Bridge construction risk assessment and data analysis system

Abstract

The application belongs to the technical field of bridge construction management, and relates to a bridge construction risk assessment and data analysis system, which comprises the following steps: a progress network plan is established according to a bridge construction design file, the progress network plan comprises process nodes and constraints before and after the process nodes; when the progress completion ratio of a process changes, a time deviation amount is propagated in the after direction, total float and free float of the affected process are updated, a criticality weight value is generated according to the updated total float, and a progress risk alarm threshold is dynamically generated in combination with a risk lag rate and a remaining free float; a risk alarm signal is triggered when the process connection lag length reaches or exceeds the progress risk alarm threshold; in response to the risk alarm signal, a contribution delay amount of the process before the current process is calculated, and the alarm threshold of the downstream process is recursively adjusted. The application overcomes the defect that a fixed threshold cannot adapt to the dynamic change of float, and realizes real-time self-adaptive adjustment of the construction period risk threshold.

Description

Technical Field

[0001] This invention belongs to the field of bridge construction management technology and relates to a bridge construction risk assessment and data analysis system. Background Technology

[0002] During bridge construction, the risk of schedule delays is one of the factors affecting the timely completion of the project. Currently, the technical solutions related to bridge construction schedule risk assessment mainly involve the following two categories: one is to use a fixed time difference threshold to judge the schedule delay, and to pre-set a fixed lag time alarm threshold, which triggers an alarm when the actual lag exceeds the threshold.

[0003] Second, risk probability assessment is based on historical data statistics. By collecting the construction period data of completed bridges, statistical analysis is used to fit the probability distribution of construction delays, and risk levels are preset accordingly.

[0004] However, the above scheme has the following defects when applied to the dynamic adjustment of the construction period duration threshold for bridge construction: 1. The risk alarm threshold for the construction period duration is fixed and fails to be dynamically linked with the schedule network time difference: The fixed threshold scheme does not consider that the actual construction period deviation of the process will cause the remaining free time difference of the subsequent process to change dynamically. The process that originally had sufficient buffer may be transformed into a high-risk state due to the consumption of buffer. The fixed threshold cannot capture this process, resulting in alarm lag.

[0005] 2. The adjustment mechanism of the schedule alarm threshold is detached from the remaining buffer capacity of the process and fails to achieve adaptive correction: The scheme based on historical statistics is based on the historical distribution characteristics of the threshold, rather than the remaining free float of the current process. The remaining free float is an important indicator for measuring the ability of a process to tolerate additional deviations. Existing technology has not used it to dynamically correct the alarm threshold.

[0006] 3. Lack of quantitative correlation between the actual performance of key construction parameters and the project schedule risk assessment: Existing solutions rely on fixed thresholds or historical accident statistics, and do not quantitatively model the deviation between the actual sampled values ​​and standard values ​​of construction parameters that affect the progress in completed stages, resulting in a lack of foresight in risk assessment. Summary of the Invention

[0007] In view of this, in order to solve the problems mentioned in the background technology, a bridge construction risk assessment and data analysis system is proposed.

[0008] The objective of this invention can be achieved through the following technical solution: This invention provides a bridge construction risk assessment and data analysis system, including: a schedule network construction module, a schedule deviation propagation module, an adaptive threshold generation module, a schedule lag alarm module, and a backtracking analysis module.

[0009] The progress network construction module is connected to the schedule deviation propagation module, the schedule deviation propagation module is connected to the adaptive threshold generation module, the adaptive threshold generation module is connected to the schedule lag alarm module, and the schedule lag alarm module is connected to the backtracking analysis module.

[0010] The schedule network construction module establishes a schedule network plan based on the bridge construction design documents, which includes multiple process nodes and the preceding and succeeding constraints between nodes. Each process node stores the expected duration, progress completion percentage, initial total float value, and initial free float value.

[0011] The schedule deviation propagation module, when the progress completion ratio of any process node changes, propagates the deviation downstream along the direction of the next process based on the schedule deviation of the process, recalculates the total float and free float of each affected next process node, and updates the float parameters in the schedule network plan.

[0012] The adaptive threshold generation module generates the criticality weight value of the corresponding process based on the total float after each process node is updated, and generates the progress risk alarm threshold for the current process by combining the risk lag rate data and the remaining free float.

