data are selected based on equipment operating status, communication status, sampling status, and time synchronization status.

[0016] A local baseline state is established based on historical stable monitoring data. The effective water quality monitoring data is compared with the local baseline state, and abnormal data segments of the stations are generated based on the deviation state within multiple consecutive sampling periods.

[0017] Preferably, the anomaly analysis module generates anomaly process markers based on the site's anomaly data segments, specifically including:

[0018] Based on the sampling times when each water quality indicator continuously meets the abnormal and recovery conditions, the time of abnormal occurrence and the time of abnormal recovery are determined.

[0019] The sequence of anomalies, the sequence of anomalies, and the phase transition relationships are determined based on the time of anomaly occurrence, the time of anomaly recovery, and the direction of continuous change, and anomaly process markers are generated.

[0020] Preferably, the event correlation analysis module performs candidate abnormal data segment filtering based on the station abnormal data segment and the water flow status data, specifically including:

[0021] The water flow transmission time is determined based on the location relationship between adjacent monitoring stations and water flow status data.

[0022] The event-related time interval is determined based on the start and end times of the abnormal data segments from the upstream monitoring stations and the water flow transmission time.

[0023] Abnormal data segments from downstream monitoring stations located within the event-related time interval are identified as candidate abnormal data segments.

[0024] Preferably, the event correlation analysis module performs correlation conservation state construction based on the abnormal process marker, specifically including:

[0025] The core and secondary abnormal indicators are determined based on the participation status of each water quality indicator in different abnormal stages.

[0026] Based on the anomaly occurrence sequence, anomaly recovery sequence, change direction, and stage transition relationship corresponding to the core anomaly indicators, an associated conservation state is constructed.

[0027] Preferably, the event correlation analysis module generates an event correlation inconsistency state based on the candidate abnormal data segment and the correlation conservation state, specifically including:

[0028] Using the core anomaly indicators as the main index, the candidate anomaly data segments are reorganized according to the data structure of the aforementioned correlation conservation state to form a candidate correlation structure;

[0029] The candidate association structure is matched with the core anomaly indicators, anomaly occurrence order, anomaly recovery order, change direction, and stage transition relationship in the association conservation state.

[0030] The anomaly occurrence time and anomaly recovery time of the candidate anomaly data segment are compared with the event-related time interval;

[0031] Based on the corresponding matching results and time comparison results, generate time-related inconsistency states, indicator-related inconsistency states, and stage structure inconsistency states.

[0032] Preferably, the reverse correction module performs classification reverse correction based on the time association inconsistency and stage structure inconsistency in the event association inconsistency states, specifically including:

[0033] The event association time interval is adjusted based on the inconsistent time association state and the actual time of anomaly occurrence and recovery of candidate abnormal data segments;

[0034] Based on the inconsistent state of the stage structure, the candidate abnormal data segment is divided into associated abnormal sub-segment and local abnormal sub-segment;

[0035] Establish event correlations based on core anomaly indicators and the order of anomaly occurrence among multiple associated conservation states.

[0036] Preferably, the reverse correction module performs association state reconstruction based on the corrected event association time interval, the associated abnormal sub-segment, and the event association relationship, specifically including:

[0037] The core anomaly indicators, anomaly occurrence order, anomaly recovery order, change direction, and stage transition relationship corresponding to the associated anomaly sub-segments are retained, while the structural information corresponding to the local anomaly sub-segments is removed.

[0038] Based on the retained data and the event relationships, the association conservation state is regenerated, and an event association confirmation state is formed.

[0039] Preferably, the association state update module performs an association conservation state update based on the event association confirmation state, specifically including:

[0040] Extract the core abnormal indicators that are common to adjacent monitoring stations;

[0041] The order of anomalies and the order of recovery are updated based on the anomaly occurrence time and the anomaly recovery time corresponding to the common core anomaly indicators.

[0042] Based on the updated core anomaly indicators, the order of anomaly occurrence, and the order of anomaly recovery, the associated conservation state corresponding to the subsequent monitoring stations is regenerated.

[0043] Preferably, the event data management module performs cross-site abnormal event data management based on the confirmation status of multiple event associations, specifically including:

[0044] Based on the event association relationships in the event association confirmation status, site abnormal data segments with continuous event association relationships are combined to form cross-site abnormal event data groups;

[0045] The cross-site abnormal event data group to be updated is determined based on the monitoring station and associated event number corresponding to the newly added water quality monitoring data, and the cross-site abnormal event data group is updated according to the updated event association confirmation status.

[0046] The present invention has the following beneficial effects:

[0047] 1. This invention constructs site-specific abnormal data segments, abnormal process markers, and associated conservation states, and maps candidate abnormal data segments to associated conservation states to perform corresponding relationship matching, generating event association inconsistency states. This can distinguish between cross-site propagation anomalies and local independent anomalies, solving the problem of inaccurate cross-site abnormal event association under the conditions of continuous pollution event superposition or local anomaly interference.

[0048] 2. This invention forms a classification reverse correction mechanism by performing event association time interval adjustment, association abnormal sub-segment division and event association relationship establishment according to the event association inconsistency status. It can adopt corresponding data processing methods for different types of event association deviations, thereby improving the consistency of multi-site abnormal event association results.

[0049] 3. This invention regenerates the association conservation state based on the event association confirmation state, and uses the regenerated association conservation state as the association constraint template for subsequent event association analysis at monitoring sites, thereby realizing the recursive update of the association state of cross-site abnormal events and ensuring the continuous organization and unified management of multi-site abnormal event data groups. Attached Figure Description

[0050] Figure 1 This is a schematic diagram of the structure of the multi-site water quality monitoring data management system based on a cloud platform proposed in this invention;

[0051] Figure 2 This is a schematic diagram of the process for generating and analyzing the correlation conservation state proposed in this invention;

[0052] Figure 3 This is a schematic diagram of the classification reverse correction and association state update process proposed in this invention. Detailed Implementation

[0053] The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0054] Example 1: Refer to Figure 1 The present invention provides a multi-site water quality monitoring data management system based on a cloud platform, including: multiple water quality monitoring stations and a cloud platform that communicates with the multiple water quality monitoring stations;

[0055] The cloud platform includes a data acquisition module, an abnormal process analysis module, an event correlation analysis module, a reverse correction module, a correlation status update module, and an event data management module.

