Water conservancy project construction progress real-time evaluation and early warning method

Abstract

The invention relates to the technical field of project management, in particular to a water conservancy project construction progress real-time evaluation and early warning method, which comprises the following steps of obtaining an operation list, comparing the operation list with real-time time, identifying lagged nodes, extracting an interval increasing path, marking a chain type path, analyzing a space overlapping and compression task, and identifying a compression area. And counting node repetition and starting frequency, identifying jump sensitive points, extracting paths and resource density, and generating a risk early warning level set. According to the method, through comprehensive analysis of a task execution interval coverage proportion, starting offset and a task asynchronous completion phenomenon, dynamic identification of key construction nodes is formed, progressive change of an interval trend between the nodes is combined, continuous lag structure labeling of a path level is formed, and intersection analysis of a spatial position and a task execution time period is introduced; and a task gathering area which is compressed and executed in the same time period is identified, so that the identification precision of potential construction progress abnormity and the rationality of early warning grade layering are effectively improved.

Description

Technical Field

[0001] This invention relates to the field of project management technology, and in particular to a method for real-time assessment and early warning of the construction progress of water conservancy projects. Background Technology

[0002] The field of water conservancy project management technology involves the management and control of water conservancy projects throughout the entire process, from preliminary planning, design, construction to post-construction maintenance. Core aspects include schedule management, cost management, quality control, safety assurance, resource allocation, and environmental protection. This technical field utilizes methods such as project scheduling, construction organization design, optimal resource allocation, and multi-level coordination mechanisms to achieve standardized and controllable management of the entire water conservancy project construction process, ensuring the smooth progress of the project within the predetermined time and resource conditions.

[0003] Traditional real-time assessment and early warning methods for water conservancy project construction progress refer to the technical approach of evaluating the completion status of construction milestones, the progress of key processes, and the overall execution of the construction plan during the construction of water conservancy projects, and issuing early warnings when deviations from the plan occur. This method relies on manually collecting construction progress data and summarizing and analyzing it based on construction logs, on-site photographic records, and regular progress reports, while also incorporating rules of thumb for deviation judgment and risk warning. This approach primarily uses paper records or Excel spreadsheets to compare and analyze the progress milestones of each construction unit, and management personnel manually assess the coordination between each process and its impact on the overall progress.

[0004] Current technologies commonly use manual data collection and paper or tabular records to compare construction progress milestones. This approach suffers from problems such as delayed data acquisition, low analysis efficiency, and strong subjectivity in judgment. In complex situations involving multiple tasks running concurrently and overlapping construction, it is difficult to reveal coordination deviations between key tasks in a timely manner, resulting in the failure to identify potential risk nodes in advance. When multiple critical paths experience compressed execution in a certain spatial area, relying on experience-based judgment cannot accurately determine the chain reaction delays on the overall project schedule. This makes it difficult for managers to adjust resource allocation and construction organization plans in a timely manner, affecting the overall response efficiency and effectiveness of construction control. Summary of the Invention

[0005] The purpose of this invention is to address the shortcomings of existing technologies by proposing a method for real-time assessment and early warning of the construction progress of water conservancy projects.

[0006] To achieve the above objectives, the present invention adopts the following technical solution: a method for real-time assessment and early warning of construction progress in water conservancy projects, comprising the following steps: S1: Obtain the work list of structures in the water conservancy project, record the planned start and end times and the real-time start and end times recorded on site, compare the coverage ratio of real-time time within the planned time period, record the advance or delay of the task start order in the planned arrangement, and generate a set of synchronous lag nodes. S2: Based on the set of synchronous lag nodes, extract the node numbers belonging to the intersection of multiple tasks, read the real-time completion time of the preceding task and the real-time start time of the node, number and mark the order of the preceding and following intervals, and generate a chain-like progress lag path table. S3: Using the chained progress lag path table, extract the spatial operation unit number of the task, extract the spatial location range and real-time execution period of the task, determine whether the task combination is a parallel process in the original plan, and generate a list of spatial compression response zone identifiers. S4: Based on the set of synchronous lag nodes, the chained progress lag path table, and the spatial compression response zone identifier list, count the number of times the node number and spatial number appear, determine whether the same node appears repeatedly in the chained progress lag path table and the spatial compression response zone identifier list, identify whether the number of times the node is started decreases within three construction cycles, and generate a multi-dimensional risk jump point list.

[0007] As a further embodiment of the present invention, the set of synchronous lag nodes includes task coverage ratio, task start offset, and synchronous lag node number; the chained progress lag path table includes chain path number, node interval trend label, and lag path sequence; the spatial compression response zone identifier list includes spatial region number, compression task combination number, and number of compression execution tasks; and the multidimensional risk jump point list includes duplicate node number, periodic start frequency trend, and jump sensitive node number.

[0008] As a further aspect of the present invention, the step of obtaining the set of synchronization lag nodes specifically includes: S111: Obtain the work list of structures in the water conservancy project, record the planned start and end times and the real-time start and end times recorded on site, calculate the time interval coverage ratio of the task work nodes, compare the overlap interval length between the real-time work time period recorded on site and the planned time period, calculate the ratio of the overlap interval length to the total length of the planned work time period, and obtain the coverage time ratio factor of the work nodes. S112: Call the coverage time ratio factor, compare the planned arrangement position of the node tasks with the corresponding start order, determine the offset direction and offset amount, filter the task nodes that start early or late, and generate the job point offset set structure. S113: Call the job point offset set structure to obtain the planned start time, start time, planned duration and duration of the task, and calculate the disturbance amplitude of the task nodes according to the task index variable. Identify the task nodes that have early start, compressed execution and asynchronous completion phenomena, calculate the offset synchronization disturbance amount of the task nodes, and compare the offset synchronization disturbance amount with the set offset intensity threshold to obtain the set of synchronous lag nodes.