[0013] The progress delay alarm module collects the delay time of the current process connection in real time. When the delay time of the process connection is greater than or equal to the progress risk alarm threshold, a risk alarm signal is triggered.

[0014] The backtracking analysis module, in response to risk alarm signals, traces backward along the preceding direction, sequentially checks the progress of each preceding process, calculates the contribution of each preceding process to the delay of the current process, and adjusts the risk alarm thresholds of each subsequent process.

[0015] Compared with the prior art, the beneficial effects of the present invention are as follows: (1) The present invention uses a schedule deviation propagation module to propagate the schedule deviation along the successor direction step by step when the progress completion ratio of any process changes, and recalculates the total float and free float of the affected process. Using this dynamic float as input, a criticality weight value is generated, so that the alarm threshold is adjusted in real time according to the actual buffer consumption of the process, solving the problem that a fixed threshold cannot capture non-critical processes turning into high-risk states, and avoiding alarm lag.

[0016] (2) This invention uses the remaining free float of the current process as a buffer index, which directly participates in the calculation and dynamic correction of the alarm threshold. When the remaining free float decreases, the threshold is automatically tightened; when the remaining free float is sufficient, the threshold is appropriately relaxed. At the same time, the contribution delay of the preceding process is transmitted downstream through the backtracking analysis module, and the threshold of the succeeding process is adjusted proportionally according to the actual absorption ratio, so as to achieve real-time matching with the actual progress status.

[0017] (3) This invention identifies key construction parameters that affect the progress, extracts the actual sampling sequence of completed stages, calculates the deviation amplitude and duration, screens parameters with high correlation coefficients with lag duration, and then calculates the risk lag rate. This makes the schedule assessment not only rely on historical statistics or fixed thresholds, but also on the real-time performance of current construction parameters, providing early warnings before significant lags occur. Attached Figure Description

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

[0019] Figure 1 This is a schematic diagram of the module connection of the present invention;

[0020] Figure 2 This is a schematic diagram illustrating the identification logic of key construction parameters affecting the progress of the current process node in this invention;

[0021] Figure 3 This is a schematic diagram illustrating the calculation logic of the delay amount contributed by each preceding process to the delay of the current process in this invention. Detailed Implementation

[0022] 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.

[0023] Please see Figure 1 As shown, the present invention provides a bridge construction risk assessment and data analysis system, including: a schedule network construction module, a schedule deviation propagation module, an adaptive threshold generation module, a schedule lag alarm module, and a backtracking analysis module.

[0024] The progress network construction module is connected to the schedule deviation propagation module, the schedule deviation propagation module is connected to the adaptive threshold generation module, the adaptive threshold generation module is connected to the schedule lag alarm module, and the schedule lag alarm module is connected to the backtracking analysis module.

[0025] The progress network construction module establishes a progress network plan based on the bridge construction design documents, which includes multiple process nodes and the preceding and succeeding constraints between nodes. Each process node stores the expected duration, progress completion percentage, initial value of total float, and initial value of free float.

[0026] In this embodiment, the schedule network plan is established as follows: The construction procedure list and the logical relationship description text between each procedure are read line by line from the bridge construction design document. The logical relationship description includes at least a "complete-start" constraint type. The unique identifier, expected duration, and set of identifiers of the preceding procedures for each procedure are parsed. The specific parsing process is as follows: The construction procedure list text is split into lines, with each line corresponding to one procedure. For each line, a regular expression is used to match the procedure number field to extract the unique identifier of the procedure. The duration field is matched to extract the expected duration, in hours or days. The procedure name field is matched as auxiliary identification information.

[0027] For the current process, find all records in the logical relationship description text that use the unique identifier of the current process as the subsequent process. The format of each record is "previous process identifier points to subsequent process identifier, constraint type = completion-start". Iterate through all records, filter out the records where the subsequent process identifier matches the unique identifier of the current process, extract the preceding process identifier from each matching record, and add all preceding process identifiers to the set of immediate preceding process identifiers of the current process.

[0028] If the preceding process information is not explicitly given in the logical relationship description text, it is supplemented according to the default rules of construction process: for the starting process, its preceding process identifier set is empty; for non-starting processes, if there is no logical relationship record pointing to the process, it is determined that the process has a default "complete-start" constraint with the previous process, and the identifier of the previous process is added to its preceding process identifier set.