[0056] The data acquisition module generates abnormal data segments for the stations based on water quality monitoring data and station operation status data;

[0057] The abnormal process analysis module determines the order of occurrence, recovery order, direction of change, and stage transition relationship of each water quality indicator based on the abnormal data segments of the site, and generates abnormal process markers.

[0058] The event correlation analysis module determines the event correlation time interval and filters candidate abnormal data segments based on the abnormal data segments and water flow status data at the site. It constructs the correlation conservation state based on the abnormal process marker, maps the candidate abnormal data segments to the correlation conservation state, performs corresponding relationship matching, and generates an event correlation inconsistency state.

[0059] The reverse correction module performs classification correction on the event association time interval and candidate abnormal data segments based on the event association inconsistency status, and generates an event association confirmation status based on the correction results;

[0060] The association status update module updates the association conservation status based on the event association confirmation status, and uses the updated association conservation status as the association constraint for subsequent event association analysis at monitoring stations.

[0061] The event data management module generates cross-site abnormal event data groups based on the associated confirmation status of multiple events.

[0062] Specifically, multiple water quality monitoring stations are deployed in different monitoring areas. Each water quality monitoring station collects water quality monitoring data for its corresponding area according to a set sampling cycle, and simultaneously collects station operation status data and water flow status data. The collected data is then sent to the cloud platform.

[0063] The data acquisition module receives data uploaded from various water quality monitoring stations, organizes the water quality monitoring data, station operation status data, and water flow status data in a unified manner, generates station anomaly data segments, and sends them to the anomaly process analysis module. The anomaly process analysis module determines the abnormal change process corresponding to each water quality indicator based on the station anomaly data segments, generates anomaly process markers, and sends them to the event correlation analysis module.

[0064] The event correlation analysis module generates event correlation time intervals based on the locational relationships between monitoring stations, water flow status data, and abnormal data segments from each station. It then generates correlation conservation states based on abnormal process markers. Within the event correlation time intervals, candidate abnormal data segments are identified and mapped to the correlation conservation states, completing the matching process and generating an event correlation inconsistency state. The reverse correction module performs classified reverse correction based on the event correlation inconsistency state, adjusting the event correlation time intervals, candidate abnormal data segments, and event correlation relationships, and generating an event correlation confirmation state.

[0065] The association status update module updates the association conservation status based on the event association confirmation status and sends the updated association conservation status to the event association analysis module as a data constraint for subsequent event association analysis at monitoring stations. The event data management module establishes cross-site abnormal event data groups based on multiple event association confirmation statuses, and stores and updates these cross-site abnormal event data groups, realizing unified organization, association analysis, and data management of water quality monitoring data from multiple stations.

[0066] Furthermore, the data acquisition module generates abnormal data segments based on the site's operational status data and water quality monitoring data, specifically including:

[0067] Valid water quality monitoring data are selected based on equipment operating status, communication status, sampling status, and time synchronization status.

[0068] A local baseline state is established based on historical stable monitoring data. The effective water quality monitoring data is compared with the local baseline state, and abnormal data segments of the station are generated based on the deviation state within multiple consecutive sampling periods.

[0069] Specifically, the data acquisition module first receives water quality monitoring data, station operation status data, and water flow status data uploaded by each water quality monitoring station. The station operation status data includes equipment operation status, communication status, sampling status, and time synchronization status. The data acquisition module determines whether the current sampled data meets preset data validity conditions based on the station operation status data. When the equipment operation status is normal, the communication status is normal, the sampling status is complete, and the time synchronization status meets the sampling requirements, the corresponding water quality monitoring data is determined to be valid water quality monitoring data; otherwise, the corresponding water quality monitoring data is marked as invalid data and stops participating in the subsequent generation of abnormal data segments from other stations.

[0070] For each monitoring station, the data acquisition module filters historical stable monitoring data from the continuously collected valid water quality monitoring data prior to the current sampling time and establishes a local baseline state. Let the first... The monitoring station number Water quality indicators at any time The monitoring value is expressed as The local reference state is represented as The local baseline state is then calculated using the moving average of historical stability monitoring data:

[0071] ;

[0072] in: Indicates the first The monitoring station number Local baseline status corresponding to each water quality indicator; Indicates the number of sampling times prior to the current sampling time. The monitoring value corresponding to each sampling time; This represents the number of historical samples used in the local baseline state calculation. In this scheme, the number of historical samples... The preferred sampling period is 30 to 60 consecutive sampling cycles to ensure that the local baseline state can reflect the stable trend of the water quality indicators corresponding to the current monitoring station.

[0073] The data acquisition module calculates the deviation state based on the difference between the current monitored value and the local baseline state. The deviation state is represented as follows:

[0074] ;

[0075] in: Indicates a deviation from the state; Indicates the current monitoring value; This indicates the corresponding local reference state.

[0076] In this scheme, to avoid anomalous data segments caused by single sampling fluctuations, the data acquisition module further determines whether the current deviation state constitutes a continuous anomaly by combining the continuous change state. The continuous change state is determined by the direction and duration of the deviation state change over multiple consecutive sampling periods; when the deviation state maintains the same direction of change over multiple consecutive sampling periods, and the deviation state continues for more than a preset duration period, the corresponding monitoring data is identified as anomalous data. The preset duration period is preferably 3 to 5 consecutive sampling periods.

[0077] Subsequently, the data acquisition module combines consecutive abnormal data according to the sampling time sequence to form a site abnormal data segment. Let the site abnormal data segment be represented as:

[0078] ;

[0079] in: Indicates the first Each monitoring station corresponds to an abnormal data segment; Indicates the start time of the abnormal data segment at the site; Indicates the end time of the abnormal data segment at the site; This indicates the deviation state at the corresponding sampling time.

[0080] When there is only normal data within one consecutive sampling period between two adjacent abnormal data segments, and the main abnormal indicators corresponding to the two abnormal data segments are consistent, the data acquisition module will merge the two abnormal data segments. When the main abnormal indicators corresponding to the two abnormal data segments are different or the interval exceeds the preset sampling period, they will be retained as independent abnormal data segments, and the corresponding start time, end time and data items involved in the anomaly will be recorded respectively, providing a unified data input for subsequent anomaly process analysis.