[0009] As a further aspect of the present invention, the step of obtaining the chained progress lag path table specifically includes: S211: Based on the set of synchronous lagging nodes, extract the node numbers at the intersection of multiple tasks, count the upstream preceding task identifiers and downstream starting task identifiers associated with the node numbers, construct the structural mapping relationship between tasks, and obtain the intersection task connection time sequence table. S212: Call the intersection task connection timing table, calculate the time difference between the real-time completion time of the preceding task and the real-time start time of the downstream task in the connection pair, and number and mark the arrangement order according to the relationship of the time difference to obtain the intersection segment time difference number sequence. S213: Based on the time difference numbering sequence of the intersection section, detect the changing trend of the time difference value corresponding to any three consecutive numbering segments, identify the numbering segments that show a continuous increasing trend, calculate the path trend intensity index, select the consecutive numbering paths whose lengthening path trend intensity index exceeds the set path mark threshold, and obtain the chain-like progress lag path table.

[0010] As a further aspect of the present invention, the step of obtaining the spatial compression response zone identifier list specifically includes: S311: Based on the chain-like progress lag path table, extract the spatial operation unit number of the task, merge the task time information and spatial range information, use the task's planned execution start and end time and spatial boundary parameters to construct a joint data frame of time interval and spatial location, and use clustering to identify the tasks according to time overlap and spatial proximity to obtain a spatiotemporally related task combination cluster. S312: Call the spatiotemporal associated task combination cluster, detect the spatial overlap rate parameter of the tasks in the combination within the same time period, and compare the spatial overlap rate parameter with the set spatial overlap rate judgment threshold. If the spatial overlap rate exceeds the spatial overlap rate judgment threshold, record the task combination identifier, and determine the process logic relationship of the task combination in the original plan. Filter the combinations with parallel logic relationship to obtain the parallel overlapping task combination set. S313: Using the aforementioned set of parallel overlapping tasks, perform a sequence comparison analysis on the trend parameters of the task duration change in the combination, identify two or more tasks that exhibit a shortened duration during the real-time execution period, and if the judgment condition is met, extract the spatial operation unit number of the corresponding combination and generate a list of spatial compression response zone identifiers.

[0011] As a further aspect of the present invention, the steps for obtaining the multidimensional risk transition point list are as follows: S411: Based on the set of synchronous lag nodes, the chained progress lag path table, and the spatial compression response zone identifier list, extract the node number and spatial number corresponding to the node, and construct the node number statistical sequence and the spatial number statistical sequence respectively. Count and accumulate the number items in the two types of sequences to generate the number frequency statistical results. S412: Call the frequency statistics results of the number, filter and judge whether the node number appears repeatedly in the chain-like progress lag path table and the spatial compression response zone identifier list, extract the duplicate node identifier by the intersection of node number index, and obtain the start-up sequence of the corresponding node in three consecutive construction cycles to generate the cycle start-up change trend. S413: Using the aforementioned periodic start-up change trend, perform a two-stage direction judgment on the period difference in the node start-up number change vector. If the node shows a continuous downward trend in the two periods, then classify the node number as a jump-sensitive type and generate a multi-dimensional risk jump point list.

[0012] As a further aspect of the present invention, a periodic threshold judgment is performed on the number items in the node number statistical sequence. The periodic threshold is a predetermined percentage threshold of the frequency of occurrence of the node number in multiple construction cycles. If the frequency of occurrence of the node number exceeds the periodic threshold, the node number is marked as an abnormal node. The abnormal nodes are filtered by the intersection of the node number index in the chained progress lag path table and the spatial compression response zone identifier list to verify whether the abnormal nodes have an abnormal change trend across cycles. If the abnormal node fluctuates in multiple cycles, the abnormal node is determined to be a potential risk node. The risk level of the potential risk node is assessed by the period difference in the periodic start-up trend. The assessment criterion is the amount of change of the period difference within a predetermined range, and the potential risk node is listed as a risk node.

[0013] As a further aspect of the present invention, the method further includes step S5: S5: Call the multidimensional risk transition point list, extract the number of task paths connected to the node and the frequency of resource calls, and classify the risk level according to whether the number of task paths exceeds three and the frequency of resource calls is more than twice a day. If both conditions are met, the risk level is marked as a level three warning. If either condition is met, it is marked as a level two warning. If neither condition is met but the node is a node in the multidimensional risk transition point list, it is marked as a level one warning. Generate a construction progress risk warning level status set. The construction progress risk warning level status set includes node risk level labels, task path quantity level, and resource call frequency level.

[0014] As a further aspect of the present invention, the steps for obtaining the construction progress risk warning level status set are specifically as follows: S511: Call the multidimensional risk jump point list, extract the set of task paths connected to the jump nodes, count the number of task paths, extract the resource call log data associated with the nodes, classify and count the resource call timestamps, and generate the node path number and resource call frequency statistics results. S512: Using the statistical results of the number of node paths and the frequency of resource calls, based on the two criteria of whether the number of task paths exceeds the threshold for determining the number of paths and whether the frequency of resource calls is higher than the threshold for determining the frequency of resources, the nodes are judged by a combination of conditions and assigned a corresponding warning level label. Nodes that meet both conditions are labeled as Level 3, and nodes that meet either condition are labeled as Level 2, thus generating a preliminary judgment set of node warning levels. S513: Using the aforementioned node early warning level preliminary judgment set, perform continuous jump state verification on nodes that are not marked as level two or three. If a node is a continuous jump node in the multidimensional risk jump point list, then supplement it with a level one early warning level. Summarize and encode the early warning identifiers of the nodes to generate a construction progress risk early warning level state set.