[0029] For each process node obtained from parsing, store its unique identifier, estimated duration, and set of identifiers of preceding processes, and initialize its progress completion ratio to zero; if there is an identifier in the set of identifiers of preceding processes that is not defined in the process list, output an error message and stop the parsing process.

[0030] Based on the set of predecessor process identifiers for each process, directed edges are created from each predecessor process node to the current process node, and directed edges are created from the current process node to each of its successors process nodes, forming an initial directed graph containing all process nodes and directed edges.

[0031] Perform topological sorting on the initial directed graph: repeatedly remove process nodes that do not currently have any preceding process pointing to them, and record the removal order; if the removal operation continues until it can no longer remove nodes and there are still nodes left to be removed, it is determined that there is a circular dependency in the initial directed graph, the establishment process is terminated and an error message is output; if all nodes are successfully removed, a progress network structure without circular dependencies is generated according to the removal order.

[0032] For each process node, store the estimated duration, progress completion percentage, initial total float, and initial free float in the corresponding node data structure to complete the establishment of the schedule network plan.

[0033] It should be noted that the initial values ​​of total float and free float are obtained as follows: starting from the starting node of the topological sort, the earliest start time and earliest end time of each process node are calculated in the forward propagation order, where the earliest start time of the starting node is set to zero, and the earliest start time of subsequent nodes is the maximum value of the earliest end times of all its immediate predecessor processes; after completing the forward propagation, the latest start time and latest end time of each process node are calculated in the reverse propagation order, where the latest end time of the terminating node is equal to its earliest end time, and the latest end time of subsequent nodes is the minimum value of the latest start times of all its immediate successor processes.

[0034] The initial value of the total float is obtained by calculating the difference between the latest start time and the earliest start time of each process node. The initial value of the free float of each process node is obtained by subtracting the earliest end time of the process node from the minimum of the earliest start times of all subsequent processes of each process node.

[0035] When the progress completion ratio of any process node changes, the schedule deviation propagation module propagates the deviation downstream along the direction of the next process based on the schedule deviation of the process, recalculates the total float and free float of each affected next process node, and updates the float parameters in the schedule network plan.

[0036] In this embodiment, the schedule deviation propagation module is implemented as follows: the difference between the actual duration and the expected duration of the current process node is taken as the schedule deviation. A positive schedule deviation indicates that the actual schedule is extended, and a negative schedule deviation indicates that the actual schedule is shortened.

[0037] Based on the successor constraints in the schedule network plan, identify all successor process nodes starting from the current process node to form the downstream propagation range.

[0038] The schedule deviation is added to the cumulative path of the remaining duration of each subsequent process node. The following judgments are performed according to the positive and negative directions of the schedule deviation: when the schedule deviation is positive, the free float of each affected subsequent process node is deducted level by level along each downstream path. When there is still an uneliminated deviation value after the free float is reduced to zero, the total float of the corresponding process node is deducted.

[0039] When the schedule deviation is negative, the free float and total float of each subsequent process node are increased step by step along each downstream path. The increased float value shall not exceed the upper limit of the difference between the initial float and the consumed float. The specific implementation method is as follows: each process node maintains a counter for consumed free float and consumed total float, which is initially zero and is accumulated each time a positive deviation is deducted during propagation. When a negative deviation is recovered, the increase shall not exceed the current value of the counter.

[0040] Write the recalculated time difference parameters back to the corresponding process node storage location in the schedule network plan.

[0041] It should be noted that the aforementioned downstream propagation range is specifically formed by taking all the identified successor process nodes starting from the current process node as the first-level downstream nodes, and expanding the downstream path layer by layer using a breadth-first traversal order. Specifically, an empty queue and an empty set are initialized, the current process node is added to the queue as the starting node and marked as visited, and the current node is recorded as level 0; the following operations are repeated until the queue is empty: a node is taken from the head of the queue, and all successor process nodes of the taken node are judged one by one. If a successor process node has not yet been marked as visited, the successor process node is added to the tail of the queue, and its level is recorded as the current node's level plus one, and the successor process node is marked as visited.

