Digital management method and system for engineering projects

Abstract

The present invention relates to the field of digital management technology, and discloses a digital management method and system for engineering projects, wherein the method comprises: collecting multi-dimensional project information of the engineering project, and performing digital modeling and real-time synchronization processing based on the multi-dimensional project information to obtain a digital model of the project; performing deep learning analysis and multi-scenario simulation on the digital model of the project to generate a project budget plan; performing fuzzy comprehensive evaluation and procurement planning on the project budget plan to obtain a dynamic supply chain plan, and performing progress management on the dynamic supply chain plan to obtain an adaptive progress plan; performing risk prediction and association rule mining on the adaptive progress plan to obtain a dynamic risk list of the project; performing knowledge graph construction and case reasoning analysis on the dynamic risk list of the project to construct an engineering project management knowledge base, and the method improves the overall efficiency and quality of project management.

Description

Technical Field

[0001] The present invention relates to the technical field of digital management, and in particular to a digital management method and system for an engineering project. Background Art

[0002] With the rapid development of information technology, engineering project management is gradually transforming towards digitalization and intelligence. Traditional engineering project management methods often seem to be unable to cope with the increasingly complex project environment. The collection and processing of project information is inefficient, making it difficult to achieve real-time monitoring and dynamic adjustment; project budget preparation lacks accuracy and flexibility, and cannot effectively respond to market changes; supply chain management and progress control are often separated, making it difficult to achieve optimal resource allocation.

[0003] In addition, project risk management remains at a static, empirical level, making it difficult to identify and respond to emerging risk factors in a timely manner. The accumulation and inheritance of project management knowledge also faces huge challenges, and a large number of valuable experiences and lessons are difficult to effectively accumulate and reuse. These problems seriously restrict the efficiency and quality of engineering projects and increase project costs and risks. Summary of the invention

[0004] The present invention provides a digital management method and system for an engineering project, which are used to improve the overall efficiency and quality of project management.

[0005] In a first aspect, the present invention provides a digital management method for an engineering project, the digital management method for an engineering project comprising:

[0006] Collect multi-dimensional project information of the engineering project, and perform digital modeling and real-time synchronization processing based on the multi-dimensional project information to obtain a digital model of the project;

[0007] Conduct deep learning analysis and multi-scenario simulation on the digital model of the project to generate a project budget plan;

[0008] Performing fuzzy comprehensive evaluation and procurement planning on the project budget plan to obtain a dynamic supply chain plan, and performing progress management on the dynamic supply chain plan to obtain an adaptive progress plan;

[0009] Perform risk prediction and association rule mining on the adaptive schedule to obtain a dynamic risk list of the project;

[0010] A knowledge graph and case reasoning analysis are performed on the dynamic risk list of the project to build an engineering project management knowledge base.

[0011] In a second aspect, the present invention provides a digital management system for an engineering project, the digital management system for the engineering project comprising:

[0012] A collection unit, used to collect multi-dimensional project information of the engineering project, and perform digital modeling and real-time synchronization processing based on the multi-dimensional project information to obtain a digital model of the project;

[0013] A simulation unit, used to perform deep learning analysis and multi-scenario simulation on the digital model of the project to generate a project budget plan;

[0014] A planning unit, used for performing fuzzy comprehensive evaluation and procurement planning on the project budget plan to obtain a dynamic supply chain plan, and performing progress management on the dynamic supply chain plan to obtain an adaptive progress plan;

[0015] A prediction unit, used to perform risk prediction and association rule mining on the adaptive schedule to obtain a dynamic risk list of the project;

[0016] A construction unit is used to construct a knowledge graph and conduct case reasoning analysis on the dynamic risk list of the project, and to build an engineering project management knowledge base.

[0017] The third aspect of the present invention provides a computer device, comprising: a memory and at least one processor, wherein the memory stores instructions; the at least one processor calls the instructions in the memory so that the computer device executes the above-mentioned digital management method for engineering projects.

[0018] A fourth aspect of the present invention provides a computer-readable storage medium, wherein the computer-readable storage medium stores instructions, which, when executed on a computer, enable the computer to execute the above-mentioned digital management method for engineering projects.

