Artificial intelligence algorithm-based capital construction site risk operation optimization management method

Abstract

The invention belongs to the technical field of optimization management, and particularly relates to an infrastructure site risk operation optimization management method based on an artificial intelligence algorithm, and the method comprises the steps: obtaining a historical infrastructure construction description set corresponding to each risk operation based on the construction monitoring record of each risk operation of a historical infrastructure site; performing data preprocessing to obtain a sample capital construction description set; performing classification based on the sample capital construction description set, obtaining a sample capital construction site feature set and a sample capital construction personnel feature set, and obtaining an intelligent decision tree model; and obtaining a real-time infrastructure construction description set corresponding to each risk operation, obtaining an intelligent risk decision value based on the real-time infrastructure construction description set and the intelligent decision tree model, and obtaining an infrastructure site risk operation scheme corresponding to each risk operation based on the intelligent risk decision value. According to the invention, accurate assessment and global coordinated optimization of capital construction site risk operation can be realized, and decision scientificity and management efficiency are improved.

Description

Technical Field

[0001] The present invention belongs to the technical field of optimization management, and specifically relates to an optimization management method for infrastructure site risk operations based on an artificial intelligence algorithm. Background Art

[0002] At present, there are many high-risk operations in infrastructure construction sites. Through real-time risk analysis, unsafe behaviors, illegal operations, and unsafe conditions of equipment and facilities, such as equipment loss and wire aging, can be discovered in a timely manner, so that measures can be taken to correct them, avoid accidents, and protect the life safety and health of construction workers. In addition, by obtaining the situation on site in real time and formulating management optimization strategies in a timely manner based on actual operation risks, managers can understand the use of resources such as personnel, equipment, and materials in real time, and make reasonable arrangements based on actual progress to avoid idleness and waste of resources, improve resource utilization efficiency, and ensure that the project proceeds as planned.

[0003] However, the existing methods for optimizing the management of infrastructure risk operations still have some shortcomings, which are specifically reflected in the following aspects: the application of traditional methods for optimizing the management of infrastructure site risk operations is relatively limited in terms of intelligent decision-making, and most of them are in the collection of information data and the single-line processing of data. On the one hand, the singleness of information data collection cannot reflect the interaction between data, and thus cannot reflect the comprehensive impact of data on the final operation risk assessment in decision-making; on the other hand, the single-line processing of data leads to slow data processing speed, which affects the timeliness of risk decisions and optimization management decisions for infrastructure site operations. Summary of the invention

[0004] In view of the shortcomings of the existing technology, the present invention provides an optimization management method for infrastructure site risk operations based on artificial intelligence algorithms, which can achieve accurate assessment and global coordination optimization of infrastructure site risk operations, and improve scientific decision-making and management efficiency.

[0005] To achieve the above objectives, the technical solution adopted by the present invention is a method for optimizing the management of infrastructure site risk operations based on artificial intelligence algorithms, comprising the following steps: Step 1: Based on the construction monitoring records of each risk operation at the historical infrastructure site, obtain the historical infrastructure construction description set corresponding to each risk operation. The process is as follows: Obtain historical infrastructure construction monitoring records, including certain historical infrastructure construction conditions; Based on any historical infrastructure construction situation, obtain a historical infrastructure construction information set, including equipment failure rate, emergency brake response time, equipment maintenance compliance rate, personnel safety equipment coverage rate, and personnel fatigue index of the historical infrastructure construction situation; Mark the historical infrastructure construction information set as a historical infrastructure construction description point corresponding to the historical infrastructure construction situation; A historical infrastructure construction description set is obtained by integrating the historical infrastructure construction description points corresponding to all historical infrastructure construction conditions; Step 2: Based on the historical infrastructure construction description set, data preprocessing is performed to obtain a sample infrastructure construction description set; Step 3: Classify based on the sample infrastructure construction description set, obtain the sample infrastructure site feature set and the sample infrastructure personnel feature set, and obtain an intelligent decision tree model based on the sample infrastructure site feature set and the sample infrastructure personnel feature set; Step 4: Obtain the real-time infrastructure construction description set corresponding to each risk operation, obtain the intelligent risk decision value based on the real-time infrastructure construction description set and the intelligent decision tree model, and obtain the infrastructure site risk operation plan corresponding to each risk operation based on the intelligent risk decision value.

[0006] Preferably, the method for obtaining the personnel fatigue index is: Obtaining the heart rate variability of construction workers based on wearable devices , construction site environmental parameters and mean square error of static load coefficient , construction site environmental parameters include ambient temperature mean square deviation , mean square error of ambient humidity and the mean square error of the construction location height ; Obtain the work data of construction personnel, including working hours , rest frequency and work intensity ; Based on heart rate variability , construction site environmental parameters, static load coefficient mean square error And the work data of construction personnel to determine the personnel fatigue index , the formula is: ; ; In the formula, , , , , , are weight factors, is the rest frequency parameter value, is the transfer function, is the heart rate variability parameter value, is the ambient temperature mean square error parameter value, is the parameter value of the mean square error of the ambient humidity, It is the parameter value of mean square deviation of construction location height.

[0007] Preferably, in step 2, the process of performing data preprocessing based on the historical infrastructure construction description set to obtain the sample infrastructure construction description set is: Obtain a collection of historical infrastructure construction descriptions; Obtaining the historical infrastructure construction information interval stored in the database, the historical infrastructure construction information interval describing the reasonable range of parameter values ​​in the historical infrastructure construction information set; Based on historical infrastructure construction information interval and 3 The rule is used to perform double cleaning of abnormal data and obtain the cleaned infrastructure construction description set: Based on the historical infrastructure construction information interval, the historical infrastructure construction description set is preliminarily cleaned of abnormal data. The preliminarily cleaning of abnormal data is to process the missing parameters of the historical infrastructure construction description set that exceed the historical infrastructure construction information interval and mark the missing parameter positions; Get all the historical infrastructure construction information sets in the historical infrastructure construction description set, based on any parameter in the historical infrastructure construction information set , i represents the parameter number, get The parameter mean And the parameter standard deviation ; Among them, if , then Mark as outliers and perform secondary cleaning of abnormal data. Perform missing processing and mark the missing position of parameters, Indicates the jth value of the i-th parameter; Get all the missing parameter positions and use the mean filling method to fill in the parameters; Get the cleaned infrastructure construction description set; Based on the cleaned infrastructure construction description set, data normalization is performed to obtain a sample infrastructure construction description set.

[0008] Preferably, the process of obtaining all missing parameter positions and filling the parameters using the mean filling method is: Get the position where any parameter is missing ; Get parameter type based on the missing position of the parameter; Based on the parameter type, obtain a collection of infrastructure construction conditions in the same industry with similar parameter types and complete data stored in the database; Based on the same industry and parameter type with similar infrastructure construction conditions and complete data, the missing position of the parameter Use mean filling method to fill.