[0042] When nodes in the queue are taken out one by one, the order in which they are taken out forms the access sequence of downstream nodes. During the traversal, all nodes marked as visited except the starting node constitute the downstream propagation range. If a node has multiple successor paths, the node is added to the queue only once to ensure that each node is processed only once.

[0043] The adaptive threshold generation module generates the criticality weight value of the corresponding process based on the total float after each process node is updated, and generates the progress risk alarm threshold for the current process by combining the risk lag rate data and the remaining free float.

[0044] In this embodiment, generating the progress risk alarm threshold for the current process includes: obtaining the total float after each process node is updated, taking the reciprocal of the total float after each process node is updated, dividing the reciprocal by the sum of the reciprocals of all process nodes, and generating the criticality weight value of each process node.

[0045] Identify the key construction parameters that affect the progress of the current process node, and calculate the risk lag rate of the current process node based on the deviation of the actual performance of the key construction parameters in the completed construction stage from the standard value range.

[0046] Among them, reference Figure 2 As shown, the identification of key construction parameters affecting the progress of the current process node includes: dividing the construction parameters affecting the process progress into multiple preset construction categories, including four categories: resource allocation, technical execution, environmental matching, and collaborative management. The specific method for classifying construction parameters into corresponding construction categories is as follows: S1. Define a multi-dimensional feature template vector for each preset construction category, where each dimension represents an influencing factor. The influencing factors include, but are not limited to: production factor dimensions (people, machines, materials), quality accuracy dimensions (tolerance, pass rate), external environment dimensions (weather, temperature, site), and time-series collaboration dimensions (handover, waiting, parallel processing).

[0047] The values ​​for each dimension were pre-calibrated using principal component analysis or the expert Delphi method on a large number of existing construction parameters, and the values ​​were normalized to... Interval. For example, the feature template for the resource allocation category takes 0.95 in the production factor dimension and 0.10 in the quality and accuracy dimension.

[0048] S2. For any construction parameter to be classified, parse its definition text, unit, typical measurement method and the type of process to which it belongs, and generate its feature vector of the same dimension in the following way: S21. Text semantic mapping: Input the parameter name and description text into the preset industry word vector model, such as the Word2Vec or BERT model trained based on bridge construction corpus, calculate the semantic similarity between the text and the keyword set of each influencing factor, and take the maximum value as the initial value of the current dimension.

[0049] S22. Unit and Dimension Rule Correction: If the unit of the parameter is "person", "unit", "ton", "%", etc., the value of the production factor dimension will be automatically increased; if the unit is "mm", "second", "MPa", etc., the value of the quality accuracy dimension will be increased; if the unit is "℃", "m / s", "mm / h", etc., the value of the external environment dimension will be increased; if the parameter description contains keywords such as "handover", "delay", "waiting", "coordination", etc., the value of the time sequence coordination dimension will be increased, specifically according to a preset step size.

[0050] S23. Comprehensive Normalization: Add the semantic similarity and the values ​​of each dimension after unit rule correction to obtain the final feature vector.

[0051] S3. Use cosine similarity to calculate the correlation strength between the construction parameter feature vector and the feature template of each category.

[0052] S4. The category with the highest correlation strength is used as the category to which the construction parameter is assigned. If the highest correlation strength is lower than the preset assignment threshold of 0.35, the construction parameter is marked as a general monitoring parameter and is not included in the subsequent priority sorting and screening process. If the correlation strength difference between two or more categories is less than 0.05, the construction parameter is allowed to be assigned to multiple categories at the same time, and duplicates will be removed in the subsequent statistics.

[0053] As an example, construction parameters belonging to the resource allocation category include, but are not limited to, labor availability rate, key equipment utilization rate, and material arrival timeliness rate; construction parameters belonging to the technical execution category include, but are not limited to, first-pass yield rate of procedures and component positioning deviation; construction parameters belonging to the environmental matching category include, but are not limited to, duration of adverse weather impact; and construction parameters belonging to the collaborative management category include, at least, the frequency of work process handover delays.

[0054] Based on the current bridge construction environment, priority is assigned to each construction category. Specifically, the operation environment identifier, technical difficulty identifier, and resource allocation requirement identifier of the current process node are read. If the operation environment identifier is one of high-altitude operation, deep-water operation, or night operation, the priority order is: environment matching, technical execution, resource allocation, and collaborative management.