[0019] In the technical solution provided by the present invention, multi-dimensional project information is collected and processed by intelligent means to realize real-time digital expression of the project, deep learning technology is used to analyze project data, and a more accurate project budget plan is generated through multi-scenario simulation. The budget plan is evaluated by a fuzzy comprehensive evaluation method, and combined with dynamic supply chain management, the optimal allocation and flexible scheduling of resources are realized. Through discrete event simulation and constraint propagation algorithm, a schedule that can dynamically adapt to project changes is generated to improve the flexibility and efficiency of project execution. Bayesian network and association rule mining technology are used to realize dynamic prediction and analysis of project risks, and improve the foresight and accuracy of risk management. By constructing knowledge graphs and case reasoning analysis, the effective accumulation and intelligent application of project management knowledge are realized, and the scientific nature of decision-making and the efficiency of experience inheritance are improved. This method covers all aspects of project management, from information collection to knowledge accumulation, realizes intelligent management of the entire process, and significantly improves the overall efficiency and quality of project management.

[0020] Other features and advantages of the present invention will be described in the following description, and partly become apparent from the description, or understood by practicing the present invention. The purpose and other advantages of the present invention are realized and obtained by the structures particularly pointed out in the description, claims and drawings.

[0021] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, preferred embodiments are given below and described in detail with reference to the accompanying drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] Figure 1 A schematic diagram of an embodiment of a digital management method for an engineering project in an embodiment of the present invention;

[0023] Figure 2 A schematic diagram of an embodiment of a digital management system for engineering projects in an embodiment of the present invention;

[0024] Figure 3 FIG. 1 is a schematic diagram of an embodiment of a computer device in an embodiment of the present invention. DETAILED DESCRIPTION

[0025] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution of the present invention will be clearly and completely described below in conjunction with the accompanying drawings. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0026] The terms "including" and "having" and any variations thereof mentioned in the embodiments of the present invention are intended to cover non-exclusive inclusions. For example, a process, method, system, product or device end including a series of steps or units is not limited to the listed steps or units, but may optionally include other steps or units that are not listed, or may optionally include other steps or units that are inherent to these processes, methods, products or device ends.

[0027] To facilitate understanding of this embodiment, a digital management method for an engineering project disclosed in an embodiment of the present invention is first described in detail. Figure 1 As shown, the method comprises the following steps:

[0028] S1. Collect multi-dimensional project information of the engineering project, and perform digital modeling and real-time synchronous processing based on the multi-dimensional project information to obtain a digital model of the project;

[0029] It is understandable that the execution subject of the present invention may be a digital management system of an engineering project, or a terminal or a server, which is not limited here. The embodiment of the present invention is described by taking a server as the execution subject as an example.

[0030] Specifically, the project contract documents of the engineering project are intelligently scanned and parsed to extract the basic parameter set of the project. The parameter set may include basic information such as the basic schedule, budget, resource requirements, and contract terms of the project. Based on the basic parameter set of the project, the engineering project is preliminarily decomposed to generate the project work breakdown structure, which breaks down the overall goals of the project into smaller and more manageable components and tasks. The project work breakdown structure is intelligently completed and optimized. The intelligent completion process uses artificial intelligence and machine learning technology to automatically identify and fill in possible omissions and deficiencies in the work breakdown structure by analyzing historical data of similar projects and industry best practices. The optimization process is to optimize the arrangement of tasks and the allocation of resources on the basis of the existing structure to improve efficiency and reduce costs. Through this process, a complete project structure tree is obtained. The project structure tree includes all necessary work components and tasks, and clarifies the hierarchical relationship and interdependence between them. According to the complete project structure tree, the project resources are preliminarily configured to generate an initial resource allocation plan. The initial resource allocation plan is a resource allocation plan based on the needs of the project and the availability of existing resources. The initial resource allocation plan is subjected to multi-dimensional constraint analysis. This includes reviewing the demand and supply of human resources, equipment resources, material resources, etc. to identify resource conflicts and bottlenecks. By analyzing the resource conflict list, intelligent adjustments are made to resource allocation to optimize the resource allocation plan. The optimized resource allocation plan is used to make a preliminary arrangement of the project schedule to obtain the initial project schedule network. The initial project schedule network is a timetable and priority arrangement of project activities based on the project structure tree and the resource allocation plan. The schedule network can use tools such as the critical path method or project review technology to determine the critical path of the project and the earliest and latest completion time of each task. According to the initial project schedule network, identify and classify the risk factors of the project and establish a project risk factor library. Project risk factors may include time delays, cost overruns, resource shortages, quality issues, etc. Quantitative and qualitative analysis is performed on the risk factor library to assess the possibility and impact of each risk and form a project risk assessment matrix. By combining the project risk assessment matrix, the optimized resource allocation plan, and the initial project schedule network, a digital model of the project is constructed.

