A long time series environmental monitoring data storage and processing method

Abstract

The application discloses a long-time sequence environmental monitoring data storage and processing method, relates to the technical field of environmental monitoring data processing, and is suitable for a long-time continuous water environmental monitoring scene of a fixed section automatic monitoring station. Without increasing additional monitoring hardware, in view of the problems that single equipment disturbance, single real water body change and a mixture of the two in a turbidity deviation process in a device action adjacent period are difficult to distinguish, and abnormal entry and removal have time asymmetry, based on turbidity measurement records, auxiliary water quality index measurement records and device action records, the starting boundary and the ending boundary of the confused time section are stably determined, and the section is separately saved from a conventional abnormal processing chain, so that accurate and consistent section basis is provided for subsequent storage, statistics and processing.

Description

Technical Field

[0001] This invention relates to the field of environmental monitoring data processing technology, and more specifically, to a method for storing and processing long-term environmental monitoring data. Background Technology

[0002] Environmental monitoring data processing technology is commonly used for automatic monitoring of the water environment. In this scenario, fixed-section automatic monitoring stations typically monitor turbidity and auxiliary water quality indicators such as conductivity and dissolved oxygen continuously, while simultaneously recording equipment action information such as maintenance, calibration, and malfunctions. For long-term continuous monitoring sequences, existing processing methods usually first identify the anomaly trigger location based on thresholds, amplitude of change, or rate of change. Then, they expand along the time axis around the trigger location to determine the corresponding abnormal time segment. Combining equipment action information or human experience, the abnormal segment is retained, removed, or verified. Finally, the processing results are written into the long-term sequence processing results for subsequent storage, statistics, and analysis.

[0003] The existing technology has the following shortcomings: On the one hand, under the condition of proximity to equipment operation, a single equipment disturbance, a single real water body change, and a mixture of both may all manifest as turbidity deviation. Relying solely on a single-point threshold, the magnitude of change, or the rate of change to identify the location of the abnormality trigger can easily lead to the confusion of deviation processes from different sources, making it difficult to reliably distinguish the source of the abnormality.

[0004] On the other hand, existing methods typically expand abnormal sections symmetrically around the trigger location according to a uniform rule. However, when the equipment disturbance occurs too early and the auxiliary water quality indicator linkage is released too late, the left background section is easily incorporated prematurely, and the right tail section is easily cut off prematurely before the auxiliary water quality indicator linkage ends. At the same time, subsequent treatment lacks a dedicated separation outlet for such sections, so the section boundary and treatment outlet are unstable. Summary of the Invention

[0005] To overcome the aforementioned deficiencies of the prior art, embodiments of the present invention provide a method for storing and processing long-term environmental monitoring data to address the problems mentioned in the background art.

[0006] To achieve the above objectives, the present invention provides the following technical solution: A method for storing and processing long-term environmental monitoring data includes: S101, Read the monitoring records and equipment action records within the target time period, and mark unusable monitoring points based on the equipment action records. The monitoring records include at least the turbidity measurement records and the auxiliary water quality index measurement records. S102, for each time point to be judged in the monitoring records after excluding unusable monitoring points, read the continuous historical segment that meets the stability conditions in front of it as the leading stable segment, and identify the local deviation starting point based on the deviation of the current turbidity relative to the leading stable segment and the rate of change relative to the previous effective monitoring point. S103, determine the entry and release conditions around each local deviation starting point and determine the confusion time segment. Specifically, search backward from the local deviation starting point for the first monitoring point that meets the condition that the time interval between the deviation and the equipment operation time does not exceed the adjacent range and the corresponding turbidity meets the equipment influence morphology condition, and designate it as the initial boundary candidate. When the change of at least one auxiliary water quality indicator reaches the auxiliary change threshold within the allowable lag time from the initial boundary candidate, and the turbidity does not meet the fallback condition again within the subsequent consecutive preset number of valid monitoring points, the entry condition is determined to be met, and the initial boundary candidate is determined to be the initial boundary. When the turbidity meets the fallback condition again and the change of at least one auxiliary water quality indicator falls back to within the auxiliary fallback threshold, the release condition is determined to be met. Search backward from the local deviation starting point for the first monitoring point that meets the release condition, and designate it as the end boundary candidate. Read the consecutive preset number of valid monitoring points backward from the end boundary candidate. When the entry condition is not met again within the consecutive preset number of valid monitoring points, the end boundary candidate is determined to be the end boundary. The segment between the initial boundary and the end boundary is determined as the confusion time segment.

[0007] In a preferred embodiment, in S101, the action type, action start time, and action end time in the device action record are read to generate the corresponding action influence interval; when the device action record contains only a single registration time, it is expanded into an action influence interval according to the preset influence duration of the corresponding action type; sampling points that fall into the maintenance action influence interval, calibration action influence interval, or fault action influence interval are marked as unavailable monitoring points.

[0008] In a preferred embodiment, in S102, effective turbidity monitoring points that reach the number of leading segment length parameters are read forward around each time to be determined to form a candidate segment of the leading stable segment; the turbidity fluctuation range and the adjacent absolute change rate between any two adjacent effective turbidity monitoring points are calculated within the candidate segment of the leading stable segment; when the turbidity fluctuation range does not exceed the stable fluctuation threshold and each adjacent absolute change rate does not exceed the stable rate threshold, the candidate segment of the leading stable segment is determined as the leading stable segment.

[0009] In a preferred embodiment, in S102, the absolute value of the difference between the measured turbidity value at the current time to be judged and the background representative value of the corresponding leading stable segment is extracted as the deviation amount, and the ratio of the absolute value of the difference between the measured turbidity value at the current time to be judged and the measured turbidity value at the previous effective turbidity monitoring point to the time interval between the two is extracted as the rate of change; when the deviation amount reaches the deviation threshold or the rate of change reaches the rate threshold, the current time to be judged is identified as a candidate for the local deviation starting point.

[0010] In a preferred embodiment, in S102, effective turbidity monitoring points that reach the number of subsequent segment length parameters are read backward from the candidate for local deviation starting point to form a subsequent continuation segment; when the absolute deviation of each effective turbidity monitoring point in the subsequent continuation segment relative to the corresponding leading stable segment background representative value remains outside the fallback threshold, the current candidate for local deviation starting point is determined as the local deviation starting point; the continuous deviation segment from the local deviation starting point to the subsequent first formation of the fallback condition is regarded as the same continuous deviation process, and even if the deviation amount trigger condition or the change rate trigger condition is met again at the subsequent judgment time in the same continuous deviation process, no new local deviation starting point will be generated.

[0011] In a preferred embodiment, in S103, candidate monitoring points are searched backward along time around the local deviation starting point; when the time interval between the candidate monitoring point and the device action time falls into the vicinity range, and the corresponding turbidity meets at least one type of device influence morphology condition, the candidate monitoring point is determined as the starting boundary candidate; the device influence morphology condition includes abrupt entry morphology, short-term offset morphology, and short-segment distortion morphology, wherein, the abrupt entry morphology is when the absolute change rate of the current monitoring point relative to the preceding effective turbidity monitoring point reaches the device transition rate threshold and the absolute deviation of the current monitoring point relative to the background representative value of the corresponding leading stable segment reaches the device transition deviation threshold; the short-term offset morphology is when, after continuously reading effective turbidity monitoring points up to the offset duration parameter number from the current monitoring point, the absolute deviation of each effective turbidity monitoring point relative to the background representative value of the corresponding leading stable segment within the observation segment all reaches the offset amplitude threshold; the short-segment distortion morphology is when, within the local observation segment corresponding to the distortion observation length parameter, the absolute change rate between at least two adjacent effective turbidity monitoring points reaches the distortion rate threshold and the number of turbidity deviation direction switching reaches the distortion switching number threshold.

[0012] In a preferred embodiment, in S103, consecutive effective turbidity monitoring points are read backward around the candidate starting boundary to reach the number of continuous observation length parameters to form a continuous observation section; when at least one auxiliary water quality indicator that has been enabled and has available quality markers reaches the auxiliary change threshold within the allowable lag time, and the absolute deviation of each effective turbidity monitoring point in the continuous observation section from the corresponding leading stable section background representative value remains outside the fallback threshold, the candidate starting boundary is determined as the starting boundary.