[0081] Through the above processing, the data acquisition module completes the screening of effective water quality monitoring data, the establishment of local baseline states, the calculation of deviation states, the identification of continuous change states, and the generation of abnormal data segments at stations, providing the abnormal process analysis module with abnormal data segments at stations that have a continuous time structure and a unified data organization format.

[0082] Furthermore, the anomaly analysis module generates anomaly process markers based on the site's anomaly data segments, specifically including:

[0083] Based on the sampling times when each water quality indicator continuously meets the abnormal and recovery conditions, the time of abnormal occurrence and the time of abnormal recovery are determined.

[0084] The sequence of anomalies, the sequence of recovery from anomalies, and the phase transition relationships are determined based on the time of anomaly occurrence, the time of anomaly recovery, and the direction of continuous change, and anomaly process markers are generated.

[0085] Specifically, the anomaly analysis module receives the site anomaly data segments output by the data acquisition module and analyzes the change process of each water quality indicator within the site anomaly data segment in chronological order. For each water quality indicator, it first determines its corresponding anomaly occurrence time and anomaly recovery time. The anomaly occurrence time is defined as the sampling time when the corresponding water quality indicator first continuously meets the anomaly judgment conditions, and the anomaly recovery time is defined as the sampling time when the corresponding water quality indicator recovers to near the local baseline state and continuously meets the recovery conditions. Let the first... The time of occurrence of the anomaly corresponding to each water quality indicator is indicated as follows: The abnormal recovery time is represented as The duration of the corresponding water quality index anomaly is expressed as follows:

[0086] ;

[0087] in: Indicates the first The duration of abnormality corresponding to each water quality indicator; Indicates the time when the anomaly occurred; This indicates the time of anomaly recovery. In this scheme, the anomaly duration is preferably no less than 3 consecutive sampling periods to avoid instantaneous fluctuations from being involved in subsequent anomaly process analysis.

[0088] The anomaly analysis module establishes an anomaly occurrence sequence based on the anomaly occurrence time corresponding to each water quality indicator, and an anomaly recovery sequence based on the anomaly recovery time corresponding to each water quality indicator. When multiple water quality indicators have the same anomaly occurrence time, the priority is determined according to the deviation of the corresponding monitoring value from largest to smallest; when multiple water quality indicators have the same anomaly recovery time, the recovery sequence is determined according to the recovery rate of the corresponding deviation state from fastest to slowest, thus ensuring that the anomaly occurrence sequence and anomaly recovery sequence have a unique data organization form.

[0089] Subsequently, the anomaly analysis module determines the direction of change for each water quality indicator based on the changing trends of the monitoring values ​​corresponding to continuous sampling times. Let the first... Water quality indicators at any time The monitoring value is expressed as The direction of change is represented as:

[0090] ;

[0091] in: Indicates the direction of change; This indicates the monitored value at the current sampling time; This represents the monitored value at the previous sampling time; Represents a symbolic function. When... When the time indicates that the monitored value is in an upward trend; when When the time indicates that the monitored value is in a downward trend; when This indicates that the monitored value remains stable.

[0092] In this scheme, the direction of change is preferably determined based on the overall trend of change over three consecutive sampling periods. When the direction of change remains consistent over three consecutive sampling periods, the corresponding direction is determined as the direction of change for that stage, so as to reduce the impact of single sampling fluctuations on the determination of the direction of change.

[0093] The anomaly process analysis module further divides the anomaly into stages based on the anomaly occurrence time, anomaly recovery time, and direction of change for each water quality indicator. The anomaly stages include the anomaly initiation stage, the anomaly development stage, the anomaly persistence stage, and the anomaly recovery stage. Specifically, the anomaly initiation stage corresponds to the time interval from the anomaly occurrence time to the start of continuous change in the main anomaly indicator; the anomaly development stage corresponds to the continuous change process of the main anomaly indicator; the anomaly persistence stage corresponds to the process of the main anomaly indicator maintaining a stable anomaly state; and the anomaly recovery stage corresponds to the process of the main anomaly indicator recovering to a local baseline state. The module determines the stage transition relationship according to the temporal connection between the above stages and records the transition position between adjacent stages, used to describe the stage evolution process of each water quality indicator within the entire site's anomaly data segment.

[0094] After completing the above processing, the anomaly analysis module organizes the core analysis results into anomaly process tags. Let the anomaly process tags be represented as:

[0095] ;

[0096] in: Indicates an abnormal process marker; Represents the set of times when an anomaly occurred; Represents the set of abnormal recovery moments; Indicates the order in which the exceptions occurred; Indicates the order of abnormal recovery; This represents the set of directions of change for various water quality indicators; This represents the set of stage transition relationships.

[0097] In this scheme, the order of anomalies, the order of anomaly recovery, the direction of change, and the relationship of stage transitions are all stored using a unified data structure. An index relationship is established according to the corresponding number of each water quality indicator, so that the anomaly process marker can fully reflect the temporal and stage change characteristics of each water quality indicator within the anomaly data segment of the same site. This provides a unified data input and structural constraints for the subsequent event correlation analysis module to construct the correlation conservation state, perform candidate anomaly data segment mapping, and match the corresponding relationship.

[0098] Reference Figure 2 Furthermore, the event correlation analysis module performs candidate abnormal data segment filtering based on the station's abnormal data segments and water flow status data, specifically including:

[0099] The water flow transmission time is determined based on the location relationship between adjacent monitoring stations and water flow status data.

[0100] The time interval of the event association is determined based on the start and end times of the abnormal data segments from the upstream monitoring stations and the water flow transmission time.

[0101] Abnormal data segments from downstream monitoring stations located within the event-related time interval were identified as candidate abnormal data segments.

[0102] Furthermore, the event correlation analysis module performs correlation conservation state construction based on the abnormal process markers, specifically including:

[0103] The core and secondary abnormal indicators are determined based on the participation status of each water quality indicator in different abnormal stages.

[0104] A correlation conservation state is constructed based on the sequence of occurrence of anomalies, the sequence of recovery of anomalies, the direction of change, and the stage transition relationship corresponding to the core anomaly indicators.