[0009] Preferably, in step 3, the process of obtaining the intelligent decision tree model is: Classify based on the sample infrastructure construction description set to obtain the sample infrastructure site feature set and the sample infrastructure personnel feature set; The sample infrastructure site feature set includes equipment failure rate, emergency brake response time, equipment maintenance compliance rate, and equipment maintenance compliance rate; The characteristic set of sample infrastructure personnel includes personnel safety equipment coverage and personnel fatigue index; Performing data normalization processing on the sample infrastructure site feature set and the sample infrastructure personnel feature set respectively, to obtain the sample infrastructure site normalized feature set and the sample infrastructure personnel normalized feature set; Based on the normalized feature set of the sample infrastructure site, data level marking is performed to obtain the level feature set of the sample infrastructure site; Based on the normalized feature set of sample infrastructure construction personnel, data level marking is performed to obtain the level feature set of sample infrastructure construction personnel; Based on the sample infrastructure site level feature set, obtain the parameter information entropy corresponding to each parameter in the sample infrastructure site level feature set; Based on the sample infrastructure personnel level feature set, obtain the parameter information entropy corresponding to each parameter in the sample infrastructure personnel level feature set; An intelligent decision tree model is obtained based on the information entropy of each parameter in the sample infrastructure personnel grade feature set and the sample infrastructure site grade feature set.

[0010] Preferably, the process of obtaining the normalized feature set of the sample infrastructure site and the normalized feature set of the sample infrastructure personnel is: Get any parameter in the sample infrastructure site feature set ; Get Parameters The corresponding maximum value and the minimum value ; Based on the maximum value and the minimum value And parameters Perform normalization operation on any value of to obtain normalized parameters; After integration, a normalized feature set of the sample infrastructure site is obtained; Get any parameter in the sample infrastructure personnel feature set ; Get Parameters The corresponding maximum value and the minimum value ; Based on the maximum value and the minimum value And parameters Perform normalization operation on any value of to obtain normalized parameters; After integration, the normalized feature set of sample infrastructure personnel is obtained.

[0011] Preferably, the process of obtaining an intelligent decision tree model based on the information entropy of each parameter is: Obtain parameter information entropy corresponding to each level parameter in the sample infrastructure site level feature set, and mark it as the infrastructure site level parameter information entropy set; Obtain parameter information entropy corresponding to each level parameter in the sample infrastructure personnel level feature set, and mark it as infrastructure personnel level parameter information entropy set; Integrate the infrastructure site level parameter information entropy set and the infrastructure personnel level parameter information entropy set into a level parameter information entropy set; Matching decision tree node parameters stored in the database based on the level parameter information entropy set; Build an intelligent decision tree model based on decision tree node parameters.

[0012] Preferably, in step 4, the process of obtaining the infrastructure site risk operation plan corresponding to each risk operation based on the intelligent risk decision value is: Based on the real-time infrastructure site situation, a real-time infrastructure construction description set is obtained, where the real-time infrastructure construction description set includes a real-time infrastructure site feature set and a real-time infrastructure personnel feature set; Perform data normalization processing based on the real-time infrastructure site feature set to obtain a real-time infrastructure site normalized feature set; Perform data normalization processing based on the real-time infrastructure construction personnel feature set to obtain a real-time infrastructure construction personnel normalized feature set; Obtain intelligent risk decision values ​​based on the normalized feature set of real-time infrastructure site, the normalized feature set of real-time infrastructure personnel, and the intelligent decision tree model; The intelligent risk decision value is stored as a risk decision designated label, the risk decision designated label is compared one by one with each risk decision setting label stored in the database, the risk decision setting label that is the same as the risk decision designated label is found, and the infrastructure site risk operation plan corresponding to the risk decision setting label is obtained from the database.

[0013] Preferably, the method further comprises: step 5, summarizing the optimization management schemes corresponding to the risks of each operation, obtaining a set of candidate schemes for each risk category, obtaining basic parameters of each optimization management scheme, and obtaining a basic parameter table of the schemes, wherein the basic parameters of the schemes include the risk value, resource requirements and execution time of each optimization management scheme; Based on the basic parameters of each optimization management scheme, the comprehensive score of each optimization management scheme corresponding to each optimization management scheme is obtained; The specific method for obtaining the comprehensive score of a single solution corresponding to one of the optimization management solutions for a certain type of risk operation is as follows: Determine the risk operation corresponding to the optimization management plan, obtain the infrastructure construction information set corresponding to the risk operation, and determine the risk value of the optimization management plan based on the infrastructure construction information set corresponding to the risk operation, specifically: ; ; ; In the formula, The risk factor of the infrastructure site for risky operations, is the equipment failure rate, is the emergency brake response time, Maintain compliance rates for equipment, , , , , is the weight factor, is the personnel risk factor of risky operations, For personnel safety equipment coverage, is the personnel fatigue index, To optimize the risk value of the management plan; Normalize the risk value of the optimized management plan; Get comprehensive rating of a single solution , specifically: ; is the risk value of the optimized management plan after normalization, To optimize the actual demand of the management scheme for the b-type resource, is the maximum demand for the bth resource among all the optimization management solutions corresponding to the risk operation, To optimize the execution time of management solutions, is the maximum execution time of all optimization management solutions corresponding to the risk operation. , , is the weight factor; Determine the risk operation types at the infrastructure construction site, and construct a global coordination management solution combination set based on the candidate solution sets of each risk category. Each global coordination management solution combination in the global coordination management solution combination set includes an optimized management solution corresponding to each risk operation type. Determine the scheme combination conflict intensity of each global coordination management scheme combination and the coordination risk index of the global coordination management scheme combination; Conflict intensity of solution combinations The method to obtain is: ; ; ; ; in, is the resource conflict coefficient, is the time conflict coefficient, is the spatial conflict coefficient, for The weight factor of for The weight factor of for The weight factor of is the total demand for the bth resource by all the optimal management schemes in the global coordinated management scheme combination, k is an index variable used to traverse the optimal management schemes in the global coordinated management scheme combination, is the overlapping duration of the g-th job, for The associated weight of is the total operation duration of the fth optimal management solution in the global coordination management solution combination, The total area of ​​the operating area overlapped by all the optimized management schemes in the global coordinated management scheme combination. is the operating area of ​​the fth optimal management scheme in the global coordinated management scheme combination, is the actual distance from the edge of the working area, The distance required by the safety distance specification; Synergistic risk indicators for the global coordinated management solution portfolio The method to obtain is: ; in, for The corresponding optimization management plan and The corresponding risk coupling coefficient between the optimization management schemes, For The risk value of an optimal management solution is associated with For The risk value of another associated optimal management solution; The optimal global coordination management solution combination is selected based on the minimization of the objective function and the constraints, where: The objective function is: ; In the formula, is the comprehensive score of the ath optimal management solution in the global coordinated management solution combination. To globally coordinate and manage the scheme combination conflict intensity, To provide a synergistic risk indicator for the overall coordinated management solution portfolio, for The weight factor of for The weight factor of The constraints are: ; ; In the formula, is the demand for the bth resource of the ath optimal management solution, It is the upper limit of the collaborative risk index of the set global coordinated management solution combination.