[0055] Otherwise, if the technical difficulty is marked as high difficulty or new process, the priority order is: technical execution, environmental matching, resource allocation, and collaborative management.

[0056] Otherwise, if the resource allocation requirement is identified as a scarce resource or a single-source device, the priority order is: resource allocation, technology execution, environment matching, and collaborative management.

[0057] Otherwise, the priority order is: collaborative management, technical execution, resource allocation, and environment matching.

[0058] The number of parameters to be filtered is assigned to each construction category according to priority, from highest to lowest: the highest priority category filters three parameters, the next highest priority category filters two parameters, and the remaining two categories each filter one parameter. If the actual number of construction parameters in a construction category is less than the number that should be allocated, the actual number shall prevail.

[0059] Based on the allocated quantity, select a corresponding number of parameters from each construction category according to the historical correlation between construction parameters and process progress, from high to low, to form a preliminary screening parameter set.

[0060] Read the actual sampling sequence and corresponding standard value range of each construction parameter in the completed construction stage from the initial screening parameter set, calculate the deviation amplitude and deviation duration for each sampling point, and mark the sampling points with positive deviation amplitude as parameter trigger events.

[0061] The method for calculating the deviation amplitude is as follows: For numerical parameters, the standard value range of the parameter is obtained according to the control target specified in the design documents or construction plan. If the current value of the construction parameter is within the standard value range, the deviation amplitude is set to 0. Otherwise, the current value is subtracted by the median value of the standard value range and half of the span of the standard value range to obtain the deviation amplitude.

[0062] For event-type parameters, such as the duration of severe weather impact or the frequency of process handover delays, the deviation amplitude is taken as the actual occurrence value when the actual occurrence value is greater than 0, otherwise it is taken as 0.

[0063] Deviation duration refers to the time elapsed from the first sampling point where the deviation amplitude is greater than 0 until the last sampling point where the deviation amplitude is equal to 0. If there is an interruption during the deviation period, the duration before the interruption is counted separately as the duration of a single trigger event.

[0064] The occurrence time series of all triggering events for each construction parameter are aligned with the actual lag time series of the current process node on the time axis, and the correlation coefficient is calculated. The correlation coefficient is specifically the Pearson correlation coefficient, which is calculated by using the covariance of the actual lag time series and the occurrence time series of the triggering events as the numerator and the product of the standard deviations of the two series as the denominator. Construction parameters with a correlation coefficient greater than a preset correlation threshold (default value is 0.6) are retained as key construction parameters.

[0065] The calculation of the risk lag rate of the current process node includes: for each construction parameter in the set of key construction parameters, extracting the deviation amplitude of each triggering event, and using the maximum and minimum value normalization method to map the deviation amplitude to the interval of 0 to 1.

[0066] Deviation levels are determined by the normalized amplitude: amplitude less than 0.3 is classified as Level 1 deviation with a weight of 0.3; amplitude between 0.3 and 0.7 is classified as Level 2 deviation with a weight of 0.6; and amplitude greater than 0.7 is classified as Level 3 deviation with a weight of 1.0.

[0067] Obtain the ratio of the duration of deviation for each triggered event to the total number of sampling periods for the process to which the event belongs. Multiply this ratio by the level weight assigned to the event to obtain the weighted impact value of the triggered event. Sum the weighted impact values ​​of all triggered events for the same key construction parameter to obtain the cumulative deviation index of this key construction parameter.

[0068] The cumulative deviation indices of each key construction parameter are accumulated or the maximum value is taken according to the preset fusion rules to obtain the delayed trigger cumulative index of the current construction process that has been completed. The fusion rules are as follows: when the triggering events of multiple parameters overlap in the same time window, the maximum value of the weighted influence value of each parameter is taken into the cumulative index and not accumulated repeatedly; when the triggering events do not overlap in time, the weighted influence values ​​of each event are accumulated.

[0069] Read the cumulative lag time of the current process node in the completed construction phase from the schedule network plan. The cumulative lag time is equal to the actual construction duration minus the expected duration, and is taken as a positive value.