[0105] Specifically, the event correlation analysis module first receives the abnormal data segments and abnormal process markers from the current monitoring station, as well as the corresponding water flow status data from adjacent monitoring stations. Based on the positional relationship between the monitoring stations, it determines the event correlation direction and calculates the event correlation time interval between adjacent monitoring stations by combining the water flow status data. Let the monitoring stations be... With monitoring stations The water flow transmission time between them is expressed as The start and end times of the abnormal data segments corresponding to the upstream monitoring stations are respectively represented as: and The time interval associated with the event is then represented as:

[0106] ;

[0107] in: Indicates monitoring stations With monitoring stations The time intervals of the events between them; Indicates the water flow transmission time; Indicates the start time of the abnormal data segment at the upstream site; This indicates the end time of the abnormal data segment at the upstream station. In this scheme, the water flow transmission time... The preferred method is to calculate based on the river channel distance between monitoring stations and the real-time average flow velocity. When the water flow status data is updated, the event association time interval is updated synchronously to ensure that the event association analysis always uses the data association range corresponding to the current water flow status.

[0108] The event correlation analysis module filters abnormal data segments from all downstream monitoring stations based on the event correlation time interval. Abnormal data segments whose start or end times fall within the event correlation time interval are identified as candidate abnormal data segments, and a correspondence is established between these candidate abnormal data segments and the event correlation time interval. Each candidate abnormal data segment records the corresponding event correlation time interval number, the associated upstream monitoring station number, and the correlation time information, forming a unified data index to provide corresponding data location relationships for subsequent correlation analysis.

[0109] After screening candidate anomaly data segments, the event correlation analysis module generates a correlation conservation state based on the anomaly process markers. First, it calculates the stage participation status of each water quality indicator based on its participation in the anomaly initiation, development, duration, and recovery stages. When a water quality indicator continuously participates in the anomaly process throughout all stages, it is identified as a core anomaly indicator; when a water quality indicator only participates in some stages, it is identified as a secondary anomaly indicator.

[0110] Let the set of all water quality indicators be represented as: The core set of abnormal indicators is represented as follows: The set of ancillary abnormal indicators is represented as follows: ;in: Represents the set of all abnormal indicators; This represents the set of core abnormal indicators; This represents a set of ancillary abnormal indicators; This indicates the number of core abnormal indicators. In this scheme, when an abnormal indicator participates in all four abnormal stages, it is preferably identified as a core abnormal indicator; when it participates in fewer than four stages, it is identified as a secondary abnormal indicator.

[0111] Subsequently, the event correlation analysis module extracts the anomaly occurrence sequence, anomaly recovery sequence, change direction, and stage transition relationship corresponding to the core anomaly indicators from the anomaly process markers. The anomaly occurrence sequence indicates the order in which multiple core anomaly indicators enter an abnormal state, the anomaly recovery sequence indicates the order in which multiple core anomaly indicators return to normal, the change direction indicates the change trend of the monitoring values ​​corresponding to each core anomaly indicator, and the stage transition relationship describes the transition sequence between the anomaly initiation stage, the anomaly development stage, the anomaly persistence stage, and the anomaly recovery stage.

[0112] The event correlation analysis module organizes the above data into a unified correlation conservation state. Let the correlation conservation state be represented as:

[0113] ;

[0114] in: Indicates a state of correlation conservation; This represents the set of core abnormal indicators; This represents a set of ancillary abnormal indicators; Indicates the order in which the exceptions occurred; Indicates the order of abnormal recovery; Represents the set of directions of change; This represents the set of stage transition relationships.

[0115] In this scheme, the correlation conservation state is stored using a fixed field structure. Each field is indexed to a corresponding core anomaly indicator, ensuring a consistent data organization for the anomaly occurrence order, recovery order, direction of change, and stage transition relationships corresponding to each core anomaly indicator. Subsequently, the event correlation analysis module loads the correlation conservation state into the corresponding candidate anomaly data segment and uses it as a correlation constraint template when performing event correlation analysis on the candidate anomaly data segment. This ensures that subsequent event correlation analysis processes perform corresponding relationship matching around a unified data structure, while auxiliary anomaly indicators participate in correlation analysis as auxiliary constraints and do not participate in the construction of the main structure of the correlation conservation state.

[0116] Through the above processing, the event correlation analysis module completes the generation of event correlation time intervals, the screening of candidate abnormal data segments, and the construction of correlation conservation states. This enables the abnormal processes of different monitoring stations to be described using a unified data organization structure, and establishes the constraint relationship between candidate abnormal data segments and correlation conservation states, providing a unified data foundation for subsequent candidate abnormal data segment mapping, correspondence matching, and generation of inconsistent event correlation states.

[0117] Furthermore, the event correlation analysis module generates an event correlation inconsistency state based on candidate abnormal data segments and correlation conservation states, specifically including:

[0118] Using core anomaly indicators as the main index, candidate anomaly data segments are reorganized according to the data structure of the correlation conservation state to form a candidate correlation structure;

[0119] Match the candidate association structure with the core anomaly indicators, anomaly occurrence order, anomaly recovery order, change direction, and stage transition relationship in the association conservation state;

[0120] Compare the time of occurrence and recovery of the candidate abnormal data segment with the time interval associated with the event;

[0121] Based on the corresponding matching results and time comparison results, generate time-related inconsistency states, indicator-related inconsistency states, and stage structure inconsistency states.

[0122] Specifically, after receiving candidate anomaly data segments and the associated conservation state, the event correlation analysis module first performs a unified structure mapping process on the candidate anomaly data segments. Because the duration of anomaly data segments, the number of participating anomaly indicators, and the length of anomaly stages vary among different monitoring stations, the event correlation analysis module reorganizes the candidate anomaly data segments according to the data organization structure of the associated conservation state, ensuring that the candidate anomaly data segments and the associated conservation state have a consistent data field order and index relationship. During the mapping process, the core anomaly indicator is used as the primary index, and the order of anomaly occurrence, the order of anomaly recovery, the direction of change, and the stage transition relationship are used as association fields to reorganize the data corresponding to the candidate anomaly data segments, forming a candidate association structure. Let the candidate association structure be represented as:

[0123] ;

[0124] in: Indicates the candidate association structure; This represents the set of core anomaly indicators corresponding to the candidate anomaly data segment. Indicates the order in which the anomalies occurred corresponding to the candidate anomaly data segments; Indicates the anomaly recovery order corresponding to the candidate anomaly data segments; This represents the set of change directions corresponding to candidate abnormal data segments; This represents the set of stage transition relationships corresponding to candidate abnormal data segments.