[0015] Compared with the prior art, the advantages and positive effects of the present invention are as follows: In this invention, by comprehensively analyzing the coverage ratio of task execution intervals, start-up offsets, and asynchronous task completion phenomena, dynamic identification of key construction nodes is formed. Combined with the progressive changes in the interval trends between nodes, continuous lag structure annotation at the path level is formed. Intersection analysis of spatial location and task execution time period is introduced to identify task clusters that are executed in a compressed manner within the same time period. The trend of node start-up frequency changes and the overlap of spatial distribution are summarized and statistically analyzed to achieve automatic classification of identification of sensitive points of abrupt change and level warning. The processing flow integrates multi-dimensional judgment criteria such as time coverage comparison, chain path determination, and spatial overlap analysis to enhance the ability to identify unplanned fluctuations in construction nodes and the dynamic response capability, effectively improving the identification accuracy of potential construction progress anomalies and the rationality of warning level stratification. Attached Figure Description

[0016] Figure 1 This is a schematic diagram of the main steps of the present invention; Figure 2 This is a flowchart illustrating the process of obtaining the set of synchronization lag nodes in this invention. Figure 3 This is a flowchart of the process for obtaining the chain-like progress lag path table of the present invention. Figure 4 This is a flowchart illustrating the process of obtaining the spatial compression response zone identifier list of the present invention. Figure 5 This is a flowchart illustrating the process of obtaining the multidimensional risk transition point list of the present invention. Figure 6 This is a flowchart illustrating the process of obtaining the construction progress risk warning level status set for this invention. Detailed Implementation

[0017] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0018] In the description of this invention, it should be understood that the terms "length," "width," "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "top," "bottom," "inner," and "outer," etc., indicating orientation or positional relationships, are based on the orientation or positional relationships shown in the accompanying drawings and are only for the convenience of describing the invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation, and therefore should not be construed as a limitation of the invention. Furthermore, in the description of this invention, "a plurality of" means two or more, unless otherwise explicitly specified.

[0019] Please see Figure 1This invention provides a technical solution: a method for real-time assessment and early warning of construction progress in water conservancy projects, comprising the following steps: S1: Obtain the work list of structures in the water conservancy project, record the planned start and end times and the real-time start and end times recorded on site, compare the coverage ratio of real-time time within the planned time period for the task construction time execution interval, record the advance or delay of the task start order in the planned arrangement, extract the construction nodes with early start, compressed execution and asynchronous completion phenomena, and generate a set of synchronous and lagging nodes. S2: Based on the set of synchronous lag nodes, extract the node numbers belonging to the intersection of multiple tasks, read the real-time completion time of the preceding task and the real-time start time of the node, mark the order of the intervals before and after with numbers, and determine whether the trend of the interval value change within three consecutive numbers shows a positive lengthening. If the condition is met, mark the path as a chain lag path and generate a chain progress lag path table. S3: Using the chained progress lag path table, extract the spatial operation unit number of the task, extract the spatial location range and real-time execution period of the task, extract task combinations with a spatial overlap rate of more than one-third within the same time period, determine whether the task combination is a parallel process in the original plan, and record the operation duration compression trend of the task combination. When two or more tasks in the combination are simultaneously executed in a compressed manner, the spatial region number is used as the response region to generate a list of spatial compression response area identifiers. S4: Based on the set of synchronous lag nodes, the chained progress lag path table, and the spatial compression response zone identifier list, count the number of times the node number and spatial number appear, determine whether the same node appears repeatedly in the chained progress lag path table and the spatial compression response zone identifier list, identify whether the number of times the node is started decreases within three construction cycles, and if the downward trend occurs twice in a row, define the node as a jump sensitive node and generate a multi-dimensional risk jump point list; S5: Call the multi-dimensional risk transition point list, extract the number of task paths connected to the node and the frequency of resource calls, and classify the risk level according to whether the number of task paths exceeds three and the frequency of resource calls is more than twice a day. If both conditions are met, the risk level is marked as a level three warning. If either condition is met, it is marked as a level two warning. If neither condition is met but the node is a node in the multi-dimensional risk transition point list, it is marked as a level one warning. Generate a construction progress risk warning level status set. The set of synchronous lag nodes includes task coverage ratio, task start offset, and synchronous lag node number; the chained progress lag path table includes chain path number, node interval trend label, and lag path sequence; the spatial compression response zone identifier list includes spatial region number, compression task combination number, and number of compression execution tasks; the multidimensional risk jump point list includes duplicate node number, cycle start frequency trend, and jump sensitive node number; and the construction progress risk warning level status set includes node risk level label, task path quantity level, and resource call frequency level.

[0020] Please see Figure 2 The specific steps for obtaining the set of synchronization lag nodes are as follows: S111: Obtain the work list of structures in the water conservancy project, record the planned start and end times and the real-time start and end times recorded on site, calculate the time interval coverage ratio of the task work nodes, compare the overlap interval length between the real-time work time period recorded on site and the planned time period, calculate the ratio of the overlap interval length to the total length of the planned work time period, and obtain the coverage time ratio factor of the work nodes. Based on the job types listed in the work list, the planned start and end times for each job task are extracted. Simultaneously, the actual start and end times of each job point are retrieved from the monitoring equipment at the construction site. For cases where the planned and actual times do not coincide, the intersection interval is calculated, and the length of the intersection interval is recorded as the actual coverage time of the job point. Using the planned job time length as a benchmark, the ratio of the intersection interval to the planned time interval is calculated using the formula... Obtain time coverage ratio parameters in the form of The project was set up during sluice gate construction, with a planned operation period from July 1st to July 10th, 2025, lasting 10 days. On-site records show that the actual operation took place from July 3rd to July 9th, resulting in an overlap of 7 days and a time coverage ratio of [missing information]. For the ratio, call the number of each work point in the work plan list. Complete the proportional calculation for all work points, and store the obtained proportional results in a mapping vector. In order to prevent data anomalies from causing the coverage ratio to become invalid, coverage ratio values ​​below 0.2 or above 1.1 are removed. The removal condition is set as follows: if or If an outlier is detected, the data at that work point is marked as an outlier and removed from subsequent calculations. The result after removal is used to calculate the work time coverage ratio factor.