[0070] The cumulative index of lag triggering is fitted with a linear regression and the cumulative lag duration. A univariate linear regression equation passing through the origin is used to calculate the slope parameter as the ratio of the influence of parameter deviation on lag. When the cumulative lag duration is zero, the influence ratio is set to zero. When the calculated influence ratio is greater than 1, the value is set to 1. The influence ratio is used as the risk lag rate of the current process.

[0071] The product of the criticality weight value and the risk lag rate is used as the comprehensive risk coefficient. The complementary value of the comprehensive risk coefficient is calculated by subtracting the comprehensive risk coefficient from 1, and then multiplied by the remaining free float of the current process node to obtain the progress risk alarm threshold of the current process.

[0072] The progress delay alarm module collects the process connection delay time of the current process in real time. When the process connection delay time is greater than or equal to the progress risk alarm threshold, a risk alarm signal is triggered.

[0073] In this embodiment, the real-time acquisition of the process connection lag time of the current process includes: reading the actual start time of the current process node and the planned completion time of the preceding process node corresponding to the current process node from the construction log.

[0074] Subtract the planned completion time from the actual start time. If the difference is positive, output the difference as the process connection delay time.

[0075] If the difference is zero or negative, the output process connection delay time is zero.

[0076] The present invention also includes a backtracking analysis module, which, in response to a risk alarm signal, traces backward along the preceding direction, sequentially checks the progress of each preceding process, calculates the contribution of each preceding process to the delay of the current process, and adjusts the risk alarm threshold of each subsequent process.

[0077] Reference Figure 3As shown, in this embodiment, the calculation of the contribution delay amount of each preceding process to the delay of the current process includes: based on the preceding constraints in the schedule network plan, backtracking to identify all preceding process nodes associated with the current process node, forming a reverse tracing path.

[0078] A separate contribution account is established for each process node. The contribution account is indexed by the unique identifier of the preceding process node and has an initial value of zero.

[0079] Read the schedule deviation records of each preceding process node, and perform the following calculations for each preceding process node: allocate the schedule deviation value level by level according to the consumption order of free float and total float of each process node in the propagation path from the preceding process node to the current process node, until all deviation values ​​are attributed to the contribution account of each process node.

[0080] Specifically: the schedule deviation value is used as the initial residual deviation value. Along the propagation path from the preceding process node to the current process node, each process node on the path is visited level by level in order from the preceding to the following, including intermediate nodes and the current process node itself.

[0081] At each node, the free float remaining amount is read first. If the remaining deviation is less than or equal to the free float remaining amount, the remaining deviation is deducted from the free float remaining amount. The deducted deviation share is fully credited to the contribution account of the immediate preceding process node corresponding to the node, and the path allocation ends.

[0082] If the remaining deviation is greater than the free float remaining amount, then all the free float remaining amount is deducted, and the deducted value is added to the contribution account. The remaining deviation is updated to the original deviation amount minus the deducted value. Then, the total float remaining amount of this node continues to be consumed, deducted and updated according to the same rules, until the remaining deviation amount is zero or the current process node is reached.

[0083] If the remaining deviation is not zero when the current process node is reached, the remaining deviation will be added to the index corresponding to the preceding process node in the contribution account of the current process node.

[0084] The contribution shares received by the current process node that belong to the same preceding process node are summed up, and the contribution delay amount of the preceding process node to the delay of the current process node is output.

[0085] In this embodiment, adjusting the progress risk alarm threshold of each successive process includes: calculating the sum of the contribution delays of each preceding process node as the total contribution delay.

[0086] Read the remaining free float of the current process node. If the remaining free float is greater than or equal to the total contribution delay, keep the existing schedule risk alarm threshold of all subsequent process nodes unchanged; otherwise, mark the current process node as being in a capacity over-limit state, and take the difference between the contribution delay of the current process node and the remaining free float as the deviation to be transmitted.

[0087] For each successor process node in the current process node that is in a capacity over-limit state, process them sequentially, specifically in ascending order of earliest start time. If the earliest start events are the same, process them in lexicographical order of process identifiers. Obtain the remaining free float of the successor process node being processed. If the remaining free float is greater than zero, deduct the share corresponding to the deviation to be transferred. The share allocation method is as follows: if the deviation to be transferred is less than or equal to the remaining free float of the successor process node being processed, deduct the full amount and clear the deviation to be transferred to zero; otherwise, deduct all remaining free float and update the deviation to be transferred to the original deviation to be transferred minus the deducted value.