[0125] In this scheme, each data field in the candidate association structure is indexed using a data organization method consistent with the association conservation state. The core anomaly indicators are arranged in ascending order according to the monitoring indicator number. The order of anomaly occurrence and the order of anomaly recovery are represented by sequential numbers, and the direction of change is represented by three states: rising, falling, and stable. The stage transition relationship is represented by the transition number between adjacent stages.

[0126] After completing the structure mapping, the event correlation analysis module performs correspondence matching according to the field correspondence principle. First, it matches the set of core anomaly indicators in the correlation conservation state with the set of core anomaly indicators in the candidate correlation structure item by item. Then, based on the corresponding core anomaly indicators, it compares the order of anomaly occurrence, the order of anomaly recovery, the direction of change, and the stage transition relationship. When the corresponding fields are consistent, they are marked as correlated consistently; when the corresponding fields differ, they are marked as correlated inconsistently. Let the first... The matching result of the corresponding field for the item is represented as follows:

[0127] ;

[0128] in: Indicates the first Matching results of item field correspondence; In the associated conservation state, the first Item corresponding field; This represents the corresponding field in the candidate association structure. In this scheme, the core anomaly indicators adopt a completely consistent matching method; the order of anomaly occurrence and the order of anomaly recovery preferably allow adjustments within one adjacent sequential position; the direction of change must remain consistent; the stage transition relationship is matched by comparing each item with the corresponding stage number.

[0129] The event correlation analysis module generates event correlation inconsistency states based on the matching results of the corresponding relationships of each field. When the overall time of occurrence of the candidate abnormal data segment is outside the event correlation time interval, it is determined to be a time correlation inconsistency state; when the core abnormal indicators in the candidate correlation structure are missing, added, or replaced with the core abnormal indicators in the correlation conservation state, it is determined to be an indicator correlation inconsistency state; when there are inconsistencies in the correspondence between the order of occurrence of abnormalities, the order of recovery of abnormalities, the direction of change, or the stage transition relationship, it is determined to be a stage structure inconsistency state.

[0130] Let the inconsistent state of event association be represented as:

[0131] ;

[0132] in: This indicates an inconsistent state of event association; This indicates a state of inconsistency in time correlation; This indicates an inconsistency in the correlation between indicators; This indicates a state of inconsistency in the stage structure.

[0133] In this scheme, the time-related inconsistency state is represented by a binary state. When the candidate abnormal data segment completely falls within the event-related time interval, the value is 0, otherwise the value is 1. The indicator-related inconsistency state is generated based on the correspondence of core abnormal indicators. The stage structure inconsistency state is determined comprehensively based on the correspondence of the abnormal occurrence order, abnormal recovery order, change direction and stage transition relationship.

[0134] After generating the event-related inconsistency states, the event correlation analysis module establishes a unified data association relationship between the candidate abnormal data segments, the correlation conservation state, and the event-related inconsistency states, and sends this relationship to the reverse correction module. The event-related inconsistency states fully retain the inconsistency type and corresponding location of the relevant fields, enabling the reverse correction module to perform corresponding data processing based on different types of inconsistency states without needing to re-analyze the correlation conservation state. Through this process, a continuous process is achieved, from unified data organization and correspondence matching of candidate abnormal data segments to the generation of event-related inconsistency states, providing unified data input and clear inconsistency types for subsequent classification and reverse correction.

[0135] Reference Figure 3 Furthermore, the reverse correction module performs classified reverse correction based on the time correlation inconsistency and stage structure inconsistency in the event correlation inconsistency states, specifically including:

[0136] Adjust the event association time interval based on the inconsistency in time association and the actual time of anomaly occurrence and recovery of candidate anomaly data segments;

[0137] Based on the inconsistent state of the stage structure, the candidate abnormal data segments are divided into associated abnormal sub-segments and local abnormal sub-segments.

[0138] Establish event correlations based on core anomaly indicators and the order of anomaly occurrence among multiple associated conservation states.

[0139] Furthermore, the reverse correction module performs association state reconstruction based on the corrected event association time interval, associated abnormal sub-segments, and event association relationships, specifically including:

[0140] Retain the core anomaly indicators, anomaly occurrence order, anomaly recovery order, change direction and stage transition relationship corresponding to the associated anomaly sub-segments, and remove the structural information corresponding to the local anomaly sub-segments;

[0141] Based on the retained data and event relationships, the association conservation state is regenerated, and an event association confirmation state is formed.

[0142] Specifically, after receiving the event correlation inconsistency status output by the event correlation analysis module, the reverse correction module first determines the corresponding correction method based on the type of event correlation inconsistency status. When the event correlation inconsistency status is a time correlation inconsistency status, the reverse correction module uses the actual occurrence time and recovery time of the candidate abnormal data segment as the correction basis to readjust the event correlation time interval. Let the original event correlation time interval be represented as... The actual time of occurrence and the time of recovery of the candidate abnormal data segment are respectively represented as: and The corrected event-related time interval is then represented as:

[0143] ;

[0144] in: Indicates the corrected time interval of the event association; Indicates the start time of the time interval associated with the original event; Indicates the end time of the time interval associated with the original event; This indicates the time when the candidate abnormal data segment corresponds to the abnormal occurrence. This indicates the time of anomaly recovery corresponding to the candidate abnormal data segment.

[0145] In this scheme, when the time deviation between the candidate abnormal data segment and the original event-related time interval does not exceed two consecutive sampling periods, the above method is preferred to update the event-related time interval; when the time deviation exceeds the preset range, the original event-related time interval is retained, and the corresponding candidate abnormal data segment is marked as a new candidate related object.

[0146] When the event correlation inconsistency state is a stage structure inconsistency state, the reverse correction module re-divides the correlated abnormal segments and local abnormal segments according to the stage transition relationship. The module first determines the range of data in the candidate abnormal data segment that maintains the correlation conservation state, identifying the data maintaining the correlation conservation state as the correlated abnormal segment; and the data that does not maintain the correlation conservation state as the local abnormal segment. The correlated abnormal segment inherits the data structure corresponding to the original correlation conservation state, while the local abnormal segment separately records the corresponding abnormal occurrence order, abnormal recovery order, change direction, and stage transition relationship, and is retained as independent abnormal data, not participating in the current event correlation analysis process.