[0013] In a preferred embodiment, in S103, monitoring points are searched in the forward time direction; when the deviation of the turbidity of the current monitoring point relative to the background representative value of the corresponding leading stable segment falls back to within the fallback threshold, and the change of at least one auxiliary water quality indicator that has been enabled and has an available quality label falls back to within the auxiliary fallback threshold, the current monitoring point is determined as a candidate for the end boundary.

[0014] In a preferred embodiment, in S103, consecutive effective turbidity monitoring points that reach the backward observation length parameter number are read backward from the end boundary candidate to form a backward observation window; when the entry condition is not met again in the backward observation window, the current end boundary candidate is determined as the end boundary; when the entry condition is met again in the backward observation window, a new end boundary candidate is searched along the forward time direction.

[0015] In a preferred embodiment, the method further includes S104, in which, for each obfuscated time segment, the abnormal processing of directly retaining or directly removing the obfuscated time segment is blocked, and the start and end positions of the obfuscated time segment and the corresponding monitoring records are written into the obfuscated segment record in the long-term series processing result; wherein, the obfuscated time segment is written with the segment-level dedicated exit status and the record-level dedicated exit status of the corresponding monitoring record within the segment, and the obfuscated segment record is constructed. The obfuscated segment record includes at least the obfuscated segment number, the start boundary time, the end boundary time, the corresponding local deviation from the start point time, the segment length, and the corresponding monitoring record within the segment.

[0016] The effects and advantages of the long-term environmental monitoring data storage and processing method of the present invention are as follows: This invention provides a method for storing and processing long-term environmental monitoring data. By processing unavailable monitoring points and monitoring points adjacent to equipment actions within the action influence interval separately, and constructing deviation entry anchor points based on leading stable segments and local deviation starting points, this invention can establish a more stable segment entry basis for turbidity deviation processes under equipment action proximity conditions without adding additional monitoring hardware. Furthermore, by performing asymmetric boundary confirmation around the local deviation starting point using forward equipment disturbance evidence and backward linkage fallback evidence, it can suppress premature left-side merging and premature right-side truncation caused by uniform symmetrical expansion around a single trigger position. Moreover, by blocking the conventional anomaly processing chain for the confirmed confusing time segments and writing it into a special record, the confusing segments can be separated and saved separately from the ordinary anomaly processing results, providing a consistent segment basis for the subsequent storage, statistics, and processing of long-term environmental monitoring data. Attached Figure Description

[0017] Figure 1 This is a schematic diagram of the method flow of the present invention; Figure 2 A schematic diagram illustrating the process for determining the confusion time segment under the constraint of local deviation from the starting point. Detailed Implementation

[0018] The technical solutions of 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.

[0019] This invention provides a method for storing and processing long-term environmental monitoring data. It is applicable to automatic water environment monitoring scenarios where fixed-section automatic monitoring stations continuously monitor turbidity and auxiliary water quality indicators such as conductivity and dissolved oxygen over long periods, while simultaneously recording equipment operation information such as maintenance, calibration, and malfunctions. Identifying local anomaly segments within the long-term continuous monitoring sequence is fundamental to subsequent sequence processing. Existing methods typically first identify the anomaly trigger location based on thresholds, amplitude, or rate of change, and then expand forward and backward around the trigger location to determine the anomaly time segment. For anomalies with a single source and relatively clear entry and resolution processes, basic segment determination is usually achievable.

[0020] However, when turbidity anomalies occur in the period adjacent to the equipment operation, the equipment operation may change the sampling conditions, measurement conditions, or instrument status in a local period. The actual water body changes are often accompanied by delayed linkage of auxiliary water quality indicators, which makes the monitoring records show mutual confusion between single equipment disturbance, single actual water body change, and the mixed process of the two. Moreover, the actual confusion process often shows that the equipment disturbance enters earlier and the auxiliary linkage is released later, resulting in an asymmetry between the abnormal entry stage and the abnormal release stage on the time axis. If the existing method still expands symmetrically around the trigger position according to a uniform rule, it is easy to prematurely incorporate the background segment on the left that has not yet entered the confusion state, and it is also easy to prematurely truncate the abnormal tail segment before the auxiliary linkage on the right has ended, thus making the positioning of the confusion time segment unstable.

[0021] Therefore, without adding additional monitoring hardware, and only using existing turbidity measurement records, auxiliary water quality index measurement records, and equipment operation records, this invention stably determines the start and end boundaries of the confusion time segment around the turbidity deviation process under the condition of equipment operation proximity, and separates the confusion time segment from the conventional anomaly handling chain.

[0022] Based on the above design, this invention constructs a complete process for storing and processing long-term environmental monitoring data, consisting of steps S101 to S104 sequentially. (Refer to...) Figure 1 , Figure 1 This is a schematic diagram of the method flow of the present invention, which includes: Step S101, Monitoring Record Organizing and Basic Monitoring Sequence Generation, is used to read, time-align, attach actions, write status, and mark unusable points from the monitoring record input information X101 and the equipment action record input information X102 within the target time period, generating the basic monitoring sequence R101 required for subsequent local deviation start point identification and confusion time segment determination. This step takes the monitoring record input information X101 and the equipment action record input information X102 as inputs and outputs the basic monitoring sequence R101. The basic monitoring sequence R101 records the measured turbidity value, measured auxiliary water quality index value, action type, action impact marker, unusable status, and quality marker corresponding to each sampling time, for direct reading in subsequent steps. The target time period refers to the start and end time interval corresponding to this long-term series processing. Unusable monitoring points refer to sampling points that are written as unusable on a unified time axis. Quality markers refer to record markers used to characterize the missing monitoring value status or predefined abnormal status identifier of each sampling point.

[0023] Step S102 generates local deviation starting points under a leading stable reference. This step uses the basic monitoring sequence R101 as input, reads previous stable historical segments around each time point to be judged as leading stable segments, and identifies local deviation starting points based on the deviation of the current turbidity relative to the leading stable segment and the rate of change of the current turbidity relative to previous effective turbidity monitoring points. Short a posteriori observation is then used to filter candidate points that experience brief declines, generating a local deviation starting point result set R102. This step takes the basic monitoring sequence R101 as input and outputs the local deviation starting point result set R102. The local deviation starting point result set R102 records the local deviation starting points identified and retained in this step, for subsequent steps to determine the confusion time interval around each retained starting point. The time point to be judged refers to the current sampling time from the monitoring point that has completed the unusable status marking in R101 and enters this step for point-by-point judgment. The leading stable segment refers to the historical reference segment formed by continuous effective turbidity monitoring points before the time point to be judged. The local deviation starting point refers to the sampling point identified under the leading stable reference and retained after subsequent filtering.

[0024] Step S103, determining the obfuscated time segment under the constraint of the local deviation starting point, is used to confirm the starting boundary that meets the entry condition forward and the ending boundary that meets the release condition backward around each local deviation starting point, and to verify the state stability after the ending boundary in combination with the backward observation window, generating the obfuscated time segment result set R103. This step takes the local deviation starting point result set R102 and the basic monitoring sequence R101 as input and outputs the obfuscated time segment result set R103. The obfuscated time segment result set R103 is used to record the obfuscated time segments established after entry confirmation, release confirmation, and backward observation verification, for subsequent steps to perform obfuscated segment record writing and processing exit control. The local deviation starting point refers to the reserved starting point from the local deviation starting point result set R102 generated in step S102. The starting boundary candidate refers to the candidate starting position formed when searching backward along time around the local deviation starting point. The ending boundary candidate refers to the candidate ending position formed when searching forward along time around the local deviation starting point.