[0021] S112: Call the coverage time scaling factor, compare the planned arrangement of node tasks with the corresponding start order, determine the offset direction and offset amount, filter task nodes that start early or late, and generate a set structure of job point offsets. Call the time coverage scaling factor sequence corresponding to the job point By comparing the ranking of each work point on the timeline, the planned ranking number can be obtained. The actual startup order Execute order difference operation ,like If it is, it means that the work point has a delayed start; otherwise, if This indicates that the work point is executed ahead of schedule, and a directional judgment is performed on the sequence offset value, setting the directionality as a variable. When the offset is greater than 0, it is set to "extend"; when it is less than 0, it is set to "lift". The offset direction is stored in an array. If the planned sequence for job point 3 is 5, but the actual start position is 3, then the offset value is... The direction is "lift", and it is stored in the offset set according to the offset value. The magnitude of the offset is divided into three levels: "mild," "moderate," and "severe," with the intervals set as follows: It is mild. Moderate. For severe cases, set This is considered a minor offset. Based on the combination of offset direction and offset degree, the job point offset type is categorized as "Minor Advance," generating a job point offset set structure. As shown in Table 1; Table 1: Offset Direction and Classification Table ; Table 1 lists the deviations that occurred after comparing the actual execution of the work points with the planned sequence.

[0022] S113: Call the job point offset set structure to obtain the planned start time, start time, planned duration, and duration of the task. Calculate the perturbation amplitude for each task node based on the task index variable to identify task nodes exhibiting early start, compressed execution, or asynchronous completion. The formula is as follows: ; Calculate the offset synchronization disturbance of the task nodes, compare the offset synchronization disturbance with the set offset intensity threshold, and obtain the set of synchronization lag nodes; in, Representing the Offset synchronization disturbance of each task node. Indicates the first The actual startup time of each task node. Indicates the first The planned start time for each task node. Indicates the first Coverage time scaling factor for each task node Indicates the first The real-time duration of each task node. Indicates the first The planned duration of each task node; The formula calculation logic is as follows: Divide the difference between the actual start time and the planned start time by the planned start time, and perform a relative offset absolute value calculation to characterize whether the task has shifted overall on the timeline. The square root of the sum of the squared deviation between the time coverage ratio and 1 and the squared deviation between the actual duration and the planned duration is used to measure the comprehensive offset of the task in terms of coverage and duration. The two are added together to form the disturbance value. The first part reflects the forward and backward movement trend of the timeline position, and the second part reflects the integrity and consistency deviation of the task execution process. The entire calculation process is standardized to achieve horizontal quantitative comparison of different tasks, ensuring the comparability of disturbance indicators under the same evaluation scale, and is used for subsequent screening and discrimination operations in the process of identifying synchronous task nodes and adjusting plans. The offset synchronization disturbance is a quantitative indicator that measures the degree of deviation of a single task node in time execution. It combines the task start time offset, actual execution coverage and execution time difference from the plan. The larger the value, the more serious the synchronization offset of the node. It can be used to screen critical operation points that need to be adjusted in advance or postponed. Parameter explanation and formula calculation process: : No. The actual start time of each task point is derived from the timestamps recorded by the construction monitoring equipment, and the unit is days. : No. The planned start time for each task point is extracted from the task list; : No. The time coverage ratio of each task point; : No. The actual duration of each task point is calculated as the actual termination time minus the actual start time. : No. The planned duration of each task point, and the planned duration in the task list; Taking work point 3 as an example, its parameters are as follows: ; ; (i.e., from July 3 to July 8); (i.e., from July 5th to July 11th); ; Substitute into the formula to calculate: ; ; ; ; The result shows that the exponent value of operation point 3 in the synchronization disturbance dimension is 0.7431. According to the preset threshold setting, if For good synchronization, It is a medium offset. If the synchronization offset is severe, then work point 3 is classified as "severe synchronization offset"; The advantage of the formula is that by introducing the coupling calculation of three factors—time offset rate, time coverage ratio deviation, and planned / actual duration cycle deviation—it can form a quantitative assessment of the comprehensive impact of time variation while ensuring the integrity of the structural unit. This is applicable to the quantitative management of time sensitivity in multi-node coordinated construction tasks.

[0023] Please see Figure 3 The specific steps for obtaining the chained progress lag path table are as follows: S211: Based on the set of synchronous lagging nodes, extract the node numbers at the intersection of multiple tasks, count the upstream predecessor task identifier and downstream startup task identifier associated with the node number, construct the structural mapping relationship between tasks, and obtain the intersection task connection time sequence table. Extract the unique identifier of each task node and locate its intersection position in the task topology. By scanning the task dependency graph structure, trace back the upstream task code of each lagging task node to confirm the preceding task number to which each node depends. Then, traverse down each upstream task node to collect the downstream task identifiers connected to it, construct a complete task dependency link, and form a bidirectional task relationship mapping table. Then, combine the task node pairs that are upstream and downstream to form a task intersection segment according to the task execution sequence, and record the numbers of the task nodes on both sides of the intersection point to obtain the path sequence of the task intersection connection. In a practical application, in a dam reinforcement formwork construction task, node number T7 is a lagging node, upstream nodes are T4 and T5, and downstream nodes are T9 and T10. Then the task intersection segment can form a path chain of T4→T7→T9 and T5→T7→T10, and the intersection task connection sequence table is obtained.