[0014] The present invention has the following beneficial effects: The present invention uses an intelligent decision tree model to conduct multi-dimensional analysis of historical and real-time data, comprehensively considers multiple key factors such as equipment failure rate, emergency braking response time, and personnel fatigue index, and achieves accurate risk assessment and intelligent optimization of operation plans. At the same time, the global coordination management mechanism is introduced to conduct conflict assessment and collaborative optimization of optimization management plans from the three dimensions of resources, time, and space, effectively reducing systemic risks and improving the scientific nature of decision-making and management efficiency.

[0015] The present invention can dynamically adjust risky operation plans according to real-time data to ensure a balance between construction safety and project progress. Through comprehensive scoring of basic parameters such as resource demand and execution time, as well as quantitative analysis of the conflict intensity and collaborative risk indicators of the global coordination management plan combination, it realizes the upgrade from single plan optimization to global combination optimization. This not only improves resource utilization efficiency and avoids idle and waste of resources, but also significantly reduces safety risks in the operation process, providing a strong guarantee for the smooth implementation of infrastructure projects. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 This is a schematic diagram of the method flow of Example 1 of the present invention; Figure 2 This is a schematic diagram of the method flow of Example 2 of the present invention. DETAILED DESCRIPTION

[0017] The technical solutions in the embodiments of the present invention will be described clearly and completely below in conjunction with the accompanying drawings in the embodiments of the present invention.

[0018] Example 1: Figure 1 As shown, the present invention provides an optimization management method for infrastructure site risk operations based on artificial intelligence algorithms, comprising the following steps: Step 1: Based on the historical infrastructure construction monitoring records of each risk operation on site, obtain the historical infrastructure construction description set corresponding to each risk operation. The historical infrastructure construction description set includes several historical infrastructure construction description points. The specific process is as follows: Obtain historical infrastructure construction monitoring records. Historical infrastructure construction monitoring records include several historical infrastructure construction situations. Historical infrastructure construction monitoring records are a collection of data obtained by comprehensively monitoring and recording past infrastructure construction processes. Historical infrastructure construction situations are a basic unit in historical infrastructure construction monitoring records, representing a specific scene or stage in the past infrastructure construction process, and recording various details contained in a specific construction scene.

[0019] Based on any historical infrastructure construction situation, a historical infrastructure construction information set is obtained. The historical infrastructure construction information set includes the equipment failure rate, emergency braking response time, equipment maintenance compliance rate, personnel safety equipment coverage rate, and personnel fatigue index of the historical infrastructure construction situation. The personnel fatigue index is an indicator that comprehensively evaluates the fatigue level of construction personnel, taking into account working hours, rest frequency, and work intensity. The work intensity can be obtained from the construction project book and is divided into light, medium, and heavy, and the work intensity is marked as 1, 2, and 3 respectively.

[0020] The method for obtaining the personnel fatigue index is: obtaining the heart rate variability of construction workers based on wearable devices , construction site environmental parameters and mean square error of static load coefficient , construction site environmental parameters include ambient temperature mean square deviation , mean square error of ambient humidity and the mean square error of the construction location height ; Obtain construction personnel work data, including working hours , rest frequency and work intensity ; Based on heart rate variability , construction site environmental parameters, static load coefficient mean square error And the work data of construction personnel to determine the personnel fatigue index , the formula is: ; ; In the formula, for The weight factor of for The weight factor of for The weight factor of is the rest frequency parameter value, is the transfer function, is the heart rate variability parameter value, is the ambient temperature mean square error parameter value, is the parameter value of the mean square error of the ambient humidity, is the parameter value of the mean square error of the construction location height, for The weight factor of for The weight factor of is the weight factor of p.

[0021] Personnel fatigue index It integrates physiological indicators (heart rate variability), environmental parameters (temperature, humidity, altitude mean square deviation), static load coefficient mean square deviation and work data (duration, frequency, intensity), comprehensively covering the main factors affecting fatigue and avoiding the one-sidedness of a single indicator.

[0022] The method for obtaining the mean square error P of the static load coefficient is to count the action postures of the construction workers during risk operations (bending, carrying, etc.), compare the action postures with the posture fatigue score mapping table stored in the database, obtain the posture fatigue score corresponding to each action posture in the posture fatigue score mapping table, and then calculate the mean square error P of the static load coefficient; different actions have different pressures on the body, and the mean square error of the static load coefficient can reflect the cumulative impact of long-term or repetitive actions on muscles and bones, and improve the accuracy of fatigue assessment. By counting the action postures of the construction workers (such as bending, carrying, etc.) and comparing them with the posture fatigue score mapping table in the database, the degree of fatigue caused by the static load of the body can be objectively quantified.

[0023] Work intensity According to the energy consumption MET, muscle load MVC, action frequency Dp and static load coefficient mean square error P of the construction personnel, the specific acquisition method is: ; In the formula, is the weight factor of MET, is the weight factor of MVC, is the weight factor of Dp, is the weight factor of P. Work intensity combines energy expenditure (MET), muscle load (MVC), movement frequency (Dp) and mean square error of static load coefficient (P) to fully reflect the physical demand of the operation. Work intensity, as one of the core input parameters of fatigue index, together with heart rate variability and environmental parameters, constructs a risk assessment model of "man-machine-environment" collaboration.

[0024] The historical infrastructure construction information set is marked as a historical infrastructure construction description point corresponding to the historical infrastructure construction situation; and the historical infrastructure construction description set is obtained by integrating the historical infrastructure construction description points corresponding to all historical infrastructure construction situations.

[0025] Fully considering the construction conditions of historical infrastructure, collecting and integrating data based on historical infrastructure construction records, and comprehensively considering the operating conditions and quantitative representations of various historical infrastructure sites can provide scientific data reference and situation verification for the formulation of construction plans, make the optimization management plan more in line with actual construction conditions, and reduce changes and delays in the construction process.

[0026] Step 2: Based on the historical infrastructure construction description set, data preprocessing is performed to obtain a sample infrastructure construction description set. The specific acquisition process is as follows: Obtain a historical infrastructure construction description set; obtain the historical infrastructure construction information interval stored in the database, the historical infrastructure construction information interval describes the reasonable range of parameter values ​​in the historical infrastructure construction information set; based on the historical infrastructure construction information interval and 3 The rule is used to perform double cleaning of abnormal data to obtain the cleaned infrastructure construction description set; data normalization is performed based on the cleaned infrastructure construction description set to obtain the sample infrastructure construction description set.

[0027] Check each data record in the historical infrastructure construction description set one by one, and compare the value of each parameter in each data record with the corresponding historical infrastructure construction information interval. If a value of a parameter exceeds its corresponding historical infrastructure construction information interval, the parameter is marked as abnormal data. For the parameters marked as abnormal, missing processing is performed, that is, the value of the abnormal parameter is set to a null value or a specific missing mark, to ensure that the abnormal data is temporarily isolated without deleting the entire data record for further processing.