[0025] Step S104, "Dedicated Exit Switching and Confusion Segment Record Writing for Confusion Time Segments," takes the confusion time segment result set R103 as input and performs dedicated exit switching, confusion segment record construction, and result writing around each confusion time segment to generate confusion segment record R104 and update the corresponding exit status in the long-term series processing results. This step takes the confusion time segment result set R103 as input and outputs confusion segment record R104. Confusion segment record R104 records the boundary position, corresponding local deviation start point, and corresponding monitoring record within the segment for subsequent manual review, specialized processing, or statistical analysis. A confusion time segment refers to a confirmed segment from the confusion time segment result set R103 generated in step S103. Confusion segment record R104 is the result record object generated in this step, used to contain the segment identifier, boundary position, corresponding local deviation start point, and corresponding monitoring record within each confusion time segment. Long-term series processing results refer to the total result object formed after anomaly identification, segment determination and result output organization of monitoring records within the target time period. It is used to accept the results of routine anomaly processing and the results of confused segment records.

[0026] The implementation process and operational effects of the method of the present invention will be described in detail below with reference to specific embodiments. It should be understood that the embodiments are only used to illustrate the technical solution of the present invention, and not to limit it. The relevant steps, parameters, and module divisions can be appropriately adjusted without changing the essence of the invention.

[0027] In one optional implementation, step S101 includes reading monitoring records, generating equipment action intervals, establishing a unified timeline and aligning monitoring items, attaching equipment actions, and identifying unavailable monitoring points and generating basic monitoring sequences.

[0028] During the monitoring record reading process, environmental monitoring data organized by sampling time in X101 are read within the target time period to obtain the basic monitoring record. The monitoring record includes at least turbidity measurement records and auxiliary water quality indicator measurement records. The auxiliary water quality indicator measurement records preferably include conductivity and dissolved oxygen measurement records; other monitoring indicator measurement records that reflect natural water body changes and have distinguishable response characteristics from equipment disturbances can also be used. For null values, missing values, or predefined abnormal state markers in X101, this step retains the original record and completes the state transition through quality marking.

[0029] In generating equipment action intervals, equipment action records in X102 within the target time period are read and converted into action impact intervals that can be used to attach sampling points. Equipment action records preferably include action type, action start time, and action end time. The action type distinguishes between maintenance, calibration, and fault actions, while the action start and end times characterize the corresponding action impact intervals. For original equipment action records that already contain action start and end times, the action impact interval is directly determined using the original records. For records containing only a single registration time, it is preferable to expand the action impact interval based on the preset impact duration corresponding to that action type. The preset impact duration is preferably pre-configured as a fixed duration for each action type and determined or corrected based on the equipment operation and maintenance manual, site maintenance procedures, or historical action impact statistics. For action types that have not yet been configured with a preset impact duration, the registration time can be used as the action start and end times.

[0030] In establishing a unified time axis and aligning monitoring items, the standard sampling time output by the monitoring system is used as the unified time axis, and the records of each monitoring item in X101 are organized under the same time reference. The standard sampling time refers to the discrete sampling time point formed by the automatic monitoring station according to the preset sampling period. Based on this unified time axis, the measured turbidity records and the measured records of each auxiliary water quality index are prioritized to be directly aligned according to the same sampling time. When there is a sampling time offset between a certain monitoring item record and the unified sampling time that does not exceed the preset time tolerance range, it is treated as a mergeable record and merged into the corresponding unified sampling time. A mergeable record refers to a monitoring record whose sampling time offset from the unified sampling time is within the preset time tolerance range and is suitable for merging into the unified sampling time. The preset time tolerance range is preferably no greater than half of the corresponding sampling period and less than the sampling interval between two adjacent standard sampling times. It can also be configured based on the sampling synchronization error range of the monitoring station or the statistical results of historical asynchronous sampling offsets. When there are mergeable records for the same monitoring item before and after the same unified sampling time, the record with the smaller time distance is selected first. If the time distance is the same, the record that appears earlier is selected first. After merging, the status of the corresponding monitoring item is transferred to the quality label through the missing measurement marker. Subsequent determinations involving the linkage of multiple monitoring items will be based on the merging results under a unified timeline.

[0031] In the device action mapping process, the action impact interval corresponding to X102 is mapped to various sampling moments on a unified time axis. Sampling moments falling within an action impact interval are written with the corresponding action type and action impact flag. For sampling moments falling within multiple action impact intervals, the action type with the highest priority is written according to the preset action priority, forming the corresponding action impact flag. The preset action priority is preferably set as fault over calibration, and calibration over maintenance. The remaining sampling moments continue to have their corresponding action impact flags written around the unified time axis.

[0032] In the process of identifying unavailable monitoring points and generating basic monitoring sequences, equipment action status identification and result writing are performed point-by-point for each sampling point on a unified time axis, generating a basic monitoring sequence R101. Sampling points falling within the influence range of maintenance actions, calibration actions, or fault actions are marked as unavailable, forming an unavailable monitoring point identifier. Sampling points located outside the influence range of equipment actions but retained in the basic monitoring sequence R101 continue to be read as reading objects for subsequent steps to perform judgments based on changes in turbidity and auxiliary water quality indicators. Simultaneously, quality markers are written for each sampling point. When the measured turbidity value of a sampling point is in a missing state, a turbidity missing marker is written. When the measured conductivity value of a sampling point is in a missing state, a conductivity missing marker is written. When the measured dissolved oxygen value of a sampling point is in a missing state, a dissolved oxygen missing marker is written. When other auxiliary water quality indicators are used, corresponding missing markers for each auxiliary indicator are written. When the original monitoring record in X101 contains invalid codes, fault codes, over-range codes, or other predefined abnormal status identifiers, an abnormal status marker is written. In subsequent steps, when reading the corresponding monitoring item, usable reading points can be preferentially selected based on the quality marker of the corresponding monitoring item. Finally, the measured turbidity value, measured auxiliary water quality index value, action type, action impact marker, unusable status, and quality marker corresponding to each sampling time are written into the basic monitoring sequence R101. R101 must at least include the sampling time, measured turbidity value, measured auxiliary water quality index value, action type, action impact marker, unusable status, and quality marker.

[0033] In an optional implementation, step S102 includes generating and traversing the time to be determined, reading and determining the leading stable segment, calculating the deviation and rate of change, identifying the local deviation starting point, filtering the subsequent continuation segment, and writing back R102.

[0034] In the generation and traversal of the time to be judged, the basic monitoring sequence R101 is first read, and monitoring points that have not been written as unusable and have usable measured turbidity values ​​are selected as the source of the time to be judged. Within this source of time to be judged, points are traversed point by point in a unified time axis order to identify local deviations from the starting point for each time to be judged. Subsequent steps in this process, including reading the leading stable segment, calculating the rate of change, and filtering the subsequent continuation segment, are all performed based on valid turbidity monitoring points. When a monitoring point has usable measured turbidity values, it can be used as the time to be judged to continue this step. Further judgments involving auxiliary water quality indicators are performed in subsequent steps.

[0035] In the process of reading and determining the leading stable segment, historical effective turbidity monitoring points are read backward around each time point to be determined, forming candidate segments for the leading stable segment. The leading segment length parameter is used to limit the minimum number of effective turbidity monitoring points required for the leading stable segment. After reading effective turbidity monitoring points backward along a unified time axis, when a candidate segment meets the leading segment length parameter requirement, the turbidity fluctuation range and adjacent absolute change rate are calculated around that candidate segment. The turbidity fluctuation range refers to the difference between the maximum and minimum turbidity values ​​within the candidate segment. The adjacent absolute change rate refers to the ratio of the absolute value of the turbidity difference between any two adjacent effective turbidity monitoring points within the candidate segment to the corresponding time interval. The stable fluctuation threshold is used to limit the upper limit of the overall turbidity fluctuation allowed in the leading background segment. The stable rate threshold is used to limit the maximum allowable single-step change rate of adjacent monitoring points within the leading background segment. When the turbidity fluctuation range within the candidate segment does not exceed the stable fluctuation threshold, and the adjacent absolute change rate between any two adjacent effective turbidity monitoring points does not exceed the stable rate threshold, the candidate segment is determined as the leading stable segment. Subsequently, the measured turbidity values ​​within the leading stable segment are statistically summarized to generate a representative background value for the leading stable segment, which serves as the background reference for the current time to be determined. The median value is preferred for the representative background value of the leading stable segment, but the mean value can also be used.