[0147] Subsequently, the reverse correction module establishes event correlations based on the correspondence between multiple associated conservation states. When the same candidate anomalous data segment corresponds to multiple associated conservation states, the module compares the core anomaly indicators, anomaly occurrence order, anomaly recovery order, and stage transition relationships in each associated conservation state. When multiple associated conservation states have the same core anomaly indicator and maintain a consistent anomaly occurrence order, a corresponding event correlation is established. Let the event correlation be represented as:

[0148] ;

[0149] in: Represents a set of event relationships; Indicates the first A series of related conservation states; This represents the event correlation established between two correlated conservation states. In this scheme, when the matching ratio of the core anomaly indicators corresponding to the two correlated conservation states is not less than 80%, and the order of anomaly occurrence is consistent, it is preferable to establish an event correlation; otherwise, the two correlated conservation states are kept to exist independently.

[0150] After completing the classification and reverse correction, the reverse correction module regenerates the associated conservation state based on the corrected event association time interval, associated abnormal sub-segments, local abnormal sub-segments, and event association relationships. During the regeneration process, the module retains the core abnormal indicators, abnormal occurrence order, abnormal recovery order, change direction, and stage transition relationships corresponding to the associated abnormal sub-segments. At the same time, it removes the structural information corresponding to the local abnormal sub-segments and updates the association numbers between the associated conservation states in conjunction with the event association relationships, so that the regenerated associated conservation state can reflect the overall abnormal structure after the current event association.

[0151] Let the regenerated associated conservation state be represented as:

[0152] ;

[0153] in: This represents the regenerated state of the associated conservation state; Represents the associated conserved state generating function; Indicates the corrected time interval of the event association; Indicates associated abnormal sub-fields; This represents a set of event relationships. In this scheme, the regenerated relationship conservation state preferably maintains the same data field structure as the original relationship conservation state, only updating the anomaly occurrence order, anomaly recovery order, change direction, and stage transition relationship corresponding to the core anomaly indicators, ensuring that the subsequent relationship analysis process adopts a unified data organization form.

[0154] Finally, the reverse correction module generates an event association confirmation status based on the regenerated association conservation status. The event association confirmation status includes the corrected event association time interval, the regenerated association conservation status, the association anomaly segment identifier, the local anomaly segment identifier, and the event association relationship identifier. After establishing a unique data index according to the monitoring station number, it is sent to the association status update module.

[0155] Through the above processing, the reverse correction module performs event association time interval adjustment, abnormal segment division and event association relationship establishment according to different types of inconsistent event association states. It also uses the correction results to regenerate the association conservation state and event association confirmation state, so that the event association analysis results can maintain a unified data organization structure, providing a consistent data foundation for subsequent association state updates and cross-site abnormal event data management.

[0156] Furthermore, the association state update module performs association conservation state updates based on the event association confirmation status, specifically including:

[0157] Extract the core abnormal indicators that are common to adjacent monitoring stations;

[0158] The order of anomalies and the order of recovery are updated based on the anomaly occurrence time and recovery time corresponding to the common core anomaly indicators.

[0159] Based on the updated core anomaly indicators, the order of anomaly occurrence, and the order of anomaly recovery, the associated conservation state corresponding to the subsequent monitoring stations is regenerated.

[0160] Furthermore, the event data management module performs cross-site abnormal event data management based on the confirmation status of multiple event associations, specifically including:

[0161] Based on the event association relationships in the event association confirmation status, site abnormal data segments with continuous event association relationships are combined to form cross-site abnormal event data groups;

[0162] The cross-site abnormal event data group to be updated is determined based on the monitoring station and associated event number corresponding to the newly added water quality monitoring data, and the cross-site abnormal event data group is updated according to the updated event association confirmation status.

[0163] Specifically, after receiving the event association confirmation status output by the reverse correction module, the association status update module first reads the association conservation status corresponding to the current monitoring station and the next monitoring station. Based on the correspondence of core abnormal indicators in the event association confirmation status, it extracts the core abnormal indicators that exist in both adjacent monitoring stations. For core abnormal indicators that only exist in the current monitoring station and do not appear in the next monitoring station, they are marked as terminated association indicators. For abnormal indicators newly added to the next monitoring station, they are retained in the set of auxiliary abnormal indicators of the current monitoring station and do not participate in the main update process of the association conservation status, thus ensuring that the association conservation status always uses continuously associated core abnormal indicators as the main data structure.

[0164] Subsequently, the association status update module reorders the order of occurrence and recovery of common core anomaly indicators based on the anomaly occurrence and recovery times recorded in the event association confirmation status, and updates the corresponding change directions and stage transition relationships. When the occurrence times of multiple core anomaly indicators change, the module regenerates the anomaly occurrence order according to the occurrence times; when the recovery times of multiple core anomaly indicators change, it regenerates the anomaly recovery order according to the recovery times. The updated anomaly occurrence and recovery orders continue to maintain a one-to-one correspondence with the core anomaly indicator numbers to ensure the data structure remains consistent for subsequent event association analysis.

[0165] Let the updated association conservation state be represented as:

[0166] ;

[0167] in: This represents the updated association conservation state; This represents the updated set of core anomaly indicators; This represents the updated set of ancillary anomaly indicators; Indicates the order in which exceptions occurred after the update; Indicates the order of exception recovery after the update; Represents the set of updated directions of change; This represents the updated set of stage transition relationships. In this scheme, the number of updated core anomaly indicators is preferably no less than two; when the number of common core anomaly indicators is less than two, the recursive update of the current association conservation state is terminated, and a new association conservation state is re-established.

[0168] After completing the correlation conservation state update, the correlation state update module sends the updated correlation conservation state to the event correlation analysis module according to the monitoring station number. This serves as the correlation constraint template for the next monitoring station when performing candidate abnormal data segment mapping and corresponding relationship matching. When processing the candidate abnormal data segment corresponding to the next monitoring station, the event correlation analysis module directly calls the corresponding correlation conservation state to perform event correlation analysis, without needing to rebuild the correlation structure corresponding to the previous monitoring station. This ensures that a unified data organization method is used to complete the correlation analysis across consecutive monitoring stations.