[0028] Based on historical infrastructure construction information interval and 3 The rule is used to perform double cleaning of abnormal data to obtain the cleaned infrastructure construction description set. The specific analysis process is: based on the historical infrastructure construction information interval, the historical infrastructure construction description set is preliminarily cleaned of abnormal data. The preliminarily cleaning of abnormal data is to process the missing of some parameters in the historical infrastructure construction description set that exceed the historical infrastructure construction information interval and mark the missing position of the parameters. Preliminary cleaning based on the historical infrastructure construction information interval can effectively remove those abnormal data that obviously deviate from the normal range. Due to the influence of data recording errors, special circumstances (such as improper use of equipment) or other interference factors on sample data, by marking the missing position of parameters, the integrity of the data can be more clearly understood in the subsequent data processing process. It helps to maintain data consistency, because the subsequent algorithm can accurately know which data has been processed according to the mark, and can decide whether to further investigate or supplement these missing data according to the specific situation. In this way, the data quality of the entire infrastructure construction description set can be ensured at a relatively reliable level.

[0029] Get all the historical infrastructure construction information sets in the historical infrastructure construction description set, based on any parameter in the historical infrastructure construction information set , i represents the parameter number, get The parameter mean And the parameter standard deviation ; Parameter mean The calculation formula is: ; In the formula, j represents the value sequence number, represents the jth value of the ith parameter, n is The number of all possible values ​​of ; Parameter standard deviation The calculation formula is: ; Among them, if , then Mark as outliers and perform secondary cleaning of abnormal data. Perform missing data processing and mark the missing parameter locations; obtain all the missing parameter locations and use the mean filling method to fill in the parameters; obtain the cleaned infrastructure construction description set, which is a historical infrastructure construction description set that has undergone double data cleaning and parameter missing location filling.

[0030] By calculating the mean and standard deviation of each parameter, we can scientifically and accurately identify abnormal data in the historical infrastructure construction description set, avoid the arbitrariness of subjective judgment, ensure the objectivity and accuracy of abnormal data identification, and make the subsequent various analyses and optimization management decision-making results based on the historical infrastructure construction description set more accurate and reliable. The abnormal data is missing and the missing position of the parameter is marked, and then the mean filling method is used to fill the parameters, making the cleaned infrastructure construction description set data more complete and reliable.

[0031] The process of obtaining all missing parameter positions and filling them with the mean value method is as follows: ; Get the parameter type based on the missing position of the parameter; Based on the parameter type, obtain the collection of infrastructure construction conditions in the same industry with similar parameter types and complete data stored in the database; Based on the collection of infrastructure construction conditions in the same industry with similar parameter types and complete data, find the missing position of the parameter Use mean filling method to fill.

[0032] For example, construction units that are engaged in the same sub-segment such as road and bridge construction, housing construction, and water conservancy project construction can be considered to be in the same industry. There are many parameters involved in the construction process of infrastructure, such as construction progress parameters, cost parameters, quality parameters, material usage parameters, etc. Similar parameter types mean that the selected other infrastructure construction situations are similar to the project with missing parameters in terms of the nature, meaning, and use of parameters.

[0033] Projects with similar parameter types in the same industry have certain commonalities and comparability. Using these complete data projects for mean filling can provide a relatively reasonable and reliable estimate for missing parameters based on the general situation in the industry and the actual data of similar projects, avoiding data irrationality caused by random filling. By filling missing parameters with the mean, the data can be made more complete, reducing the impact of missing data on subsequent data analysis results, improving the accuracy and credibility of data analysis, and helping to make more scientific decisions and judgments.

[0034] Step 3: Classify based on the sample infrastructure construction description set, obtain the sample infrastructure site feature set and the sample infrastructure personnel feature set, and obtain the intelligent decision tree model based on the sample infrastructure site feature set and the sample infrastructure personnel feature set. The specific process is as follows: Based on the sample infrastructure construction description set, classification is performed to obtain a sample infrastructure site feature set and a sample infrastructure personnel feature set; the sample infrastructure site feature set includes equipment failure rate, emergency brake response time, equipment maintenance compliance rate, and equipment maintenance compliance rate; The sample infrastructure personnel feature set includes personnel safety equipment coverage and personnel fatigue index. Personnel safety equipment coverage and personnel fatigue index are important indicators that directly affect the safety and working status of infrastructure personnel. Separately classifying these personnel-related features can help artificial intelligence algorithms more focused on analyzing the impact of personnel factors on infrastructure construction, provide a basis for formulating safety management measures (such as strengthening the deployment and inspection of safety equipment, and reasonably arranging working hours and rest, etc.), and improve the accuracy and reliability of the intelligent decision tree model.

[0035] Performing data normalization processing on the sample infrastructure site feature set and the sample infrastructure personnel feature set respectively, to obtain the sample infrastructure site normalized feature set and the sample infrastructure personnel normalized feature set; Based on the sample infrastructure site normalized feature set, data level marking is performed to obtain the sample infrastructure site level feature set. The specific process is to perform level marking on each normalized infrastructure site feature in each sample infrastructure site normalized feature set. For example, the normalized equipment failure rate is marked. If the level classification rules stored in the database are low level: normalized equipment failure rate belongs to [0-0.3], medium level: normalized equipment failure rate belongs to (0.3-0.7), high level: normalized equipment failure rate belongs to [0.7-1], then when the normalized equipment failure rate is 0.4, the level is medium. At the same time, the level distribution of the normalized equipment failure rate is statistically analyzed (how many low levels, how many medium levels, and how many high levels there are), and the level distribution of all normalized infrastructure site features is summarized to obtain the sample infrastructure site level feature set.

[0036] Based on the normalized feature set of sample infrastructure personnel, data level labeling is performed to obtain the sample infrastructure personnel level feature set. The method of obtaining the sample infrastructure personnel level feature set is the same as that of the sample infrastructure site level feature set. In the infrastructure site and infrastructure personnel feature sets, the dimensions and value ranges of different features may vary greatly. Through data normalization, these features are unified to the same scale to eliminate the impact of dimensions, so that the model can treat each feature equally and avoid the model being dominated by features with larger values, thereby improving the accuracy and stability of the model. In addition, the normalized data can make the gradient descent algorithm converge faster. When training the model, data normalization can make the step size of parameter updates more reasonable, avoid the problem of gradient vanishing or exploding due to large differences in feature value ranges, reduce the number of iterations of model training, and improve training efficiency.

[0037] Based on the sample infrastructure site grade feature set, the parameter information entropy corresponding to each parameter in the sample infrastructure site grade feature set is obtained; based on the sample infrastructure personnel grade feature set, the parameter information entropy corresponding to each parameter in the sample infrastructure personnel grade feature set is obtained; based on the sample infrastructure personnel grade feature set and the information entropy of each parameter in the sample infrastructure site grade feature set, an intelligent decision tree model is obtained.

[0038] The decision tree model itself has an intuitive tree structure. The decision tree built based on information entropy has clear basis and meaning for the splitting of each node, which corresponds to the characteristic parameters of specific infrastructure personnel or infrastructure sites. This makes the decision-making process and results of the model easy to understand. Professionals and non-professionals can clearly understand how the model makes operational risk assessment and optimization management decisions based on different characteristic parameters through the structure and node information of the decision tree, providing clear guidance for the analysis and decision-making of infrastructure projects.