[0024] S212: Call the intersection task connection timing table, calculate the time difference between the real-time completion time of the preceding task and the real-time start time of the downstream task in the connection pair, and number and mark the arrangement order according to the relationship of the time difference to obtain the time difference number sequence of the intersection segment. For each task connection path, the time difference between adjacent task nodes is calculated pairwise. The actual time interval of the connection segment is obtained by subtracting the actual completion time of the preceding node from the actual start time of the subsequent node. The time intervals of the path segments are then vectorized and stored to form a difference sequence. The paths are sorted using their path numbers as indexes to generate a time difference coding sequence for intersection segments. In actual construction projects, for path T4→T7, if T4 is completed on the 18th day and T7 starts on the 20th day, the time difference for this path segment is 2 days. Similarly, if T5→T7 is 1 day, T7→T9 is 3 days, and T7→T10 is 2 days, the time difference sequence is 2, 1, 3, 2, recorded as node connection numbers T4-T7, T5-T7, T7-T9, T7-T10. A continuous coding label is assigned to the sequence according to the path topology. This is used to quickly locate the direction of time difference changes between path segments in the subsequent trend judgment stage, and obtain the time difference number sequence of the intersection segment.

[0025] S213: Based on the time difference numbering sequence of the intersection segment, detect the changing trend of the time difference value corresponding to any three consecutive numbered segments, identify the numbered segments that show a continuous increasing trend, using the formula: ; Calculate the path trend intensity index, select consecutive numbered paths whose extended path trend intensity index exceeds the set path marking threshold, and obtain a chain-like progress lag path table. in, Indicators representing the strength of path trends Indicates the number of paragraphs. The time difference between each task connection pair To indicate the numbered paragraphs The logical value when the condition is met is 1, and the value when the condition is not met is 0. The length of the time difference number sequence; The calculation logic of the formula is as follows: Based on the analysis of the changing trend of the time difference of the path segments in the task intersection area, three consecutive path segments are used as a sliding window. The sum of the absolute values ​​of the time difference between the first two segments and the last two segments is calculated to reflect the total magnitude of the time difference change. This magnitude is then normalized to the denominator, which is the square root of the sum of the squares of the time difference between the two segments, forming a standardized coefficient to avoid disturbance expansion due to the large absolute time magnitude. The overall result is a fractional growth rate expression, which is then multiplied by the judgment logic quantity. Ensure that only Only when the conditions are met The role of the effective contribution logic quantity is to filter out the trend segments with continuous growth, and to sum the indicator values ​​of the effective trend segments to form the overall path trend strength indicator. This indicates whether the path exhibits a continuously increasing time difference expansion phenomenon. The larger the value, the more segments in the path show a continuous lengthening trend, and the stronger the increase, reflecting the significant extension of the path in the time dimension. The path trend strength index is a numerical parameter that measures whether the continuous time difference in a task path shows an increasing trend. This index is calculated by the direction and magnitude of the change in three consecutive time differences. Only when the time difference increases continuously is the result included. The higher the value, the more obvious the time delay trend of the path segment. It can be used to identify potential chain-like lag paths. Based on the time difference sequence of consecutive path segments recorded in the intersection section Retrieve three consecutive numbered path segments using a sliding window method, and then... , , The direction is determined by the time difference of the three path segments. The comparison operation between two differences is used to convert the magnitude of the time difference trend change into a parameter in the formula, and combined with logic to determine whether a continuous increasing trend has formed. Parameter meaning and calculation process: For the first The time difference between each connection segment For Boolean logic variables, if the following conditions are met... Conditions, then set Otherwise, it is 0, in the instance sequence. For example, Then substitute into the formula to calculate: Group 1: ; ; ; Group 2: ; ; ; Substitute into the formula to calculate: ; The results show that the path trend strength index is 1.3416, which is a quantitative measure of the extent of path trend extension. A higher value indicates a more significant trend increase between path segments. To clearly identify chain-like delayed paths, a trend strength threshold is set in this embodiment. ,when The time marker is used to mark the trend extension path and added to the set of chained lag paths, as in this example. Therefore, this path sequence was identified as a trend-elongating path that needs to be closely monitored. Table 2 lists the parameters and calculation process for this example: Table 2: Calculation Table of Path Segment Trend Indicators ; Table 2 lists the parameters and calculation indicators required for trend path judgment, which can be used as a basis for screening and judging chain-delayed task paths.

[0026] Please see Figure 4 The specific steps for obtaining the list of space compression response area identifiers are as follows: S311: Based on the chained progress lag path table, extract the spatial operation unit number of the task, merge the task time information and spatial range information, use the task's planned execution start and end time and spatial boundary parameters to construct a joint data frame of time interval and spatial location, and use clustering to identify the tasks according to time overlap and spatial proximity to obtain a spatiotemporally related task combination cluster. Extracting the spatial operation unit number of the task, the time information and spatial range information of the task need to be merged. The main purpose is to associate the execution time interval of each task with the spatial coverage area, and construct a joint data frame combining time and space information. By analyzing the planned execution start and end time and spatial boundary parameters of the task, the time period of each task is associated with the spatial location, forming a spatiotemporal joint data frame. In the construction project, the construction task is carried out within the time period from September 1 to September 5, 2025. At the same time, the spatial area of ​​the task execution is located in the southeast corner of the building. The specific spatiotemporal attribute data of each task can be constructed. The tasks are combined and identified by the temporal overlap and spatial proximity of the tasks. Set up on the construction site, the construction and monitoring tasks have intersections in time and space. In this way, the overlapping time periods and close spatial locations in the task execution process can be identified, and spatiotemporally associated task combination clusters can be obtained.