[0169] After receiving event association confirmation statuses from multiple monitoring stations, the event data management module establishes cross-site abnormal event data groups based on the event association relationships in the confirmation statuses. The module first combines abnormal data segments from stations with consecutive event associations based on the association numbers in the confirmation statuses, and then establishes a unified data organization structure according to the monitoring station numbers, event association time intervals, and association conservation states recorded in the event association confirmation statuses. Let the cross-site abnormal event data group be represented as:

[0170] ;

[0171] in: Indicates a cross-site exception event data group; This represents a collection of abnormal data segments from the site. Represents the set of event-related confirmation statuses; This represents the set of associated conservation states. In this scheme, a cross-site abnormal event data group preferably contains event association confirmation states corresponding to two or more monitoring sites. When there are event association relationships between consecutive monitoring sites, the abnormal data segments of the corresponding sites are divided into the same cross-site abnormal event data group.

[0172] The event data management module establishes a site anomaly data segment library, an event association confirmation status library, and a cross-site anomaly event data group library, and establishes a data index relationship between the three based on a unified event number. Specifically, the site anomaly data segment stores the original anomaly process data of each monitoring site; the event association confirmation status stores the corresponding association relationship and the updated association conservation status; and the cross-site anomaly event data group stores the association results of all monitoring sites corresponding to the same anomaly event.

[0173] When new water quality monitoring data is uploaded to the cloud platform, the event data management module first determines the monitoring station and associated event number corresponding to the new water quality monitoring data. Then, it calls the data acquisition module to regenerate the corresponding abnormal data segment. Subsequently, it calls the event association analysis module, reverse correction module, and association status update module to regenerate the association conservation status and event association confirmation status. Based on the updated association conservation status, it synchronously updates the corresponding cross-site abnormal event data group, while the unaffected cross-site abnormal event data group retains its original data content.

[0174] Through the above processing, the association status update module realizes the recursive update of the association conservation status between continuous monitoring stations, so that subsequent event association analysis of monitoring stations always adopts a unified data constraint structure; the event data management module completes the data organization, indexing and updating between site abnormal data segments, event association confirmation status and cross-site abnormal event data groups, so that the entire cloud platform forms a complete cross-site abnormal event data management process, providing a unified data organization foundation for subsequent abnormal event query, association analysis and historical event reuse.

[0175] Example 2: In this example, the invention is illustrated using a multi-site monitoring system consisting of the main stream and two tributaries of a river basin. The monitoring system includes an upstream monitoring station A, a tributary monitoring station B, a confluence monitoring station C, and a downstream monitoring station D. Each monitoring station uploads water quality monitoring data, station operation status data, and water flow status data to the cloud platform according to a fixed sampling cycle.

[0176] The data acquisition module receives data uploaded by each monitoring station, filters valid water quality monitoring data based on station operational status data, and establishes local baseline states based on historical stable monitoring data for each station. When monitoring station A and monitoring station B detect continuous abnormal changes, the data acquisition module generates corresponding abnormal data segments for each station and sends them to the abnormal process analysis module.

[0177] The anomaly process analysis module analyzes the anomaly data segments corresponding to monitoring stations A and B, extracting the anomaly occurrence time, anomaly recovery time, direction of change, and stage transition relationship for each water quality indicator, forming corresponding anomaly process markers. The event correlation analysis module generates event correlation time intervals corresponding to monitoring stations A to C, B to C, and C to D based on the location relationship between monitoring stations and water flow state data, and constructs correlation conservation states for each.

[0178] Since the confluence region simultaneously receives data from two tributaries, the candidate abnormal data segment corresponding to monitoring station C is simultaneously mapped to two sets of associated conservation states. The event correlation analysis module performs corresponding relationship matching to generate corresponding event correlation inconsistency states; the reverse correction module performs event correlation time interval adjustment, correlation abnormal sub-segment division, and event correlation relationship establishment according to different event correlation inconsistency states to generate corresponding event correlation confirmation states. The correlation state update module regenerates the correlation conservation state corresponding to monitoring station D based on the event correlation confirmation state, and uses the updated correlation conservation state as the correlation constraint template for event correlation analysis at monitoring station D.

[0179] The event data management module establishes cross-site abnormal event data groups based on the event association confirmation status of monitoring sites A, B, C, and D, and records the abnormal data segments, association conservation status, and event association relationships for each site. When new monitoring data is uploaded, only the corresponding event association confirmation status and cross-site abnormal event data groups are updated; the remaining unaffected event data remain unchanged.

[0180] Table 1 shows a comparison between this scheme and the traditional multi-site water quality monitoring data management method.

[0181] Comparison Projects Traditional data management methods This invention Cross-site event association methods Directly associate abnormal data based on time windows Execute event association based on the associated time interval and the association conservation state. abnormal event description method Describe the abnormal data independently Exception process labeling is used to describe the sequence of exception occurrence, recovery sequence, direction of change, and stage transition relationships. Candidate anomaly data analysis methods Directly compare abnormal data Map candidate outlier data segments to the correlation conservation state and perform corresponding relationship matching. Abnormal association correction method Unified correction of correlation results Perform classification-based reverse correction based on inconsistent correlations between different events. Multi-site associated update method Each time, all relationships are re-established The association conservation state is updated recursively based on the event association confirmation status. Data management methods Independently store abnormal data from each monitoring station Establish a unified management system for cross-site anomaly event data groups.

[0182] Through the above embodiments, abnormal events corresponding to multiple monitoring stations are analyzed by adopting a unified data organization structure. The event association confirmation status is used as input data for the association status update module to regenerate the association conservation status corresponding to subsequent monitoring stations, and further used as the association constraint template for the event association analysis of the next monitoring station, so as to realize the recursive update of the association status between continuous monitoring stations. The event data management module constructs cross-site abnormal event data groups based on the event association confirmation status, so as to realize the unified organization and data management of abnormal events among multiple monitoring stations.