[0039] The process of obtaining the normalized feature set of the sample infrastructure site by performing data normalization on the feature set of the sample infrastructure site is as follows: obtaining any parameter in the feature set of the sample infrastructure site ; Get parameters The corresponding maximum value and the minimum value ; Based on the maximum value and the minimum value And parameters Any value of is normalized to obtain normalized parameters, and then the normalized feature set of the sample infrastructure site is obtained after integration and arrangement.

[0040] Get any parameter in the sample infrastructure personnel feature set ; Get parameters The corresponding maximum value and the minimum value ; Based on the maximum value and the minimum value And parameters Any value of is normalized to obtain normalized parameters; after integration, the normalized feature set of sample infrastructure personnel is obtained.

[0041] By obtaining the maximum and minimum values ​​of the parameters in the sample infrastructure site feature set, using a specific formula for normalization, and finally integrating and arranging the normalized parameters to obtain the sample infrastructure site normalized feature set, this normalization method maps all parameters to a unified range, eliminating the impact of dimensional differences, allowing different types of parameters to be compared and analyzed on the same scale. It is more intuitive to understand the contribution of different parameters to the infrastructure site characteristics and operational risk descriptions, providing more valuable information for subsequent data analysis and model building.

[0042] The process of obtaining the parameter information entropy corresponding to each parameter in the sample infrastructure site level feature set based on the sample infrastructure site level feature set is as follows: first determine the number of calibrated levels, then for any infrastructure site characteristic parameter, count the number of levels, that is, the level distribution, and then calculate the information entropy based on the information entropy formula. For example, for the equipment failure rate, assuming that the statistics show that there are 3 low levels, 5 medium levels, and 2 high levels, for a total of 3 different level categories, according to the information entropy formula , G is the total number of level categories, in this embodiment, G is 3, l is the number of the level category, then is 0.3, is 0.5, is 0.2, then H is approximately 1.48.

[0043] By counting the grade parameters and their corresponding grade numbers, the information entropy formula is used to quantify the uncertainty or confusion of each parameter in the grade feature set of the sample infrastructure site, providing a basis and basis for subsequent data analysis and model construction based on information entropy. By calculating the information entropy of each parameter, the amount of information contained in each parameter can be intuitively understood.

[0044] An intelligent decision tree model is obtained based on the information entropy of each parameter. The specific analysis process is as follows: obtain the parameter information entropy corresponding to each level parameter in the sample infrastructure site level feature set, and mark it as the infrastructure site level parameter information entropy set; obtain the parameter information entropy corresponding to each level parameter in the sample infrastructure personnel level feature set, and mark it as the infrastructure personnel level parameter information entropy set; integrate the infrastructure site level parameter information entropy set and the infrastructure personnel level parameter information entropy set into a level parameter information entropy set; match the decision tree node parameters stored in the database based on the level parameter information entropy set; and build an intelligent decision tree model based on the decision tree node parameters.

[0045] Specifically, a table structure is specially set up in the database to store the node parameters of the decision tree. These tables will record in detail the mapping relationship between different information entropy intervals and the corresponding decision tree node parameters. For example, a "decision tree node parameter table for infrastructure site" is created, which contains fields such as "lower limit of information entropy interval", "upper limit of information entropy interval", and "node parameter of infrastructure site decision tree". Similarly, a "node parameter table for infrastructure personnel decision tree" is created with a similar structure. Taking the information entropy of the equipment failure rate at the infrastructure site as an example, assuming that the information entropy is in the interval [0,0.5], the corresponding decision tree node parameter may represent a decision branch condition, such as "if the equipment failure rate is lower than a certain threshold, enter the low-risk branch". The corresponding relationship between these information entropy intervals and node parameters is determined based on a large amount of historical data and expert experience. After collecting enough infrastructure site data and corresponding risk decision results, the node parameters that are most suitable for decision-making under different information entropy intervals are found through data analysis and expert evaluation, and stored in the database.

[0046] Take out an information entropy value from the set, for example, the calculated information entropy of the equipment failure rate is 0.35. In the "Infrastructure Site Decision Tree Node Parameter Table", traverse the "Information Entropy Interval Lower Limit" and "Information Entropy Interval Upper Limit" fields to find the interval where the information entropy value is located. Assuming that 0.35 is found to fall in the [0.3,0.5] interval, obtain the "Infrastructure Site Decision Tree Node Parameter" corresponding to the interval. This infrastructure site decision tree node parameter includes a specific condition (such as "When the equipment failure rate is greater than 10%, perform a certain decision operation") and node type (root node, internal node, and leaf node).

[0047] Step 4: Obtain the real-time infrastructure construction description set corresponding to each risk operation, obtain the intelligent risk decision value based on the real-time infrastructure construction description set and the intelligent decision tree model, and obtain the infrastructure site risk operation plan corresponding to each risk operation based on the intelligent risk decision value. The specific process is as follows: Based on the real-time infrastructure site conditions, a real-time infrastructure construction description set is obtained, which includes a real-time infrastructure site feature set and a real-time infrastructure personnel feature set; based on the real-time infrastructure site feature set, data normalization processing is performed to obtain a real-time infrastructure site normalized feature set; based on the real-time infrastructure personnel feature set, data normalization processing is performed to obtain a real-time infrastructure personnel normalized feature set. The real-time infrastructure site normalized feature set and the real-time infrastructure personnel normalized feature set can provide managers of infrastructure projects with a clearer and more intuitive data view. By real-time monitoring and analysis of the normalized data, the deviation between the current infrastructure site risk operations and the standard construction conditions and the average construction operation effect can be quickly understood.

[0048] Based on the real-time infrastructure site normalized feature set, the real-time infrastructure personnel normalized feature set and the intelligent decision tree model, an intelligent risk decision value is obtained; the intelligent risk decision value is stored as a risk decision designated label, the risk decision designated label is compared one by one with each risk decision setting label stored in the database, and the risk decision setting label that is the same as the risk decision designated label is found, and the infrastructure site risk operation plan corresponding to the risk decision setting label is obtained from the database.

[0049] The root node of the decision tree will set a partition condition based on a certain feature (such as equipment failure rate). For example, the partition condition of the root node is "whether the equipment failure rate is greater than 5%." Substitute the equipment failure rate value in the measured data into this condition for judgment.

[0050] Assuming that the equipment failure rate in the measured data is 7%, it satisfies the condition that "equipment failure rate is greater than 5%", so the data goes down along the branch corresponding to this condition.

[0051] When walking down the branch, you will encounter multiple internal nodes. Each internal node will set a division condition based on the corresponding characteristics. For example, the division condition of the next internal node is "whether the personnel fatigue index is greater than 0.6". Then substitute the personnel fatigue index value in the measured data for judgment.

[0052] Assuming that the personnel fatigue index in the measured data is 0.7, the condition of "personnel fatigue index is greater than 0.6" is met, and the data continues to move down along the branch corresponding to this condition.