[0027] S312: Call the spatiotemporal related task combination cluster, detect the spatial overlap rate parameter of the tasks in the combination within the same time period, and compare the spatial overlap rate parameter with the set spatial overlap rate judgment threshold. If the spatial overlap rate exceeds the spatial overlap rate judgment threshold, record the task combination identifier, determine the process logic relationship of the task combination in the original plan, filter the combinations with parallel logic relationship, and obtain the parallel overlapping task combination set. The detection function checks the spatial overlap rate of tasks within the same time period. Spatial overlap rate indicates whether and to what extent the spatial areas of multiple tasks overlap within the same time frame. For example, if task A and task B are scheduled for September 1st to September 5th, and task A's spatial area is the southeast corner of a building while task B's spatial area is the northeast corner, then the spatial overlap rate is 0%. If the spatial areas of the two tasks intersect, the spatial overlap rate needs to be calculated. This can be done using geometric methods, such as calculating the intersection of the two spatial areas. The spatial overlap rate is then compared to a pre-set threshold of 30%. If the overlap rate exceeds the threshold, the task combination identifier is recorded, and the task's procedural logic is analyzed. Task combinations with parallel logical relationships are selected, meaning that tasks can be executed in parallel without conflict, reducing waiting time during construction and improving work efficiency. In a construction project, if task A and task B have 30% spatial overlap within the same time period and have no dependency relationship, a set of parallel overlapping task combinations is obtained.

[0028] S313: Using a set of parallel overlapping tasks, perform a sequence comparison analysis on the trend parameters of the task duration change in the combination, identify two or more tasks that show a shortened duration during the real-time execution period, and if the judgment condition is met, extract the spatial operation unit number of the corresponding combination and generate a list of spatial compression response area identifiers. It is necessary to perform a sequential comparative analysis of the duration change trends of each task in the task combination. For each task, the actual execution time can be recorded, and then the duration change trend in different time periods can be observed. Set the execution time of task A to 5 hours on September 1, 6 hours on September 2, and 4 hours on September 3. If some tasks show a trend of shortening duration during execution, this method can be used to identify them. For example, the execution time of task A is long at the beginning, but as the project progresses, task A is gradually optimized and the duration shortens. By comparing the duration of tasks, it is possible to identify which tasks meet the condition of shortening duration. If the duration shortening trend of the task combination meets the judgment condition, the spatial operation unit number of the corresponding combination can be extracted to identify the task that shows a compressed time period during execution. In subsequent scheduling, spatial optimization and time rearrangement can be carried out to improve work efficiency and task completion accuracy. Set task A and task B to be executed in the same time period. After comparing the time change trend, it is found that the duration of task B is gradually shortening. After meeting the condition, the corresponding spatial operation unit number is extracted to generate a list of spatial compression response area identifiers.

[0029] Please see Figure 5 The specific steps for obtaining the multidimensional risk transition point list are as follows: S411: Based on the set of synchronous lag nodes, the chained progress lag path table, and the spatial compression response zone identifier list, extract the node number and spatial number corresponding to the node, and construct the node number statistical sequence and the spatial number statistical sequence respectively. Count and accumulate the number items in the two types of sequences to generate the number frequency statistical results. Extract the node number and spatial number corresponding to each node. The node number represents a unique identifier for a task or process, while the spatial number represents the physical spatial location of the node. In this process, the task obtains the number and corresponding spatial information of each node by parsing the set of synchronous lagging nodes. Two statistical sequences are constructed: one is the node number statistical sequence, and the other is the spatial number statistical sequence. The two sequences are used to record the frequency of each node number and spatial number in task scheduling. The number items in these two sequences are counted and accumulated to determine the number of times each node and spatial location appears in the entire plan. For example, in a construction task, the task with node number "N001" appears 5 times at the spatial number "S001", and the task with node number "N002" appears 3 times at the spatial number "S001". By counting, the frequency of the number can accurately reflect the usage of each node and space, generating the number frequency statistics.

[0030] S412: Call the frequency statistics results of the node number, filter and judge whether the node number appears repeatedly in the chained progress lag path table and the spatial compression response area identifier list, extract the duplicate node identifier by using the intersection of node number index, and obtain the start-up sequence of the corresponding node in three consecutive construction cycles to generate the cycle start-up change trend. The system filters and determines whether each node number appears repeatedly in the delayed path and compressed region. The delayed path refers to the path where there is a delay during task execution, and the compressed region refers to the region where time compression occurs during task scheduling. By filtering the path and region, it can be analyzed whether a node appears repeatedly in the region. If a node is found to appear multiple times in the same cycle and in the delayed path or compressed region, then the node needs to be marked as a duplicate node. The duplicate node identifier is extracted by the intersection of the node number index, and the start-up sequence of the node in three consecutive construction cycles is obtained. If node "N001" starts 3 times in the first cycle, 4 times in the second cycle, and 2 times in the third cycle, then the start-up sequence of the node is [3, 4, 2]. Using the start-up sequence, the start-up trend of the node in each cycle can be analyzed to generate the cycle start-up change trend.

[0031] S413: Utilize the trend of periodic start-up changes to make a two-stage directional judgment on the period difference in the vector of node start-up frequency changes. If a node shows a continuous downward trend in two periods, the node number is classified as a jump-sensitive type, and a multi-dimensional risk jump point list is generated. By performing a two-stage directional judgment on the period difference in the vector of node start-up count changes, the period difference refers to the difference in the number of node start-ups between two adjacent periods. By calculating the difference, it can be determined whether there is a downward trend in the number of node start-ups between two periods. If the start-up count sequence of node "N001" is set to [3, 4, 2], then the difference between the first and second periods is 4-3=1, indicating an increase in the number of start-ups; while between the second and third periods, the difference is 2-4=-2, indicating a decrease in the number of start-ups. If the number of node start-ups shows a downward trend in two consecutive periods, then the node will be judged as a jump-sensitive type, the node number will be classified into the jump-sensitive type, and a multi-dimensional risk jump point list will be generated.