[0031] S2. Conduct deep learning analysis and multi-scenario simulation on the project digital model to generate a project budget plan;

[0032] Specifically, feature extraction is performed on the project digital model to obtain a project feature vector set. Relevant feature data such as project scale, complexity, required resources, and time schedule are extracted from the model. Based on the project feature vector set, similarity calculation is performed on historical project data to identify historical projects with the most similar features to the current project, and these projects are grouped into similar project sets. Cost data of similar project sets are normalized to eliminate cost differences between different projects due to factors such as scale, location, and time, so that data are comparable on the same basis, and a standardized cost data set is obtained. Based on the standardized cost data set, a project cost prediction model is constructed. The model analyzes the cost data of similar projects, learns their laws and trends, and predicts the cost of the current project. The project cost prediction model is cross-validated and parameterized. In the cross-validation process, the data is divided into multiple subsets, and the stability and reliability of the model are ensured through repeated training and validation. At the same time, the hyperparameters of the model are adjusted to find the best parameter combination to obtain an optimized cost prediction model. Using the optimized cost prediction model, the cost of each stage of the project is preliminarily predicted to obtain the initial cost estimation result. Perform Monte Carlo simulation on the initial cost estimate. Monte Carlo simulation is a technique for estimating cost uncertainty through random sampling and statistical analysis. Through a large number of simulation runs, the probability distribution of costs is obtained. Based on the cost probability distribution, multiple cost scenarios are generated to reflect the possible cost situations under different assumptions and conditions, forming a cost scenario set. Perform sensitivity analysis on the cost scenario set to identify the key factors that have the greatest impact on the total cost of the project. By analyzing the key cost influencing factors and combining the data in the cost scenario set, a dynamic programming algorithm is used to develop a project budget plan.

[0033] S3. Perform fuzzy comprehensive evaluation and procurement planning on the project budget plan to obtain a dynamic supply chain plan, and perform progress management on the dynamic supply chain plan to obtain an adaptive progress plan;

[0034] Specifically, a hierarchical analysis is conducted on the project budget plan. Through the hierarchical analysis method, a budget evaluation index system is constructed, which includes evaluation indicators in multiple aspects such as project cost, quality, time, and risk. On this basis, a fuzzy evaluation matrix is ​​constructed, and fuzzy mathematics methods are used to conduct fuzzy comprehensive evaluation of each indicator to obtain a comprehensive fuzzy evaluation result. The membership degree of the fuzzy comprehensive evaluation results is calculated, and different evaluation results are quantified into specific feasibility scores. Based on the feasibility score, the material requirements of the project are classified and clustered to form a material procurement classification list. The classification and clustering of materials not only helps to identify key materials and ordinary materials, but also helps to determine the procurement priorities and strategies of different materials. Multi-objective optimization calculations are performed on the material procurement classification list to optimize procurement decisions in multiple dimensions such as cost, time, and quality, and generate a preliminary procurement plan. Based on the preliminary procurement plan, suppliers are evaluated and screened. The evaluation criteria may include multiple factors such as the supplier's historical performance, supply capacity, and cost-effectiveness. Through comprehensive evaluation, a list of target suppliers that meet project needs is selected. Game theory analysis is performed on the target supplier list to formulate the optimal procurement strategy, including contract negotiation, price negotiation, delivery date negotiation, etc. Game theory methods are used to balance the interests of all parties to ensure the optimality and executability of the procurement strategy. According to the optimal procurement strategy, a dynamic supply chain network model is constructed. This network model not only includes the selection of suppliers and the path of material flow, but also includes the dynamic adjustment capability of each link in the supply chain, so as to generate a dynamic supply chain solution. Discrete event simulation is performed on the dynamic supply chain solution to simulate the response and performance of each link in the supply chain under different circumstances, and obtain the key node schedule of the supply chain. Key nodes include the issuance of purchase orders, the arrival of materials, the distribution of materials, etc., and potential delays and bottlenecks are estimated through simulation. According to the key node schedule of the supply chain, the constraint propagation algorithm is used to generate an adaptive schedule. The constraint propagation algorithm dynamically adjusts the project schedule by considering the time constraints and resource limitations of each task in the project, ensuring that the project can flexibly respond to various changes and uncertainties.