[0053] After a series of internal node judgments, the final data will reach a leaf node. The leaf node represents the final decision result. The leaf node of this application will store a value. For example, if the value stored in the leaf node is 0.7, then the intelligent risk decision value is 0.7, which represents the risk score of the infrastructure project.

[0054] The intelligent decision tree model infrastructure site risk operation plan is a series of response measures and operating procedures formulated for various risk situations that may occur at the infrastructure site. It aims to ensure the safety and smooth progress of infrastructure construction, reduce the possibility of risk occurrence and the losses it brings. According to different risk levels, it stipulates the number of personal protective equipment that construction workers must wear, such as safety helmets, safety belts, protective glasses, protective gloves, etc., and clarifies their specifications and usage requirements. For high-risk operation areas (such as high-altitude operations, deep foundation pit operations, etc.), special safety protection facilities are set up, and the number and layout density of safety protection facilities are set according to different intelligent risk decision values, such as guardrails and safety nets of different heights, and an increase in the number of warning signs, to ensure the safety of operators.

[0055] Since the situation at the infrastructure site changes in real time, the risk operation plan also needs to be dynamically adjusted according to real-time data and actual conditions. When new risk situations are found or the original risk level changes through the comparison of intelligent risk decision values, the corresponding updated infrastructure site risk operation plan is obtained from the database in a timely manner to ensure that the plan can always effectively respond to the risk operation situation on site.

[0056] Example 2: Figure 2 As shown, on the basis of Example 1, it also includes: Step 5, summarizing the optimization management plans corresponding to each operation risk to obtain a set of candidate plans for each risk category, obtaining basic parameters of each optimization management plan, and obtaining a basic parameter table of plans, wherein the basic parameters of the plans include the risk value, resource requirements, and execution time of each optimization management plan; and obtaining a comprehensive score of a single plan corresponding to each optimization management plan based on the basic parameters of each optimization management plan; The method for obtaining the comprehensive score of a single scheme corresponding to one of the optimization management schemes for a certain type of risk operation is as follows: determine the risk operation corresponding to the optimization management scheme, obtain the infrastructure construction information set corresponding to the risk operation, and determine the risk value of the optimization management scheme based on the infrastructure construction information set corresponding to the risk operation, specifically: ; ; ; In the formula, The risk factor of the infrastructure site for risky operations, is the equipment failure rate, is the emergency brake response time, Maintain compliance rates for equipment, for The weight factor of for The weight factor of for The weight factor of is the personnel risk factor of risky operations, For personnel safety equipment coverage, is the personnel fatigue index, for The weight factor of for The weight factor of To optimize the risk value of the management plan; Normalize the risk value of the optimized management plan; Get comprehensive rating of a single solution , specifically: ; is the risk value of the optimized management plan after normalization, To optimize the actual demand of the management scheme for the b-type resource, is the maximum demand for the bth resource among all the optimization management solutions corresponding to the risk operation, To optimize the execution time of management solutions, is the maximum execution time of all optimization management solutions corresponding to the risk operation. for The weight factor of for The weight factor of for The weight factor of Determine the risk operation types at the infrastructure construction site, and construct a global coordination management solution combination set based on the candidate solution sets of each risk category. Each global coordination management solution combination in the global coordination management solution combination set includes an optimized management solution corresponding to each risk operation type. Determine the scheme combination conflict intensity of each global coordination management scheme combination and the coordination risk index of the global coordination management scheme combination; Conflict intensity of solution combinations The method to obtain is: ; ; ; ; in, is the resource conflict coefficient, is the time conflict coefficient, is the spatial conflict coefficient, for The weight factor of for The weight factor of for The weight factor of is the total demand for the bth resource by all the optimal management schemes in the global coordinated management scheme combination, k is an index variable used to traverse the optimal management schemes in the global coordinated management scheme combination, is the g-th job overlap duration, which refers to the overlap of the job times of any two different optimization management schemes in the global coordination management scheme combination. for The associated weight of is the total operation duration of the fth optimal management solution in the global coordination management solution combination, The total area of ​​the operating area overlapped by all the optimized management schemes in the global coordinated management scheme combination. is the operating area of ​​the fth optimal management scheme in the global coordinated management scheme combination, is the actual distance from the edge of the working area, The distance required by the safety distance specification; Synergistic risk indicators for the global coordinated management solution portfolio The method to obtain is: ; in, for The corresponding optimization management plan and The corresponding risk coupling coefficient between the optimization management schemes, For The risk value of an optimal management solution is associated with For The risk value of another associated optimal management solution; The optimal global coordination management solution combination is selected based on the minimization of the objective function and the constraints, where: The objective function is: ; In the formula, is the comprehensive score of the ath optimal management solution in the global coordinated management solution combination. To globally coordinate and manage the scheme combination conflict intensity, To provide a synergistic risk indicator for the overall coordinated management solution portfolio, for The weight factor of for The weight factor of The constraints are: ; ; In the formula, is the demand for the bth resource of the ath optimal management solution, is the maximum available amount of the bth resource, It is the upper limit of the collaborative risk index of the set global coordinated management solution combination.

[0057] Through data-driven quantitative analysis and global coordinated optimization, we upgrade from single-solution optimization to global portfolio optimization and reduce systemic risks.