[0032] Please see Figure 6 The specific steps for obtaining the construction progress risk warning level status set are as follows: S511: Call the multi-dimensional risk jump point list, extract the set of task paths connected to the jump nodes, count the number of task paths, extract the resource call log data associated with the nodes, classify and count the resource call timestamps, and generate the node path quantity and resource call frequency statistics results. Extract the set of task paths connected to the jump node. The task path connected to the jump node refers to the path involved in the task execution process of the node. The path is the dependency relationship between nodes or the temporal order relationship in task execution. The number of task paths will be counted, that is, how many task paths are passed through or affected by the jump node. If node "N003" is set to connect three task paths, then the number of paths is 3. Resource call log data associated with the node will be extracted. The resource call log contains the resource usage of the node during execution. By classifying and statistically analyzing the resource call timestamps, we can understand the frequency and time distribution of resource usage of the node. In a construction project, node "N003" needs to call different equipment or personnel resources at different times. The frequency of resource calls within a specific time period will be counted, and the results of the node path count and resource call frequency statistics will be generated.

[0033] S512: Using the statistical results of the number of node paths and the frequency of resource calls, based on two criteria, whether the number of task paths exceeds the threshold for determining the number of paths and whether the frequency of resource calls is higher than the threshold for determining the frequency of resources, the nodes are judged by a combination of conditions and assigned a corresponding warning level label. Nodes that meet both conditions are labeled as Level 3, and nodes that meet either condition are labeled as Level 2, thus generating a preliminary judgment set of node warning levels. The system uses two criteria for judgment. If the number of task paths exceeds a threshold, it indicates high task path complexity and execution risk for that node. The threshold is set to 5. If a node has 6 task paths, it exceeds this threshold, indicating significant task pressure. The system also checks if the resource call frequency exceeds a threshold. Excessive resource call frequency indicates resource conflicts or delays due to high-density resource calls. A preset threshold (more than 10 calls per hour) indicates overload risk. These two criteria are used to judge task complexity and resource pressure, respectively. A binary logic classifier is built based on these two conditions. The classifier performs conditional judgments on nodes, assigning different warning levels based on the degree to which the conditions are met. If a node meets both the path count and resource call frequency criteria, it is marked as a Level 3 warning, indicating high risk. If either condition is met, the node is marked as a Level 2 warning, indicating some risk. This generates an initial warning level set for the nodes.

[0034] S513: Using the node early warning level preliminary judgment set, perform continuous jump state verification on nodes that are not marked as level two or three. If a node is a continuous jump node in the multidimensional risk jump point list, then mark it as level one early warning level. Summarize and encode the early warning identifiers of the nodes to generate a construction progress risk early warning level status set. For nodes not marked as Level 2 or Level 3, continuous jump status verification is performed. Continuous jump status verification refers to determining whether a node has a continuous jump trend in the multi-dimensional risk jump point list. If node "N003" shows continuous state jumps (the number of activations fluctuates continuously) in multiple cycles, the node will be identified as a continuous jump node. For continuous jump nodes, they will be supplemented to Level 1 warning level, indicating that the node has potential risks that will have a significant impact on the overall construction progress. After being supplemented as Level 1 warning, the warning identifiers of the nodes are summarized, and the warning information of the nodes in the project is summarized. Through the summarized codes, managers can clearly understand the risk status and construction progress risks of each node, providing a comprehensive basis for the monitoring, adjustment and emergency decision-making of the construction project, which helps to take timely measures to prevent potential risks and generate a construction progress risk warning level status set.

[0035] The above are merely preferred embodiments of the present invention and are not intended to limit the present invention in any other way. Any person skilled in the art may make changes or modifications to the above-disclosed technical content to create equivalent embodiments that can be applied to other fields. However, any simple modifications, equivalent changes, and modifications made to the above embodiments based on the technical essence of the present invention without departing from the scope of the present invention shall still fall within the protection scope of the present invention.