[0035] Event decomposition of dynamic supply chain solutions. The complex process in the supply chain is divided into a series of discrete events, such as material procurement, transportation, and acceptance. Each event has a specific start and end time, forming a discrete event sequence. Based on the discrete event sequence, an event-driven state transition model is constructed, which describes how the state of each event in the supply chain changes over time to form a supply chain simulation model. According to the supply chain simulation model, multiple sets of simulation results are generated to simulate the behavior of each part of the supply chain under different conditions and assumptions. By statistically analyzing the simulation results, the time distribution of key nodes in the supply chain is obtained. Key nodes are key time points in the supply chain that affect the progress of the project, such as material arrival, production completion, and product delivery. By analyzing the time distribution, possible bottlenecks and delays are identified. Cluster analysis is performed on the time distribution of key nodes in the supply chain to determine the time window set of different events. The time window set represents the possible completion time range of each event. Based on the time window set, a time constraint network is constructed to form an initial constraint graph. The initial constraint graph describes the time relationship and interdependence between events. The initial constraint graph is checked for arc consistency to ensure that all time constraints in the graph are compatible with each other. The inspection process helps to identify and eliminate potential contradictions and inconsistencies, and obtain the target constraint graph. According to the target constraint graph, variables are sorted and values ​​are selected to form a constraint satisfaction problem. The constraint satisfaction problem is a mathematical problem that aims to find an event schedule that satisfies all constraints. To solve this problem, backtracking search is performed. Backtracking search is a recursive algorithm that systematically tries different variable assignments to find a solution set that satisfies all constraints. Through the search process, a set of feasible solution sets are obtained, which represent possible event schedules under given constraints. Optimization is performed based on the feasible solution set to select the optimal event schedule and generate an adaptive schedule.

[0036] S4. Perform risk prediction and association rule mining on the adaptive schedule to obtain a dynamic risk list for the project;

[0037] Specifically, the adaptive schedule is task-decomposed to identify key tasks in the project, which usually have a significant impact on the overall progress and success of the project. A Bayesian network model is constructed based on the key task set obtained by task decomposition. The model is used to represent the dependencies between tasks and their association with potential risk factors. Based on the Bayesian network model, an initial risk dependency graph is generated. Parameter learning is performed on the initial risk dependency graph to determine the conditional probabilities between nodes. The conditional probability table reflects the probability of different tasks and events occurring under specific conditions. The probability distribution of various risks is calculated through these tables. The probability distribution of risk occurrence provides a quantitative assessment of potential risks, and then a threshold analysis is performed on it to identify high-risk events. The list of high-risk events includes tasks and events whose probabilities exceed the set threshold, which is a part that requires special attention in project management. A transaction database for association rule mining is constructed based on the list of high-risk events. The database records high-risk events and their associated tasks and conditions to form risk item sets. The risk item sets are processed using the Apriori algorithm to identify frequent item sets. Frequent item sets refer to tasks or events that often appear at the same time in the risk item set, which represent possible associations. Confidence and lift are calculated for frequent item sets to determine the strength and importance of association rules and obtain a set of strong association rules. Semantic analysis is performed on the set of strong association rules to build a risk association network by understanding and interpreting the meaning and context of these rules. The risk association network reveals the interrelationships and potential chain reactions between high-risk events. Based on the risk association network and the list of high-risk events, a dynamic risk list for the project is generated through a graph traversal algorithm. The graph traversal algorithm is used to identify and evaluate the key paths and nodes in the risk association network to ensure that all possible risk factors and impacts are taken into account.