[0036] In the calculation of deviation and rate of change, a background deviation reference and a single-step change reference are constructed around the current time to be judged. The deviation is the absolute value of the difference between the measured turbidity value at the current time to be judged and the representative value of the preceding stable background, used to characterize the degree of separation of the current turbidity from the preceding stable background. The deviation threshold is used to limit the minimum deviation level that can be considered a significant separation from the background. The preceding effective turbidity monitoring point refers to the most recent effective turbidity monitoring point before the current time to be judged. The current rate of change is the ratio of the absolute value of the difference between the measured turbidity value at the current time to be judged and the measured turbidity value at the preceding effective turbidity monitoring point to the time interval between the two, used to characterize the speed of turbidity change at the current monitoring point relative to the most recent effective historical point. The rate threshold is used to limit the minimum rate of change level that can be considered a significant transition. In this embodiment, the deviation corresponding to the current time to be judged is first calculated and compared with the deviation threshold. When a preceding effective turbidity monitoring point has been formed before the current time to be judged, the current rate of change is then calculated and compared with the rate threshold.

[0037] In the identification of local deviation starting points, deviation amount trigger judgment and change rate trigger judgment are performed in parallel for each time point to be judged. When the current deviation amount reaches the deviation threshold, or the current change rate reaches the rate threshold, the current time point to be judged is identified as a candidate for local deviation starting point. A dual-trigger entry point with parallel deviation amount trigger and change rate trigger is used to identify deviation processes that have clearly deviated from the leading background after slow accumulation, and to identify deviation processes that have already shown a significant transition. To maintain a single entry point for the same continuous deviation process, only the first time point to be judged that meets the entry trigger condition and is retained after being filtered by the subsequent continuation segment is retained as the local deviation starting point within the same continuous deviation process. The same continuous deviation process refers to a continuous deviation segment in which, after the emergence of a candidate local deviation starting point, the absolute deviation of each effective turbidity monitoring point relative to the corresponding leading stable segment background representative value among the effective turbidity monitoring points read continuously along a unified time axis remains outside the fallback threshold, and a leading background reading starting point that meets the stability condition has not yet been re-formed during this process. When the deviation amount trigger condition or the rate of change trigger condition is met again at subsequent pending moments within the continuous deviation segment, it is not registered as a new candidate for local deviation starting point, but is treated as a subsequent monitoring point in the current continuous deviation process. After filtering the subsequent continuation segments around the current candidate point and determining which ones to retain, the candidate point is then used as the sole entry anchor point for the current continuous deviation process.

[0038] In the subsequent continuation segment filtering, short-posterior persistence confirmation is performed on the identified local deviation starting point candidates. The subsequent segment length parameter is used to limit the minimum number of valid turbidity monitoring points read from the current candidate point. The subsequent continuation segment refers to the observation segment formed by reading valid turbidity monitoring points in chronological order after the local deviation starting point candidate until the number of subsequent segment length parameters is reached. When reading the subsequent continuation segment, valid turbidity monitoring points are accumulated in chronological order. When an intermediate monitoring point is in an unavailable state or a turbidity missing state, subsequent valid turbidity monitoring points are read again until the subsequent segment length parameter requirement is met. The fallback threshold is used to limit the maximum allowable deviation level that can be considered as having exited the local deviation state. The fallback condition is that there are monitoring points in the subsequent continuation segment whose absolute deviation of turbidity relative to the corresponding leading stable segment background representative value falls back to within the fallback threshold. When the absolute deviation corresponding to each valid turbidity monitoring point in the subsequent continuation segment remains outside the fallback threshold, the current local deviation starting point candidate is confirmed as having persistence and proceeds to R102 write-back. After the current candidate point is written into R102, the continuous deviation segment from that candidate point until the first formation of the fallback condition is treated as the same continuous deviation process. Within this continuous deviation segment, even if the deviation amount trigger condition or the rate of change trigger condition is met again at subsequent decision points, no new local deviation starting point will be generated. When the fallback condition is formed in the subsequent continuation segment, the brief fallback identification is completed, and the continuous confirmation of the current candidate point ends. Around the subsequent effective turbidity monitoring points after the fallback condition, decision points are regenerated in a unified time axis order, and subsequent local deviation starting point identification is performed.

[0039] In the R102 write-back, the local deviation starting point, identified by the preceding stable reference and filtered by the subsequent continuation segment, is written into the local deviation starting point result set R102. R102 includes at least the following fields: local deviation starting point time, corresponding preceding stable segment identifier, preceding stable segment background representative value, deviation amount, rate of change, and subsequent continuation segment persistence confirmation. Specifically, the local deviation starting point time field records the starting point's position on a unified time axis. The corresponding preceding stable segment identifier and preceding stable segment background representative value fields record the historical background reference corresponding to this starting point. The deviation amount and rate of change fields record the entry determination quantity corresponding to this starting point. The subsequent continuation segment persistence confirmation field records the result of persistence confirmation formed by this starting point within the corresponding subsequent continuation segment.

[0040] In some implementations, the leading segment length parameter, stable fluctuation threshold, stable rate threshold, deviation threshold, rate threshold, subsequent segment length parameter, and fallback threshold can be pre-configured and determined based on the statistical distribution of the target monitoring station's historical normal time period, known deviation samples, and short-term disturbance samples. This facilitates parameter configuration, engineering implementation, and adaptation to different sites. The leading segment length parameter and subsequent segment length parameter can preferably be calculated as the number of effective turbidity monitoring points based on the sampling period and corresponding observation duration. The stable fluctuation threshold and stable rate threshold can preferably be determined based on the overall fluctuation upper bound and single-step change upper bound of the historical normal time period. The deviation threshold and rate threshold can preferably be determined based on the distinguishable distribution between normal samples and known deviation samples. The fallback threshold can preferably be determined based on the deviation regression range of short-term disturbance samples during the recovery phase. In some implementations, the identification of the same continuous deviation process preferably uses the fallback condition as the process end marker, and the re-entry into the judgment time generation chain after the fallback condition as the starting point for identifying the new round of local deviations. This facilitates entry deduplication, engineering implementation, and scan advancement control under different sampling periods. When the sampling period is short and local fluctuations are frequent, it is preferable to use the fallback threshold and subsequent segment length parameters to jointly constrain the end position of the same continuous deviation process. When the sampling period is long and single-step changes are large, the starting point of the generation chain for re-entering the judgment time can also be modified by combining the rate threshold and deviation threshold to avoid the same continuous deviation process being repeatedly registered as multiple entry points. When historical data from sub-sites is available, the above parameters can preferably be determined based on the historical statistical results of the current monitoring station. When historical data from multiple sites is available, basic parameters can be formed first based on the historical statistical results of similar monitoring stations, and then boundary corrections can be performed by combining the current monitoring station's sampling period and turbidity fluctuation level to form a configuration suitable for the current monitoring station.

[0041] In an optional implementation, step S103 includes reading the local deviation from the starting point and establishing the segment basis, reverse search of the starting boundary candidate, confirmation of the entry condition, forward search of the ending boundary candidate, verification through the backward observation window, and writing back in R103. (Refer to...) Figure 2 The specific process is as follows.

[0042] In the process of reading local deviation starting points and establishing the segment basis, the local deviation starting points retained in the local deviation starting point result set R102 are first read. Then, around each local deviation starting point, the measured turbidity values, measured auxiliary water quality index values, action type, action impact marker, unusable status, and quality markers before and after the corresponding time position are extracted from the basic monitoring sequence R101 to form the segment determination basis corresponding to the current local deviation starting point. In this step, the local deviation starting point is used as the central anchor point. The segment in front of it is used for the initial boundary candidate search and entry condition confirmation, and the segment behind it is used for the end boundary candidate search and backward observation window verification.