[0058] Embodiment 3: Based on Embodiment 2, this embodiment further includes a dynamic conflict compensation mechanism in step 5, specifically: When it is detected that the synergy risk index of the global coordination management solution combination exceeds the preset safety threshold, the following dynamic compensation process is executed: Based on the IoT sensor equipment deployed at the construction site, multi-dimensional data streams are collected in real time to build a three-dimensional spatial topology map: Obtain the real-time coordinate vector set of all construction equipment through RFID positioning. The coordinate vector set includes the equipment number, three-dimensional space position coordinates and movement direction angle; Combined with the UWB positioning system, the motion trajectory matrix of the construction workers is collected. The matrix records the position coordinates, movement speed and stay time of each person at different time points; Use LiDAR to scan the distribution of obstacles at the construction site and generate a dynamically updated obstacle heat map, which includes obstacle boundary coordinate sets, height gradient data, and predicted trajectories of moving obstacles. Use spatiotemporal convolutional neural network to extract multi-scale features and predict risks of three-dimensional spatial topology graphs: The neural network input layer receives the spatiotemporal sequence data of the equipment coordinate vector set, the personnel motion trajectory matrix, and the obstacle heat map; The first convolutional layer extracts the relative motion features between devices and calculates the device collision probability tensor: The number of times the trajectories of device x and device y intersect in the zth time window is determined by the spatial overlap of the two motion trajectories and the temporal overlap window; The standard deviation of the real-time motion acceleration of device x is calculated based on the second-order derivative of its positioning coordinates in continuous time frames; The inverse value of the dynamic safety radius of the equipment is matched with the safe operation radius data in the equipment specification table through the equipment type parameter database; The second convolutional layer extracts the risk characteristics of personnel gathering and generates the personnel gathering risk matrix: The density change rate of the personnel group o is calculated by the distribution dispersion of the personnel coordinates in the group per unit time; The average safety equipment completeness within the group is the weighted average of the protective equipment presence rate and tightness detection values ​​uploaded by the wearable device sensors; The communication delay time between groups is obtained by extracting the time difference of the most recent data synchronization between adjacent groups from the construction communication system log; Establish a dynamic compensation function based on reinforcement learning to perform nonlinear correction on the conflict intensity: Construct a training sample set of historical conflict events. Each sample contains the original conflict intensity, equipment collision probability correction factor, personnel gathering risk correction factor, and actual accident rate. The output layer of the reinforcement learning model generates dynamic weight parameters, which are automatically adjusted according to the complexity of the real-time construction environment. The complexity is determined by the abnormal index of equipment operation status, the mutation coefficient of personnel turnover rate and the environmental noise level. The maximum value of the equipment collision probability tensor is coupled with the singular value decomposition result of the personnel gathering risk matrix to output the equipment collision probability correction coefficient and the personnel gathering risk correction coefficient; Input the corrected conflict intensity into the digital twin system to perform multi-dimensional simulation verification: Load all parameters of the current construction scheme combination in the virtual twin scene and inject the corrected conflict intensity value; Perform Monte Carlo simulation to generate X groups of random interference events, including sudden equipment failure, abnormal personnel flow, and sudden changes in weather parameters; When the proportion of scenarios that meet safety constraints in the simulation results exceeds 95%, the compensation scheme is determined to be effective and the compensation instruction is generated; The compensation instruction set includes: Equipment operation phase delay instruction: According to the equipment collision probability tensor sorting results, the operation beat phase difference control is implemented for high-risk equipment. The delay time is calculated in inverse proportion to the dynamic safety radius of the equipment and the current movement speed. Personnel movement reconstruction instructions: re-plan personnel movement paths based on the improved ant colony algorithm to avoid high-density areas marked in the personnel gathering risk matrix; Resource supply path optimization instructions: Use the dynamic A* algorithm combined with obstacle heat map data to generate a multi-level buffer supply route. The path safety factor is dynamically adjusted by the ratio of the supply vehicle size to the channel width.

Claims

1. An optimization management method for infrastructure site risk operations based on artificial intelligence algorithms, characterized in that: The following steps are involved: Step 1: Based on the construction monitoring records of each risk operation at the historical infrastructure site, obtain the historical infrastructure construction description set corresponding to each risk operation. The process is as follows: Obtain historical infrastructure construction monitoring records, including certain historical infrastructure construction conditions; Based on any historical infrastructure construction situation, obtain a historical infrastructure construction information set, including equipment failure rate, emergency brake response time, equipment maintenance compliance rate, personnel safety equipment coverage rate, and personnel fatigue index of the historical infrastructure construction situation; Mark the historical infrastructure construction information set as a historical infrastructure construction description point corresponding to the historical infrastructure construction situation; A historical infrastructure construction description set is obtained by integrating the historical infrastructure construction description points corresponding to all historical infrastructure construction conditions; Step 2: Based on the historical infrastructure construction description set, data preprocessing is performed to obtain a sample infrastructure construction description set; Step 3: Classify based on the sample infrastructure construction description set, obtain the sample infrastructure site feature set and the sample infrastructure personnel feature set, and obtain an intelligent decision tree model based on the sample infrastructure site feature set and the sample infrastructure personnel feature set; Step 4: Obtain the real-time infrastructure construction description set corresponding to each risk operation, obtain the intelligent risk decision value based on the real-time infrastructure construction description set and the intelligent decision tree model, and obtain the infrastructure site risk operation plan corresponding to each risk operation based on the intelligent risk decision value.

2. The method for optimizing and managing infrastructure site risk operations based on artificial intelligence algorithm according to claim 1 is characterized in that: The method for obtaining the personnel fatigue index is as follows: Obtaining the heart rate variability of construction workers based on wearable devices , construction site environmental parameters and mean square error of static load coefficient , construction site environmental parameters include ambient temperature mean square deviation , mean square error of ambient humidity and the mean square error of the construction location height ; Obtain the work data of construction personnel, including working hours , rest frequency and work intensity ; Based on heart rate variability , construction site environmental parameters, static load coefficient mean square error And the work data of construction personnel to determine the personnel fatigue index , the formula is: ； ； In the formula, , , , , , are weight factors, is the rest frequency parameter value, is the transfer function, is the heart rate variability parameter value, is the ambient temperature mean square error parameter value, is the parameter value of the mean square error of the ambient humidity, It is the parameter value of mean square deviation of construction location height.

3. The method for optimizing and managing infrastructure site risk operations based on artificial intelligence algorithm according to claim 1 is characterized in that: In the step 2, the process of performing data preprocessing based on the historical infrastructure construction description set to obtain the sample infrastructure construction description set is as follows: Obtain a collection of historical infrastructure construction descriptions; Obtaining the historical infrastructure construction information interval stored in the database, the historical infrastructure construction information interval describing the reasonable range of parameter values ​​in the historical infrastructure construction information set; Based on historical infrastructure construction information interval and 3 The rule is used to perform double cleaning of abnormal data and obtain the cleaned infrastructure construction description set: Based on the historical infrastructure construction information interval, the historical infrastructure construction description set is preliminarily cleaned of abnormal data. The preliminarily cleaning of abnormal data is to process the missing parameters of the historical infrastructure construction description set that exceed the historical infrastructure construction information interval and mark the missing parameter positions; Get all the historical infrastructure construction information sets in the historical infrastructure construction description set, based on any parameter in the historical infrastructure construction information set , i represents the parameter number, get The parameter mean And the parameter standard deviation ; Among them, if , then Mark as outliers and perform secondary cleaning of abnormal data. Perform missing processing and mark the missing position of parameters, Indicates the jth value of the i-th parameter; Get all the missing parameter positions and use the mean filling method to fill in the parameters; Get the cleaned infrastructure construction description set; Based on the cleaned infrastructure construction description set, data normalization is performed to obtain a sample infrastructure construction description set.

4. The method for optimizing and managing infrastructure site risk operations based on artificial intelligence algorithm according to claim 3 is characterized by: The process of obtaining all missing parameter locations and filling in the parameters using the mean filling method is: Get the position where any parameter is missing ; Get parameter type based on the missing position of the parameter; Based on the parameter type, obtain a collection of infrastructure construction conditions in the same industry with similar parameter types and complete data stored in the database; Based on the same industry and parameter type with similar infrastructure construction conditions and complete data, the missing position of the parameter Use mean filling method to fill.