Claims

1. A method for real-time assessment and early warning of construction progress in water conservancy projects, characterized in that, Includes the following steps: S1: Obtain the work list of structures in the water conservancy project, record the planned start and end times and the real-time start and end times recorded on site, compare the coverage ratio of real-time time within the planned time period, record whether the task start order is ahead or behind in the planned arrangement, and generate a set of synchronous lag nodes. The specific steps are as follows: S111: Obtain the work list of structures in the water conservancy project, record the planned start and end times and the real-time start and end times recorded on site, calculate the time interval coverage ratio of the task work nodes, compare the overlap interval length between the real-time work time period recorded on site and the planned time period, calculate the ratio of the overlap interval length to the total length of the planned work time period, and obtain the coverage time ratio factor. S112: Call the coverage time ratio factor, compare the planned arrangement position of the node tasks with the corresponding start order, determine the offset direction and offset amount, filter the task nodes that start early or late, and generate the job point offset set structure. S113: Call the work point offset set structure to obtain the planned start time, start time, planned duration and duration of the task, and calculate the disturbance amplitude of the task nodes according to the task index variable. Identify the task nodes that have early start, compressed execution and asynchronous completion phenomena, calculate the offset synchronization disturbance amount of the task nodes, and compare the offset synchronization disturbance amount with the set offset intensity threshold to obtain the set of synchronous lag nodes. S2: Based on the set of synchronous lag nodes, extract the node numbers belonging to the multi-task intersection point, read the real-time completion time of the preceding task and the real-time start time of the node, mark the order of the preceding and following intervals with numbers, and generate a chain-like progress lag path table. The specific steps are as follows: S211: Based on the set of synchronous lagging nodes, extract the node numbers at the intersection of multiple tasks, count the upstream preceding task identifiers and downstream starting task identifiers associated with the node numbers, construct the structural mapping relationship between tasks, and obtain the intersection task connection time sequence table. S212: Call the intersection task connection timing table, calculate the time difference between the real-time completion time of the preceding task and the real-time start time of the downstream task in the connection pair, and number and mark the arrangement order according to the relationship of the time difference to obtain the intersection segment time difference number sequence. S213: Based on the time difference numbering sequence of the intersection section, detect the changing trend of the time difference value corresponding to any three consecutive numbering segments, identify numbering segments that show a continuous increasing trend, calculate the path trend intensity index, select consecutive numbering paths whose lengthened path trend intensity index exceeds the set path marking threshold, and obtain a chain-like progress lag path table. S3: Using the chained schedule lag path table, extract the spatial work unit number of the task, extract the spatial location range and real-time execution time of the task, determine whether the task combination is a parallel process in the original plan, and generate a spatial compression response zone identifier list. The specific steps are as follows: S311: Based on the chain-like progress lag path table, extract the spatial operation unit number of the task, merge the task time information and spatial range information, use the task's planned execution start and end time and spatial boundary parameters to construct a joint data frame of time interval and spatial location, and use clustering to identify the tasks according to time overlap and spatial proximity to obtain a spatiotemporally related task combination cluster. S312: Call the spatiotemporal associated task combination cluster, detect the spatial overlap rate parameter of the tasks in the combination within the same time period, and compare the spatial overlap rate parameter with the set spatial overlap rate judgment threshold. If the spatial overlap rate exceeds the spatial overlap rate judgment threshold, record the task combination identifier, and determine the process logic relationship of the task combination in the original plan. Filter the combinations with parallel logic relationship to obtain the parallel overlapping task combination set. S313: Using the aforementioned set of parallel overlapping tasks, perform a sequence comparison analysis on the trend parameters of the task duration change in the combination, identify two or more tasks that exhibit a shortened duration during the real-time execution period, and if the judgment condition is met, extract the spatial operation unit number of the corresponding combination and generate a list of spatial compression response area identifiers. S4: Based on the set of synchronous lag nodes, the chained progress lag path table, and the spatial compression response zone identifier list, count the occurrences of node numbers and spatial numbers, determine whether the same node appears repeatedly in the chained progress lag path table and the spatial compression response zone identifier list, identify whether the number of times a node is started decreases within three construction cycles, and generate a multidimensional risk jump point list. The specific steps are as follows: S411: Based on the set of synchronous lag nodes, the chained progress lag path table, and the spatial compression response zone identifier list, extract the node number and spatial number corresponding to the node, and construct the node number statistical sequence and the spatial number statistical sequence respectively. Count and accumulate the number items in the two types of sequences to generate the number frequency statistical results. S412: Call the frequency statistics results of the number, filter and judge whether the node number appears repeatedly in the chain-like progress lag path table and the spatial compression response zone identifier list, extract the duplicate node identifier by the intersection of node number index, and obtain the start-up sequence of the corresponding node in three consecutive construction cycles to generate the cycle start-up change trend. S413: Using the aforementioned periodic start-up change trend, perform a two-stage direction judgment on the period difference in the node start-up number change vector. If the node shows a continuous downward trend in the two periods, then classify the node number as a jump-sensitive type and generate a multi-dimensional risk jump point list. S5: Call the multidimensional risk transition point list, extract the number of task paths connected to the node and the frequency of resource calls, and classify the risk level according to whether the number of task paths exceeds three and the frequency of resource calls is more than twice a day. If both conditions are met, the risk level is marked as a level three warning. If either condition is met, it is marked as a level two warning. If neither condition is met but the node is a node in the multidimensional risk transition point list, it is marked as a level one warning. Generate a construction progress risk warning level status set. The set of synchronous lag nodes includes task coverage ratio, task start offset, and synchronous lag node number; the chained progress lag path table includes chain path number, node interval trend label, and lag path sequence; the spatial compression response zone identifier list includes spatial region number, compression task combination number, and number of compression execution tasks; the multidimensional risk jump point list includes repeating node number, periodic start frequency trend, and jump sensitive node number; and the construction progress risk warning level status set includes node risk level label, task path quantity level, and resource call frequency level.

2. The method for real-time assessment and early warning of construction progress of water conservancy projects according to claim 1, characterized in that, A periodic threshold is applied to the number items in the node number statistical sequence. The periodic threshold is a predetermined percentage threshold of the frequency of occurrence of the node number in multiple construction cycles. If the frequency of occurrence of the node number exceeds the periodic threshold, the node number is marked as an abnormal node. The abnormal nodes are filtered by the intersection of the node number index in the chained progress lag path table and the spatial compression response zone identifier list to verify whether the abnormal nodes have an abnormal change trend across cycles. If the abnormal node fluctuates in multiple cycles, the abnormal node is determined to be a potential risk node. The risk level of the potential risk node is assessed by the period difference in the periodic start-up trend. The assessment criterion is the amount of change of the period difference within a predetermined range, and the potential risk node is listed as a risk node.

3. The method for real-time assessment and early warning of construction progress of water conservancy projects according to claim 1, characterized in that, The specific steps for obtaining the construction progress risk warning level status set are as follows: S511: Call the multidimensional risk jump point list, extract the set of task paths connected to the jump nodes, count the number of task paths, extract the resource call log data associated with the nodes, classify and count the resource call timestamps, and generate the node path number and resource call frequency statistics results. S512: Using the statistical results of the number of node paths and the frequency of resource calls, based on the two criteria of whether the number of task paths exceeds the threshold for determining the number of paths and whether the frequency of resource calls is higher than the threshold for determining the frequency of resources, the nodes are judged by a combination of conditions and assigned a corresponding warning level label. Nodes that meet both conditions are labeled as Level 3, and nodes that meet either condition are labeled as Level 2, thus generating a preliminary judgment set of node warning levels. S513: Using the aforementioned node early warning level preliminary judgment set, if the multidimensional risk jump point list is a continuously jumping node, then it is additionally marked as a first-level early warning level. The early warning identifiers of the nodes are summarized and coded to generate a construction progress risk early warning level status set.

Images