[0043] In the reverse search of the initial boundary candidate, monitoring points are searched backward along time around the current local deviation starting point, and the action influence marker, measured turbidity value, and deviation status relative to the corresponding leading stable segment background representative value of each monitoring point are read. The proximity range is used to limit the maximum time distance that allows a temporal proximity relationship between the equipment action and the local deviation process. The equipment action time is preferably the boundary time of the corresponding action influence interval with the smallest time distance from the current candidate monitoring point. When the current candidate monitoring point is adjacent to multiple action influence intervals, the boundary time of the corresponding action influence interval with the smallest time distance is preferably taken as the equipment action time corresponding to the current candidate monitoring point. The equipment influence morphology condition is used to confirm whether the current local deviation process has equipment-side disturbance signs. In this embodiment, the equipment influence morphology condition is preferably confirmed by the fulfillment of at least one of the following: abrupt entry morphology, short-term offset morphology, and short-segment distortion morphology. When the same candidate monitoring point simultaneously meets two or three types of morphology, the candidate monitoring point is still treated as a single initial boundary candidate, and the results of the fulfillment of multiple morphology types corresponding to it are written into the entry confirmation basis field. When the same candidate monitoring point does not form any morphology fulfillment result, the candidate monitoring point is not determined as an initial boundary candidate. The first type is the abrupt entry pattern. An abrupt entry pattern is confirmed when the absolute rate of change of the current monitoring point relative to the preceding effective turbidity monitoring point reaches the equipment transition rate threshold, and the absolute deviation of the current monitoring point relative to the background representative value of the corresponding leading stable segment reaches the equipment transition deviation threshold. The second type is the short-term offset pattern. After continuously reading effective turbidity monitoring points up to the offset duration parameter number from the current monitoring point, a short-term offset pattern is confirmed when the absolute deviation of each effective turbidity monitoring point within the observation segment relative to the background representative value of the corresponding leading stable segment reaches the offset amplitude threshold. The third type is the short-segment distortion pattern. Within the local observation segment corresponding to the distortion observation length parameter, a short-segment distortion pattern is confirmed when the absolute rate of change between at least two adjacent effective turbidity monitoring points reaches the distortion rate threshold, and the number of turbidity deviation direction switching times reaches the distortion switching number threshold. When the time interval between a candidate monitoring point and the equipment action time falls within the adjacent range, and the corresponding turbidity meets at least one of the aforementioned equipment influence pattern conditions, the candidate monitoring point is determined as the initial boundary candidate, and the entry condition confirmation continues around this initial boundary candidate.

[0044] During the entry condition confirmation process, auxiliary water quality index linkage confirmation and turbidity continuous deviation confirmation are performed around the current initial boundary candidate. The allowable lag time is used to limit the maximum observation time for auxiliary water quality indicators to reach the linkage confirmation condition. The change in auxiliary water quality indicators refers to the difference between the measured value of the auxiliary water quality indicator at each monitoring point after the initial boundary candidate and the corresponding background representative value of the auxiliary water quality indicator. In this embodiment, the change in auxiliary water quality indicators is preferably the absolute value of the difference between the measured value of the corresponding auxiliary water quality indicator and the corresponding background representative value of the auxiliary water quality indicator. In the same entry condition confirmation, the same auxiliary water quality indicator preferably maintains a uniform calculation method. The auxiliary change threshold is used to limit the minimum change level required for the linkage change of auxiliary water quality indicators to be valid. The background representative value of the auxiliary water quality indicator is preferably taken from the statistical representative value of the corresponding auxiliary water quality indicator within the leading stable segment corresponding to the current local deviation starting point. In this embodiment, the background representative value of each auxiliary water quality indicator is preferably its median value within the corresponding leading stable segment, or the mean value can be used, but in the same entry condition confirmation, the same auxiliary water quality indicator preferably maintains a uniform statistical method. Entry condition confirmation is preferably based on activated auxiliary water quality indicators with available quality markers. Activated auxiliary water quality indicators are preferably derived from the preset activated auxiliary water quality indicator configuration of the current monitoring station, and the activated set remains consistent within the same entry condition confirmation. When multiple auxiliary water quality indicators are activated, the condition for the activation of auxiliary water quality indicator linkage confirmation is that at least one auxiliary water quality indicator reaches the auxiliary change threshold within the allowable lag time. When reading auxiliary water quality indicators around the allowable lag time, if the auxiliary water quality indicator corresponding to the current monitoring point has a missing measurement marker or an abnormal status marker, the system continues to read subsequent auxiliary water quality indicators with available quality markers within the allowable lag time range. If no auxiliary water quality indicator reading value meeting the quality marker requirements is formed within the allowable lag time range, the current initial boundary candidate will not form an auxiliary water quality indicator linkage confirmation.

[0045] The entry duration observation length parameter limits the minimum number of consecutive effective turbidity monitoring points required for confirmation of persistent turbidity deviation. Around the current candidate starting boundary, consecutive effective turbidity monitoring points reaching the number specified by the entry duration observation length parameter are read sequentially over time, forming an entry duration observation segment. Confirmation of persistent turbidity deviation means that the absolute deviation of each effective turbidity monitoring point within this entry duration from the corresponding leading stable background representative value remains outside the fallback threshold. When at least one auxiliary water quality indicator change reaches the auxiliary change threshold within the allowable lag time, auxiliary water quality indicator linkage confirmation is formed. When the current candidate starting boundary simultaneously forms both auxiliary water quality indicator linkage confirmation and persistent turbidity deviation confirmation, the entry condition is determined to be met, and this candidate starting boundary is designated as the starting boundary. When the current candidate starting boundary does not form a result indicating that the entry condition is met, a new candidate monitoring point is searched backwards over time, and the current entry condition confirmation is re-executed for the new candidate starting boundary. The upper limit of the reverse search limits the maximum backtracking range for performing a reverse search of the starting boundary candidate around the current local deviation starting point. In this embodiment, the upper limit of the reverse search is preferably determined according to the maximum backtracking time on a unified time axis. When an unavailable state or turbidity measurement failure is encountered during the reverse search, the reverse search continues along the unified time axis until the upper limit of the reverse search is reached. Within the same local deviation process, the candidate starting boundary that is first confirmed by the entry conditions is preferred as the starting boundary, and the reverse search corresponding to the current local deviation process is stopped. If the current local deviation process reaches the upper limit of the reverse search but has not yet formed a starting boundary confirmed by the entry conditions, no confusion time segment is generated for the current local deviation starting point, and the current step continues to be executed around the next local deviation starting point.

[0046] In the forward search for the candidate termination boundary, for the current local deviation process where the starting boundary has been determined, monitoring points are searched forward along time from the corresponding local deviation starting point, and it is determined whether a candidate termination boundary is formed around each monitoring point. The turbidity decline condition refers to the deviation of the turbidity at the current monitoring point relative to the corresponding leading stable segment background representative value falling back to within the decline threshold. The auxiliary decline threshold is used to limit the maximum allowable change level of auxiliary water quality indicators exiting the linked change state. The auxiliary indicator decline condition refers to the change of at least one auxiliary water quality indicator falling back to within the auxiliary decline threshold. The release condition refers to the joint determination result of the turbidity decline condition and the auxiliary indicator decline condition being met. Confirmation of the release condition is preferably based on an activated auxiliary water quality indicator with available quality markers. The observation range corresponding to the release condition confirmation refers to the range of auxiliary water quality indicator readings formed from the current monitoring point along time sequence within a range not exceeding the allowable lag time. The change in auxiliary water quality indicators in the auxiliary indicator decline condition preferably uses the same calculation method as the change in auxiliary water quality indicators in the entry condition confirmation. When reading auxiliary water quality indicators within the corresponding observation range to confirm the conditions for lifting the restrictions, if the auxiliary water quality indicator corresponding to the current monitoring point has a missing measurement mark or an abnormal status mark, continue reading subsequent auxiliary water quality indicators with usable quality marks within the corresponding observation range. If no auxiliary water quality indicator reading value meeting the quality mark requirements is formed within the corresponding observation range, the current monitoring point does not form an auxiliary indicator decline condition.