[0183] Finally, it should be noted that the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments or make equivalent substitutions for some of the technical features. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A multi-site water quality monitoring data management system based on a cloud platform, characterized in that: include: Multiple water quality monitoring stations and a cloud platform that communicates with the multiple water quality monitoring stations; The cloud platform includes a data acquisition module, an abnormal process analysis module, an event correlation analysis module, a reverse correction module, a correlation status update module, and an event data management module. The data acquisition module generates abnormal data segments for the stations based on water quality monitoring data and station operation status data. The abnormal process analysis module determines the order of occurrence, recovery order, direction of change, and stage transition relationship of each water quality indicator based on the abnormal data segment of the station, and generates an abnormal process marker. The event correlation analysis module determines the event correlation time interval and filters candidate abnormal data segments based on the abnormal data segments and water flow status data at the site. It constructs a correlation conservation state based on the abnormal process marker, maps the candidate abnormal data segments to the correlation conservation state, performs corresponding relationship matching, and generates an event correlation inconsistency state. The reverse correction module performs classification correction on the event association time interval and candidate abnormal data segment according to the event association inconsistency status, and generates an event association confirmation status based on the correction result; The association status update module updates the association conservation status according to the event association confirmation status, and uses the updated association conservation status as the association constraint for subsequent event association analysis at monitoring stations. The event data management module generates cross-site abnormal event data groups based on the associated confirmation status of multiple events.

2. The multi-site water quality monitoring data management system based on a cloud platform according to claim 1, characterized in that, The data acquisition module generates abnormal data segments for the site based on the site's operational status data and the water quality monitoring data, specifically including: Valid water quality monitoring data are selected based on equipment operating status, communication status, sampling status, and time synchronization status. A local baseline state is established based on historical stable monitoring data. The effective water quality monitoring data is compared with the local baseline state, and abnormal data segments of the stations are generated based on the deviation state within multiple consecutive sampling periods.

3. The multi-site water quality monitoring data management system based on a cloud platform according to claim 2, characterized in that, The anomaly analysis module generates anomaly process markers based on the site's anomaly data segments, specifically including: Based on the sampling times when each water quality indicator continuously meets the abnormal and recovery conditions, the time of abnormal occurrence and the time of abnormal recovery are determined. The sequence of occurrence, sequence of recovery, and phase transition relationship of the anomaly are determined based on the time of occurrence, the time of recovery, and the direction of continuous change, and an anomaly process marker is generated.

4. The multi-site water quality monitoring data management system based on a cloud platform according to claim 3, characterized in that, The event correlation analysis module performs candidate abnormal data segment filtering based on the station abnormal data segments and the water flow status data, specifically including: The water flow transmission time is determined based on the location relationship between adjacent monitoring stations and water flow status data. The event-related time interval is determined based on the start and end times of the abnormal data segments from the upstream monitoring stations and the water flow transmission time. Abnormal data segments from downstream monitoring stations located within the event-related time interval are identified as candidate abnormal data segments.

5. The multi-site water quality monitoring data management system based on a cloud platform according to claim 4, characterized in that, The event correlation analysis module performs correlation conservation state construction based on the abnormal process markers, specifically including: The core and secondary abnormal indicators are determined based on the participation status of each water quality indicator in different abnormal stages. Based on the anomaly occurrence sequence, anomaly recovery sequence, change direction, and stage transition relationship corresponding to the core anomaly indicators, an associated conservation state is constructed.

6. The multi-site water quality monitoring data management system based on a cloud platform according to claim 5, characterized in that, The event correlation analysis module generates an event correlation inconsistency state based on the candidate abnormal data segments and the correlation conservation state, specifically including: Using the core anomaly indicators as the main index, the candidate anomaly data segments are reorganized according to the data structure of the aforementioned correlation conservation state to form a candidate correlation structure; The candidate association structure is matched with the core anomaly indicators, anomaly occurrence order, anomaly recovery order, change direction, and stage transition relationship in the association conservation state. The time of occurrence and recovery of the anomaly in the candidate anomaly data segment are compared with the time interval associated with the event. Based on the corresponding matching results and time comparison results, generate time-related inconsistency states, indicator-related inconsistency states, and stage structure inconsistency states.

7. The multi-site water quality monitoring data management system based on a cloud platform according to claim 1, characterized in that, The reverse correction module performs classification reverse correction based on the time association inconsistency and stage structure inconsistency in the event association inconsistency states, specifically including: The event association time interval is adjusted based on the inconsistent time association state and the actual time of anomaly occurrence and recovery of candidate abnormal data segments; Based on the inconsistent state of the stage structure, the candidate abnormal data segment is divided into associated abnormal sub-segment and local abnormal sub-segment; Establish event correlations based on core anomaly indicators and the order of anomaly occurrence among multiple associated conservation states.

8. The multi-site water quality monitoring data management system based on a cloud platform according to claim 7, characterized in that, The reverse correction module performs association state reconstruction based on the corrected event association time interval, the associated abnormal sub-segment, and the event association relationship, specifically including: The core anomaly indicators, anomaly occurrence order, anomaly recovery order, change direction, and stage transition relationship corresponding to the associated anomaly sub-segments are retained, while the structural information corresponding to the local anomaly sub-segments is removed. Based on the retained data and the event relationships, the association conservation state is regenerated, and an event association confirmation state is formed.

9. The multi-site water quality monitoring data management system based on a cloud platform according to claim 8, characterized in that, The association state update module performs an association conservation state update based on the event association confirmation state, specifically including: Extract the core abnormal indicators that are common to adjacent monitoring stations; The order of anomalies and the order of recovery are updated based on the anomaly occurrence time and the anomaly recovery time corresponding to the common core anomaly indicators. Based on the updated core anomaly indicators, the order of anomaly occurrence, and the order of anomaly recovery, the associated conservation state corresponding to the subsequent monitoring stations is regenerated.

10. The multi-site water quality monitoring data management system based on a cloud platform according to claim 9, characterized in that, The event data management module performs cross-site abnormal event data management based on the confirmation status of multiple event associations, specifically including: Based on the event association relationships in the event association confirmation status, site abnormal data segments with continuous event association relationships are combined to form cross-site abnormal event data groups; The cross-site abnormal event data group to be updated is determined based on the monitoring station and associated event number corresponding to the newly added water quality monitoring data, and the cross-site abnormal event data group is updated according to the updated event association confirmation status.

Images