[0088] At the same time, based on the proportion of the actual deduction share to the original remaining free float, the schedule risk alarm threshold of the current processing successor node is tightened downward. Specifically, 1 is subtracted from the proportion of the actual deduction share to the original remaining free float, and the calculated difference is multiplied by the schedule risk alarm threshold of the current processing successor node to obtain the tightened threshold result.

[0089] If the remaining free float is zero, the schedule risk alarm threshold of the current successor process node will not be adjusted, and the deviation to be transferred will continue to be transferred to the next successor process node until the deviation to be transferred is completely absorbed by the remaining free float of a certain process node or transferred to the end of the schedule network.

[0090] The above content is merely an example and illustration of the concept of the present invention. Those skilled in the art can make various modifications or additions to the specific embodiments described, or use similar methods to replace them, as long as they do not deviate from the concept of the invention or exceed the scope defined by the present invention, and all such modifications and additions should fall within the protection scope of the present invention.

Claims

1. A bridge construction risk assessment and data analysis system, characterized in that, include: The schedule network construction module establishes a schedule network plan based on the bridge construction design documents, which includes multiple process nodes and the preceding and succeeding constraints between nodes. Each process node stores the expected duration, progress completion percentage, initial value of total float, and initial value of free float. The schedule deviation propagation module, when the progress completion ratio of any process node changes, propagates the deviation downstream along the direction of the next process based on the schedule deviation of the process, recalculates the total float and free float of each affected next process node, and updates the float parameters in the schedule network plan. The schedule deviation propagation module includes: The difference between the actual duration and the expected duration of the current process node is taken as the schedule deviation. Based on the successor constraints in the schedule network plan, identify all successor process nodes starting from the current process node to form the downstream propagation range; The schedule deviation is added to the cumulative path of the remaining duration of each subsequent process node, and the following judgments are performed according to the positive and negative directions of the schedule deviation: When the schedule deviation is positive, the free float of each affected subsequent process node is deducted level by level along each downstream path. If there is still an uneliminated deviation value after the free float is reduced to zero, the total float of the corresponding process node is deducted. When the schedule deviation is negative, the free float and total float of each subsequent process node are increased step by step along each downstream path, and the increased float value does not exceed the upper limit of the difference between the initial float and the consumed float. Write the recalculated time difference parameters back to the corresponding process node storage location in the schedule network plan; The adaptive threshold generation module generates the criticality weight value of the corresponding process based on the total float after each process node is updated, and generates the progress risk alarm threshold for the current process by combining the risk lag rate data and the remaining free float. The progress delay alarm module collects the process connection delay time of the current process in real time. When the process connection delay time is greater than or equal to the progress risk alarm threshold, a risk alarm signal is triggered. It also includes a backtracking analysis module, which, in response to risk alarm signals, traces backward along the preceding direction, checks the progress of each preceding process in sequence, calculates the contribution of each preceding process to the delay of the current process, and adjusts the progress risk alarm threshold of each subsequent process.

2. The bridge construction risk assessment and data analysis system according to claim 1, characterized in that, The schedule network plan is established in the following way: Read the construction procedure list and the text describing the logical relationship between each procedure from the bridge construction design documents, and parse the unique identifier, expected duration and the set of identifiers of the immediate preceding procedures for each procedure. Based on the set of predecessor process identifiers for each process, directed edges are established from each predecessor process node to the current process node, and directed edges are established from the current process node to each of its successors process nodes, forming an initial directed graph containing all process nodes and directed edges. Perform topological sorting on the initial directed graph: repeatedly perform the operation of removing process nodes that do not currently have any preceding process pointing to them, and record the removal order; if the removal operation continues until it can no longer remove nodes and there are still nodes that have not been removed, it is determined that there is a circular dependency in the initial directed graph, the establishment process is terminated and an error message is output; if all nodes are successfully removed, a progress network structure without circular dependencies is generated according to the removal order. For each process node, store the estimated duration, progress completion percentage, initial total float, and initial free float in the corresponding node data structure to complete the establishment of the schedule network plan.