[0047] In this embodiment, when searching for monitoring points in the forward time direction, the deviation of the turbidity of the current monitoring point relative to the corresponding leading stable segment background representative value is first calculated, and a turbidity decline condition is determined accordingly. Then, the change of at least one enabled auxiliary water quality indicator relative to the corresponding auxiliary water quality indicator background representative value is calculated, and an auxiliary indicator decline condition is determined accordingly. When the current monitoring point simultaneously meets both the turbidity decline condition and the auxiliary indicator decline condition, the monitoring point is designated as a candidate for the end boundary, and a backward observation window verification is performed around this candidate. If the current candidate for the end boundary still meets the entry condition after the backward observation window verification, the candidate for the end boundary is not designated as the end boundary, and a new candidate for the end boundary continues to be searched in the forward time direction. Within the same local deviation process, the candidate for the end boundary that first simultaneously passes the release condition confirmation and the backward observation window verification is preferably taken as the end boundary, and the forward search corresponding to the current local deviation process is stopped. When continuing the search around the current local deviation process, the background representative value of the leading stable segment and the background representative value of the auxiliary water quality index both adopt the background reference corresponding to the current local deviation starting point, so as to maintain the consistency of the judgment caliber before and after the current confusion time segment.

[0048] In the backward observation window verification, stability confirmation after removal is performed on the already formed end boundary candidate. The backward observation length parameter is used to limit the minimum number of consecutive effective turbidity monitoring points required for stability verification after the end boundary candidate. The backward observation window refers to the verification segment formed by reading consecutive effective turbidity monitoring points in chronological order from the end boundary candidate, reaching the number of backward observation length parameters. The tail of the current processing range refers to the last effective turbidity monitoring point that can be read on the unified time axis when continuing to perform forward search of the end boundary candidate and backward observation window verification around the current local deviation starting point. When reading the backward observation window, effective turbidity monitoring points are continuously accumulated around the unified time axis. When an intermediate monitoring point has an unavailable state or a turbidity missing state, subsequent effective turbidity monitoring points continue to be accumulated backward until the backward observation length parameter requirement is reached. A re-entry event refers to a situation in the backward observation window after the current end boundary candidate where the re-entry condition is met again. The situation where the re-entry condition is met should be located after the current end boundary candidate and correspond to a new time position. When a monitoring point within the backward observation window is located at a new time position, and a new starting boundary is formed after re-performing the initial boundary candidate search and entry condition confirmation around that monitoring point, it is considered that the entry condition has been met again. If no new starting boundary is formed, it is treated as a continuation of the current local deviation process. If the current local deviation process continues within the backward observation window without meeting the entry condition again, it is not considered a re-entry event. When performing a review around the backward observation window, each monitoring point within the window is checked again for whether the entry condition is met again according to the aforementioned initial boundary candidate search and entry condition confirmation criteria in the current step. If the entry condition is not met again within the backward observation window, the current end boundary candidate is determined as the end boundary. If the entry condition is met again within the backward observation window, a new end boundary candidate is searched forward along time around the current local deviation process, and the current backward observation window review is re-performed on the new end boundary candidate. Within the same local deviation process, the end boundary candidate that first simultaneously passes the release condition confirmation and backward observation window review is preferably taken as the end boundary, and the forward search corresponding to the current local deviation process is stopped. If the search for the current local deviation process reaches the end of the current treatment range without forming a valid termination boundary, the last effective turbidity monitoring point at the end of the current treatment range will not be used as the termination boundary, and the determination of the confusion time segment corresponding to the starting point of the current local deviation will end. When continuing the search around the current local deviation process, the background representative value of the leading stable segment and the background representative value of the auxiliary water quality indicators will both use the background reference corresponding to the starting point of the current local deviation to maintain consistency in the judgment criteria before and after the current confusion time segment.

[0049] In the R103 write-back, the segment between the start and end boundaries is only defined as the confusion time segment and written to the confusion time segment result set R103 after the start and end boundaries have been confirmed. R103 includes at least the following fields: corresponding local deviation start point field, start boundary field, end boundary field, confusion time segment record field, entry confirmation basis field, deconfirmation basis field, and backward observation verification result field. Specifically, the corresponding local deviation start point field records the local deviation start point corresponding to the current confusion time segment. The start boundary field records the starting position of the segment after entry condition confirmation. The end boundary field records the ending position of the segment after deconfirmation and backward observation window verification. The confusion time segment record field records the segment range and corresponding monitoring record of the current confusion time segment. The entry confirmation basis field records the entry confirmation result of the current segment. The deconfirmation basis field records the deconfirmation result of the current segment. The backward observation verification result field records the stability verification result after the end boundary of the current segment. The backward observation verification result field is used to record the backward observation window verification result corresponding to the current end boundary. When the current local deviation process has reached the end of the current processing range and has not yet formed an end boundary that has passed verification, the obfuscation time segment record is not written to R103.

[0050] In some implementations, the aforementioned proximity range, equipment transition rate threshold, equipment transition deviation threshold, offset duration parameter, offset amplitude threshold, distortion observation length parameter, distortion rate threshold, distortion switching count threshold, allowable lag duration, auxiliary change threshold, entry duration observation length parameter, fallback threshold, auxiliary fallback threshold, and backward observation length parameter can be pre-configured and determined based on the statistical distribution of the target monitoring station's historical normal time period, known equipment movement disturbance samples, and known confused samples. This facilitates parameter configuration, engineering implementation, and adaptation to different sites. Preferably, the configuration of parameters related to equipment influence morphology follows the principle of first distinguishing between normal water body change samples and equipment movement disturbance samples, and then determining the threshold for each morphology. The equipment transition rate threshold and equipment transition deviation threshold are preferably determined based on the change rate distribution and deviation amplitude distribution of rapid entry characteristics in the equipment movement proximity samples, facilitating parameter configuration for abrupt entry morphologies. The offset duration parameter and offset amplitude threshold are preferably determined based on the distribution of the number of continuous points and the offset amplitude distribution of short-segment overall offset processes in the equipment movement proximity samples, facilitating parameter configuration for short-term offset morphologies. The distortion observation length parameter, distortion rate threshold, and distortion switching frequency threshold are preferably determined based on the length distribution, single-step absolute change rate distribution, and deviation direction switching distribution of the local distortion segment in the equipment action disturbance sample, to facilitate parameter configuration for short-segment distortion patterns. The parameters related to entry confirmation can be preferably determined based on the historical lag distribution of relative turbidity changes in auxiliary water quality indicators, the differential distribution of auxiliary change amounts, and the continuous observation requirements under the sampling period. The parameters related to release and verification can be preferably determined based on the turbidity regression range, the regression distribution of auxiliary indicators, and the common occurrence interval of re-entry events after the confusion state is released. When historical data from sub-sites is available, the above parameters can preferably be determined based on the historical statistical results of the current monitoring station. When historical data from multiple sites is available, basic parameters can be first formed based on the historical statistical results of similar monitoring stations, and then boundary corrections can be performed in conjunction with the current monitoring station's sampling period, monitoring indicator fluctuation levels, and equipment action frequency. For various length parameters configured according to the number of continuous effective turbidity monitoring points, it is preferable to first convert the corresponding number of effective turbidity monitoring points based on the preset time length and sampling period, and then perform corrections around the stable intervals in historical samples to form a configuration caliber suitable for the current monitoring station.

[0051] In an optional implementation, step S104 includes reading the obfuscated time segment and extracting the segment's basic information, switching the dedicated exit, constructing the obfuscated segment record, writing the segment boundary and corresponding monitoring records within the segment, and updating the long-term series processing results.

[0052] In the process of reading and extracting basic information of confused time segments, the confused time segment result set R103 generated in step S103 is first read, and the start boundary time, end boundary time, corresponding local deviation from the start point time, and corresponding monitoring records within the segment are extracted one by one for each confused time segment. For each confused time segment, the measured turbidity records, measured auxiliary water quality index records, action type, action impact markers, unusable status, and quality markers falling within the segment's range are further extracted from the basic monitoring sequence R101, forming the record construction basis for the current confused time segment. By uniformly extracting the boundary position, corresponding start point, and corresponding monitoring records within the segment for the confirmed confused time segments, this step ensures that subsequent dedicated outlet switching and confused segment record writing are completed within the same segment object scope.