5. The method for optimizing and managing infrastructure site risk operations based on artificial intelligence algorithm according to claim 1 is characterized in that: In step 3, the process of obtaining the intelligent decision tree model is as follows: Classify based on the sample infrastructure construction description set to obtain the sample infrastructure site feature set and the sample infrastructure personnel feature set; The sample infrastructure site feature set includes equipment failure rate, emergency brake response time, equipment maintenance compliance rate, and equipment maintenance compliance rate; The characteristic set of sample infrastructure personnel includes personnel safety equipment coverage and personnel fatigue index; Performing data normalization processing on the sample infrastructure site feature set and the sample infrastructure personnel feature set respectively, to obtain the sample infrastructure site normalized feature set and the sample infrastructure personnel normalized feature set; Based on the normalized feature set of the sample infrastructure site, data level marking is performed to obtain the level feature set of the sample infrastructure site; Based on the normalized feature set of sample infrastructure construction personnel, data level marking is performed to obtain the level feature set of sample infrastructure construction personnel; Based on the sample infrastructure site level feature set, obtain the parameter information entropy corresponding to each parameter in the sample infrastructure site level feature set; Based on the sample infrastructure personnel level feature set, obtain the parameter information entropy corresponding to each parameter in the sample infrastructure personnel level feature set; An intelligent decision tree model is obtained based on the information entropy of each parameter in the sample infrastructure personnel grade feature set and the sample infrastructure site grade feature set.

6. The method for optimizing and managing infrastructure site risk operations based on artificial intelligence algorithm according to claim 5 is characterized in that: The process of obtaining the normalized feature set of the sample infrastructure site and the normalized feature set of the sample infrastructure personnel is as follows: Get any parameter in the sample infrastructure site feature set ; Get Parameters The corresponding maximum value and the minimum value ; Based on the maximum value and the minimum value And parameters Perform normalization operation on any value of to obtain normalized parameters; After integration, a normalized feature set of the sample infrastructure site is obtained; Get any parameter in the sample infrastructure personnel feature set ; Get Parameters The corresponding maximum value and the minimum value ; Based on the maximum value and the minimum value And parameters Perform normalization operation on any value of to obtain normalized parameters; After integration, the normalized feature set of sample infrastructure personnel is obtained.

7. The method for optimizing and managing infrastructure site risk operations based on artificial intelligence algorithm according to claim 6 is characterized in that: The process of obtaining an intelligent decision tree model based on the information entropy of each parameter is as follows: Obtain parameter information entropy corresponding to each level parameter in the sample infrastructure site level feature set, and mark it as the infrastructure site level parameter information entropy set; Obtain parameter information entropy corresponding to each level parameter in the sample infrastructure personnel level feature set, and mark it as infrastructure personnel level parameter information entropy set; Integrate the infrastructure site level parameter information entropy set and the infrastructure personnel level parameter information entropy set into a level parameter information entropy set; Matching decision tree node parameters stored in the database based on the level parameter information entropy set; Build an intelligent decision tree model based on decision tree node parameters.

8. The method for optimizing and managing infrastructure site risk operations based on artificial intelligence algorithm according to claim 1 is characterized in that: In step 4, the process of obtaining the infrastructure site risk operation plan corresponding to each risk operation based on the intelligent risk decision value is as follows: Based on the real-time infrastructure site situation, a real-time infrastructure construction description set is obtained, where the real-time infrastructure construction description set includes a real-time infrastructure site feature set and a real-time infrastructure personnel feature set; Perform data normalization processing based on the real-time infrastructure site feature set to obtain a real-time infrastructure site normalized feature set; Performing data normalization processing based on the real-time infrastructure construction personnel feature set to obtain a real-time infrastructure construction personnel normalized feature set; Obtain intelligent risk decision values ​​based on the normalized feature set of real-time infrastructure site, the normalized feature set of real-time infrastructure personnel, and the intelligent decision tree model; The intelligent risk decision value is stored as a risk decision designated label, the risk decision designated label is compared one by one with each risk decision setting label stored in the database, the risk decision setting label that is the same as the risk decision designated label is found, and the infrastructure site risk operation plan corresponding to the risk decision setting label is obtained from the database.

9. The method for optimizing and managing infrastructure site risk operations based on artificial intelligence algorithm according to claim 1 is characterized in that: The method also includes: Step 5, summarizing the optimization management plans corresponding to each operation risk, obtaining a set of candidate plans for each risk category, obtaining basic parameters of each optimization management plan, and obtaining a basic parameter table of the plan, wherein the basic parameters of the plan include the risk value, resource requirements, and execution time of each optimization management plan; Based on the basic parameters of each optimization management scheme, the comprehensive score of each optimization management scheme corresponding to each optimization management scheme is obtained; The specific method for obtaining the comprehensive score of a single solution corresponding to one of the optimization management solutions for a certain type of risk operation is as follows: Determine the risk operation corresponding to the optimization management plan, obtain the infrastructure construction information set corresponding to the risk operation, and determine the risk value of the optimization management plan based on the infrastructure construction information set corresponding to the risk operation, specifically: ； ； ； In the formula, The risk factor of the infrastructure site for risky operations, is the equipment failure rate, is the emergency brake response time, Maintain compliance rates for equipment, , , , , is the weight factor, is the personnel risk factor of risky operations, For personnel safety equipment coverage, is the personnel fatigue index, To optimize the risk value of the management plan; Normalize the risk value of the optimized management plan; Get comprehensive rating of a single solution , specifically: ； is the risk value of the optimized management plan after normalization, To optimize the actual demand of the management scheme for the b-type resource, is the maximum demand for the bth resource among all the optimization management solutions corresponding to the risk operation, To optimize the execution time of management solutions, is the maximum execution time of all optimization management solutions corresponding to the risk operation. , , is the weight factor; Determine the risk operation types at the infrastructure construction site, and construct a global coordination management solution combination set based on the candidate solution sets of each risk category. Each global coordination management solution combination in the global coordination management solution combination set includes an optimized management solution corresponding to each risk operation type. Determine the scheme combination conflict intensity of each global coordination management scheme combination and the coordination risk index of the global coordination management scheme combination; Conflict intensity of solution combinations The method to obtain is: ； ； ； ； in, is the resource conflict coefficient, is the time conflict coefficient, is the spatial conflict coefficient, for The weight factor of for The weight factor of for The weight factor of is the total demand for the bth resource by all the optimal management schemes in the global coordinated management scheme combination, k is an index variable used to traverse the optimal management schemes in the global coordinated management scheme combination, is the overlapping duration of the g-th job, for The associated weight of is the total operation duration of the fth optimal management solution in the global coordination management solution combination, The total area of ​​the operating area overlapped by all the optimized management schemes in the global coordinated management scheme combination. is the operating area of ​​the fth optimal management scheme in the global coordinated management scheme combination, is the actual distance from the edge of the working area, The distance required by the safety distance specification; Synergistic risk indicators for the global coordinated management solution portfolio The method to obtain is: ； in, for The corresponding optimization management plan and The corresponding risk coupling coefficient between the optimization management schemes, For The risk value of an optimal management solution is associated with For The risk value of another associated optimal management solution; The optimal global coordination management solution combination is selected based on the minimization of the objective function and the constraints, where: The objective function is: ； In the formula, is the comprehensive score of the ath optimal management solution in the global coordinated management solution combination. To globally coordinate and manage the scheme combination conflict intensity, To provide a synergistic risk indicator for the overall coordinated management solution portfolio, for The weight factor of for The weight factor of The constraints are: ； ； In the formula, is the demand for the bth resource of the ath optimal management solution, It is the upper limit of the collaborative risk index of the set global coordinated management solution combination.

Images