3. The bridge construction risk assessment and data analysis system according to claim 1, characterized in that, The generation of the progress risk alarm threshold for the current process includes: Obtain the total float after each process node is updated, take the reciprocal of the total float and perform normalization calculation to generate the criticality weight value of each process node; Identify the key construction parameters that affect the progress of the current process node, and calculate the risk lag rate of the current process node based on the deviation of the actual performance of the key construction parameters in the completed construction stage from the standard value range. The product of the criticality weight value and the risk lag rate is used as the comprehensive risk coefficient. The complementary value of the comprehensive risk coefficient is calculated and multiplied by the remaining free float of the current process node to obtain the progress risk alarm threshold of the current process.

4. The bridge construction risk assessment and data analysis system according to claim 3, characterized in that, The identification of key construction parameters affecting the progress of the current process node includes: The construction parameters that affect the progress of the process are divided into multiple preset construction categories. Based on the current bridge construction environment, each construction category is prioritized and sorted. Different number of parameters are assigned to each construction category from high to low priority to obtain the initial set of parameters. Read the actual sampling sequence and corresponding standard value range of each construction parameter in the completed construction stage from the initial screening parameter set, calculate the deviation amplitude and deviation duration for each sampling point, and mark the sampling points with positive deviation amplitude as parameter trigger events; Align the occurrence time series of all triggering events for each construction parameter with the actual lag time series of the current process node on the time axis, calculate the correlation coefficient, and retain the construction parameters with a correlation coefficient greater than the preset correlation threshold as key construction parameters.

5. The bridge construction risk assessment and data analysis system according to claim 1, characterized in that, The real-time acquisition of the process connection lag time of the current process includes: Read the actual start time of the current process node and the planned completion time of the preceding process node corresponding to the current process node from the construction log; Subtract the planned completion time from the actual start time. If the difference is positive, output the difference as the process connection delay time. If the difference is zero or negative, the output process connection delay time is zero.

6. The bridge construction risk assessment and data analysis system according to claim 1, characterized in that, The calculation of the contribution delay amount of each preceding process to the delay of the current process includes: Based on the preceding constraints in the schedule network plan, backtrack to identify all preceding process nodes associated with the current process node, forming a reverse tracing path; Read the schedule deviation records of each preceding process node, and perform the following calculations for each preceding process node: allocate the schedule deviation value according to the consumption order of free float and total float of each process node in the propagation path from the preceding process node to the current process node, and allocate the deviation contribution share level by level until all deviation values ​​are attributed to the contribution account of each process node. Add up the shares of the same preceding process node among all the contribution shares received by the current process node, and output the delay contribution of the preceding process node to the delay of the current process node.

7. The bridge construction risk assessment and data analysis system according to claim 6, characterized in that, The adjustment of the progress risk alarm thresholds for each subsequent process includes: Calculate the sum of the contribution delays of each preceding process node as the total contribution delay; Read the remaining free float of the current process node. If the remaining free float is greater than or equal to the total contribution delay, keep the existing schedule risk alarm threshold of all subsequent process nodes unchanged; otherwise, mark the current process node as being in a capacity overrun state and take the difference between the contribution delay of the current process node and the remaining free float as the deviation to be transferred. For each successor process node that is in a capacity over-limit state, process them sequentially: obtain the remaining free float of the successor process node being processed; if the remaining free float is greater than zero, deduct the corresponding share of the deviation to be transferred from it; at the same time, tighten the progress risk alarm threshold of the successor process node being processed downward according to the proportion of the actual deducted share to the original remaining free float. If the remaining free float is zero, the schedule risk alarm threshold of the current successor process node will not be adjusted, and the deviation to be transferred will continue to be transferred to the next successor process node until the deviation to be transferred is completely absorbed by the remaining free float of a certain process node or transferred to the end of the schedule network.

8. The bridge construction risk assessment and data analysis system according to claim 7, characterized in that, The deduction of the share corresponding to the deviation to be transmitted includes the following share allocation methods: If the deviation to be transferred is less than or equal to the remaining free float of the immediate successor node being processed, then the entire amount is deducted and the deviation to be transferred is cleared to zero; otherwise, all remaining free float is deducted and the deviation to be transferred is updated to the original deviation to be transferred minus the deducted value.

Images