[0053] During dedicated exit switching, the result destination is updated around the current obfuscated time segment, switching the current segment from the regular anomaly handling exit in the long-term series processing results to the dedicated record exit for obfuscated segments. The regular anomaly handling chain refers to the default result destination in long-term series processing where regular retention or removal is performed directly for anomaly periods. Dedicated exit switching refers to the result switching process of updating the exit status for the current obfuscated time segment, causing it to enter the dedicated record exit for obfuscated segments.

[0054] In this embodiment, the dedicated exit switching is preferably achieved through a two-layer approach: segment-level dedicated exit status writing and record-level dedicated exit status writing. First, segment-level dedicated exit status is written around the current obfuscated time segment to indicate that the entire current obfuscated time segment uses the obfuscated segment dedicated record exit. Then, record-level dedicated exit status is written around the corresponding monitoring records within the segment, ensuring that each monitoring record within the segment maintains the same dedicated exit destination as the current segment. Through these two layers of dedicated exit status writing, the current obfuscated time segment and its corresponding monitoring records jointly complete the switching to the obfuscated segment dedicated record exit.

[0055] In the construction of the confusion segment record, a confusion segment record R104 is constructed for the current confusion time segment that has completed the dedicated outlet switch. R104 includes at least the confusion segment number field, the start boundary time field, the end boundary time field, the corresponding local deviation start time field, the segment length field, the corresponding monitoring record field within the segment, and the dedicated outlet status field. The confusion segment number field identifies the independent identity of each confusion time segment in the long-term series processing results. The start and end boundary time fields record the boundary positions of the current confusion time segment. The corresponding local deviation start time field records the local deviation start point around which the current confusion time segment is located. The segment length field records the segment span between the start and end boundaries. The corresponding monitoring record field within the segment records the measured turbidity records, measured auxiliary water quality index records, action types, action impact markers, unavailable status, and quality markers falling within the current confusion time segment range. The dedicated outlet status field records the result status of the current confusion time segment having completed the dedicated outlet switch.

[0056] In some implementations, the obfuscated segment record R104 may further include an entry confirmation basis field, a deconfirmation confirmation basis field, and an exit status update result field. The entry confirmation basis field records the entry confirmation result corresponding to the current obfuscated time segment. The deconfirmation confirmation basis field records the deconfirmation confirmation result corresponding to the current obfuscated time segment. The exit status update result field records the result of the dedicated exit switch completed for the current obfuscated time segment and its corresponding monitoring records. These fields are preferred supplementary fields used to improve the traceability of the obfuscated segment record and the interpretability of subsequent processing.

[0057] During the writing of monitoring records at the segment boundaries and within the segment, the start and end boundary times are simultaneously written around the current confused time segment, enabling subsequent readers to directly recover the corresponding range of the confused time segment in the overall time series based on the boundary positions. The start and end boundary time fields preferably maintain a time expression consistent with the unified time axis and, together with the corresponding local deviation start time field, constitute the location information of the current confused time segment. The segment length field is preferably determined based on the time span between the start and end boundary times or the corresponding sampling point span. All monitoring point records falling within the segment range between the start and end boundaries are extracted around the current confused time segment, maintaining the original chronological order of these records under the unified time axis. The corresponding monitoring record fields within the segment preferably include turbidity measurement records, auxiliary water quality indicator measurement records, action type, action impact markers, unavailable status, and quality markers to maintain consistency between the recorded content within the segment and the recording scope of the basic monitoring sequence R101.

[0058] During the long-time series processing result update, the constructed obfuscated segment record R104 is written into the obfuscated segment record area of ​​the long-time series processing result, and the dedicated exit status corresponding to the current obfuscated time segment is written simultaneously. Through this result update process, the long-time series processing result simultaneously inherits the results of regular anomaly handling and the obfuscated segment record, while maintaining a clear boundary of responsibility between the two types of results in the result layer. After the write update, the current obfuscated time segment becomes an obfuscated segment record result that can be directly accessed by subsequent readers.

[0059] In some implementations, the obfuscated segment number field, start boundary time field, end boundary time field, and corresponding local deviation from the start point time field can preferably be organized according to a unified timeline order and the writing order of the current segment in the result set, to facilitate result sorting and subsequent reading. The segment length field can preferably be determined based on the boundary time difference under a unified timeline or the corresponding number of valid sampling points, to adapt to the segment length representation under different sampling periods. The dedicated exit status field and exit status update result field can preferably be written using a unified status code or a unified status name, to facilitate result organization, engineering implementation, and status alignment between different processing modules.

[0060] In one specific embodiment, the application process of this method is illustrated using a simulated experimental scenario of an automatic monitoring station at a fixed cross-section of a river as an example. This simulated experimental scenario generates turbidity and conductivity records at a 5-minute sampling cycle, and sets an automatic calibration action to occur between 09:40 and 09:50. The example time period is from 09:15 to 10:35.

[0061] During this example time period, the period from 09:15 to 09:35 is the stable background segment before calibration, with corresponding measured turbidity values ​​of 11.8 NTU, 12.1 NTU, 11.9 NTU, 12.0 NTU, and 12.2 NTU, and corresponding measured conductivity values ​​of 428 μS / cm, 430 μS / cm, 429 μS / cm, 431 μS / cm, and 430 μS / cm, which can be considered as a stable background. Step S101 generates the action influence interval based on the start and end times of the automatic calibration action, and writes 09:40 and 09:45 into the unavailable state, forming the basic monitoring sequence R101.

[0062] Step S102: Five effective turbidity monitoring points from 09:15 to 09:35 are used to form a leading stable segment, and the median value of 12.0 NTU is taken as the representative value of the leading stable segment. The turbidity value at 09:50 is 14.2 NTU, which does not reach the local deviation start point identification threshold. The turbidity value at 09:55 is 16.8 NTU, with an absolute deviation of 4.8 NTU relative to the background representative value, and the rate of change relative to the previous effective turbidity monitoring point reaches the rate threshold. Therefore, 09:55 is identified as the local deviation start point. Continuing to read the values ​​at 10:00, 10:05, and 10:10, the corresponding turbidity values ​​are 19.6 NTU, 18.9 NTU, and 17.4 NTU, respectively. None of these values ​​have fallen back to within the fallback threshold. Therefore, 09:55 is determined as the local deviation start point and written into R102.

[0063] Step S103: Confirm the starting boundary forward from 09:55. Since the time interval between 09:50 and the end of the automatic calibration is 0 minutes, and the turbidity continues to increase and the conductivity continues to shift thereafter, 09:50 is determined as the starting boundary. The measured conductivity values ​​at 09:50, 09:55, and 10:00 are 440 μS / cm, 447 μS / cm, and 452 μS / cm, respectively, with changes relative to the background value of 430 μS / cm of 10 μS / cm, 17 μS / cm, and 22 μS / cm, respectively, reaching the auxiliary change threshold. Subsequently, searching forward in time, the turbidity value at 10:20 is 12.8 NTU, and the conductivity value is 433 μS / cm, both having fallen back to within the corresponding thresholds. Therefore, 10:20 is determined as the ending boundary. Continuing to read 10:25, 10:30 and 10:35, no new entry point that meets the entry conditions was found again. Therefore, 09:50 to 10:20 was determined as the obfuscated time segment and written to R103.

[0064] Step S104 involves writing segment-level and record-level dedicated exit states to the confused time segment and constructing a confused segment record R104. This record includes at least the confused segment number, the start boundary time of 09:50, the end boundary time of 10:20, the corresponding local deviation start time of 09:55, and the corresponding monitoring records within the segment. Ultimately, this time segment is not directly retained or removed as a normal abnormal time segment, but is written separately as a confused segment into the long-term series processing results.

[0065] Therefore, in this simulated experimental scenario, this method can identify the local deviation starting point, determine the confusion time segment, and write the confusion segment record around a single local increase in turbidity that occurs in the period adjacent to the automatic calibration action, and separate this local process from the ordinary anomaly handling chain.

[0066] The above embodiments can be implemented, in whole or in part, by software, hardware, firmware, or any other combination thereof. When implemented using software, the above embodiments can be implemented, in whole or in part, in the form of a computer program product.

[0067] Those skilled in the art will recognize that the modules and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and inventive constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0068] In addition, the functional modules in the various embodiments of this application can be integrated into one processing module, or each module can exist physically separately, or two or more modules can be integrated into one module.

[0069] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

[0070] In conclusion, the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. 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 method for storing and processing long-term environmental monitoring data, characterized in that, include: S101, Read the monitoring records and equipment action records within the target time period, and mark unusable monitoring points based on the equipment action records. The monitoring records include at least the turbidity measurement records and the auxiliary water quality index measurement records. S102, for each time point to be judged in the monitoring records after excluding unusable monitoring points, read the continuous historical segment that meets the stability conditions in front of it as the leading stable segment, and identify the local deviation starting point based on the deviation of the current turbidity relative to the leading stable segment and the rate of change relative to the previous effective monitoring point. S103, determine the entry and release conditions around each local deviation starting point and determine the confusion time segment. Specifically, search backward from the local deviation starting point for the first monitoring point that meets the condition that the time interval between the deviation and the equipment operation time does not exceed the adjacent range and the corresponding turbidity meets the equipment influence morphology condition, and designate it as the initial boundary candidate. When the change of at least one auxiliary water quality indicator reaches the auxiliary change threshold within the allowable lag time from the initial boundary candidate, and the turbidity does not meet the fallback condition again within the subsequent consecutive preset number of valid monitoring points, the entry condition is determined to be met, and the initial boundary candidate is determined to be the initial boundary. When the turbidity meets the fallback condition again and the change of at least one auxiliary water quality indicator falls back to within the auxiliary fallback threshold, the release condition is determined to be met. Search backward from the local deviation starting point for the first monitoring point that meets the release condition, and designate it as the end boundary candidate. Read the consecutive preset number of valid monitoring points backward from the end boundary candidate. When the entry condition is not met again within the consecutive preset number of valid monitoring points, the end boundary candidate is determined to be the end boundary. The segment between the initial boundary and the end boundary is determined as the confusion time segment.

2. The method for storing and processing long-term environmental monitoring data according to claim 1, characterized in that, In S101, the action type, action start time, and action end time in the device action record are read to generate the corresponding action influence interval; when the device action record contains only a single registration time, it is expanded into an action influence interval according to the preset influence duration of the corresponding action type; sampling points that fall into the maintenance action influence interval, calibration action influence interval, or fault action influence interval are marked as unavailable monitoring points.

3. The method for storing and processing long-term environmental monitoring data according to claim 2, characterized in that, In S102, effective turbidity monitoring points that reach the number of leading segment length parameters are read forward around each time to be judged to form a candidate segment of the leading stable segment; the turbidity fluctuation range and the adjacent absolute change rate between any two adjacent effective turbidity monitoring points are calculated within the candidate segment of the leading stable segment; when the turbidity fluctuation range does not exceed the stable fluctuation threshold and each adjacent absolute change rate does not exceed the stable rate threshold, the candidate segment of the leading stable segment is determined as the leading stable segment.

4. The method for storing and processing long-term environmental monitoring data according to claim 3, characterized in that, In S102, the absolute value of the difference between the measured turbidity value at the current time to be judged and the background representative value of the corresponding leading stable segment is extracted as the deviation amount, and the ratio of the absolute value of the difference between the measured turbidity value at the current time to be judged and the measured turbidity value at the previous effective turbidity monitoring point to the time interval between the two is extracted as the rate of change; when the deviation amount reaches the deviation threshold or the rate of change reaches the rate threshold, the current time to be judged is identified as a candidate for the local deviation starting point.

5. The method for storing and processing long-term environmental monitoring data according to claim 4, characterized in that, In S102, effective turbidity monitoring points that reach the number of subsequent segment length parameters are read backward from the candidate for local deviation starting point to form a subsequent continuation segment; when the absolute deviation of each effective turbidity monitoring point in the subsequent continuation segment relative to the background representative value of the corresponding leading stable segment remains outside the fallback threshold, the current candidate for local deviation starting point is determined as the local deviation starting point; the continuous deviation segment from the local deviation starting point to the subsequent first formation of the fallback condition is regarded as the same continuous deviation process. Even if the deviation amount trigger condition or the change rate trigger condition is met again at the subsequent judgment time in the same continuous deviation process, no new local deviation starting point will be generated.

6. The method for storing and processing long-term environmental monitoring data according to claim 5, characterized in that, In S103, candidate monitoring points are searched in reverse time around the local deviation starting point. When the time interval between the candidate monitoring point and the device action time falls into the vicinity range, and the corresponding turbidity meets at least one type of device influence morphology condition, the candidate monitoring point is determined as the starting boundary candidate. The device influence morphology conditions include abrupt entry morphology, short-term offset morphology, and short-segment distortion morphology. The abrupt entry morphology is when the absolute change rate of the current monitoring point relative to the preceding effective turbidity monitoring point reaches the device transition rate threshold and the absolute deviation of the current monitoring point relative to the background representative value of the corresponding leading stable segment reaches the device transition deviation threshold. The short-term offset morphology is when, within the observation segment formed by continuously reading effective turbidity monitoring points up to the number of offset duration parameters starting from the current monitoring point, the absolute deviation of each effective turbidity monitoring point relative to the background representative value of the corresponding leading stable segment reaches the offset amplitude threshold. The short-segment distortion morphology is when, within the local observation segment corresponding to the distortion observation length parameter, the absolute change rate between at least two adjacent effective turbidity monitoring points reaches the distortion rate threshold and the number of turbidity deviation direction switching times reaches the distortion switching number threshold.

7. The method for storing and processing long-term environmental monitoring data according to claim 6, characterized in that, In S103, consecutive effective turbidity monitoring points are read backward around the candidate starting boundary to reach the number of continuous observation length parameters, forming a continuous observation section; when at least one auxiliary water quality indicator that has been enabled and has available quality markers reaches the auxiliary change threshold within the allowable lag time, and the absolute deviation of each effective turbidity monitoring point in the continuous observation section relative to the corresponding leading stable section background representative value remains outside the fallback threshold, the candidate starting boundary is determined as the starting boundary.

8. The method for storing and processing long-term environmental monitoring data according to claim 7, characterized in that, In S103, the monitoring point is searched in the forward time direction; when the deviation of the turbidity of the current monitoring point from the background representative value of the corresponding leading stable segment falls back to within the fallback threshold, and the change of at least one auxiliary water quality indicator that has been enabled and has available quality markers falls back to within the auxiliary fallback threshold, the current monitoring point is determined as a candidate for the end boundary.

9. A method for storing and processing long-term environmental monitoring data according to claim 8, characterized in that, In S103, consecutive valid turbidity monitoring points that reach the backward observation length parameter number are read backward from the candidate end boundary to form a backward observation window; when the entry condition is not met again in the backward observation window, the current candidate end boundary is determined as the end boundary; when the entry condition is met again in the backward observation window, a new candidate end boundary is searched along the forward time direction.

10. A method for storing and processing long-term environmental monitoring data according to claim 1, characterized in that, This method also includes S104, in which, for each obfuscated time segment, the abnormal processing of directly retaining or directly removing the obfuscated time segment is blocked, and the start and end positions of the obfuscated time segment and the corresponding monitoring records are written into the obfuscated segment record in the long-term series processing result; wherein, the obfuscated time segment is written with the segment-level dedicated exit status and the record-level dedicated exit status of the corresponding monitoring record within the segment, and the obfuscated segment record is constructed. The obfuscated segment record includes at least the obfuscated segment number, the start boundary time, the end boundary time, the corresponding local deviation from the start point time, the segment length, and the corresponding monitoring record within the segment.

Images