[0038] S5. Construct a knowledge graph and conduct case reasoning analysis on the project dynamic risk list to build an engineering project management knowledge base.

[0039] Specifically, the dynamic risk list of the project is processed, risk entities and relationship sets are extracted, each risk element in the list is converted into an independent entity, and the relationship between these entities is defined. In this way, an initial knowledge graph is constructed to form a risk knowledge network containing risk entities and their mutual relationships. The risk knowledge network is ontologically mapped and aligned, and different risk entities and relationships are standardized to form a consistent standardized risk ontology. The standardized risk ontology is a standardized description of risk domain knowledge, ensuring that risk information from different sources and forms can be compatible and understandable with each other. According to the standardized risk ontology, historical project cases are semantically annotated, and the key information in each case is associated with the standardized risk ontology to generate a structured case library. Feature extraction and vectorization are performed on the structured case library to obtain a multi-dimensional case feature space. Using the case feature space, a k-nearest neighbor index is constructed, which is a fast retrieval model that can effectively find the historical cases that are most similar to specific problems or risk features. Through similarity calculation, a set of similar cases is extracted from the case library, which have similar risk scenarios or problem descriptions as the current project. Case adaptability analysis is performed on the similar case set to determine which historical solutions can be applied to the risk management of the current project. This includes evaluating the adaptability and effectiveness of solutions in historical cases in new situations to obtain candidate solutions. Multi-criteria decision-making evaluation is performed on candidate solutions, considering multiple factors such as cost, time, and effect, to select the optimal risk response strategy. Based on the optimal risk response strategy and standardized risk ontology, the engineering project management knowledge base is constructed using graph database technology. Graph database technology can effectively store and manage complex knowledge graph data, ensuring that the knowledge base can be quickly retrieved and analyzed.

[0040] In the embodiment of the present invention, multi-dimensional project information is collected and processed by intelligent means to realize real-time digital expression of the project, deep learning technology is used to analyze project data, and a more accurate project budget plan is generated through multi-scenario simulation. The budget plan is evaluated by a fuzzy comprehensive evaluation method, and combined with dynamic supply chain management, the optimal allocation and flexible scheduling of resources are realized. Through discrete event simulation and constraint propagation algorithm, a schedule that can dynamically adapt to project changes is generated to improve the flexibility and efficiency of project execution. Bayesian network and association rule mining technology are used to realize dynamic prediction and analysis of project risks, and improve the foresight and accuracy of risk management. By constructing knowledge graphs and case reasoning analysis, the effective accumulation and intelligent application of project management knowledge are realized, and the scientific nature of decision-making and the efficiency of experience inheritance are improved. This method covers all aspects of project management, from information collection to knowledge accumulation, realizes intelligent management of the entire process, and significantly improves the overall efficiency and quality of project management.

[0041] In a specific embodiment, the process of executing step S1 may specifically include the following steps:

[0042] Intelligently scan and parse the project contract documents of the engineering project to obtain the project basic parameter set, and perform preliminary structural decomposition of the engineering project based on the project basic parameter set to obtain the project work breakdown structure;

[0043] Intelligently complete and optimize the project work breakdown structure to obtain a complete project structure tree, and perform preliminary configuration of project resources based on the complete project structure tree to obtain an initial resource allocation plan;

[0044] Perform multi-dimensional constraint analysis on the initial resource allocation plan to obtain a resource conflict list, and make intelligent adjustments to resource allocation based on the resource conflict list to obtain an optimized resource allocation plan;

[0045] Make preliminary arrangements for the project schedule for the optimized resource allocation plan to obtain an initial project schedule network, and identify and classify project risk factors based on the initial project schedule network to obtain a project risk factor library;

[0046] Conduct quantitative and qualitative analysis on the project risk factor library to obtain the project risk assessment matrix, and build a project digital model based on the project risk assessment matrix, the optimized resource allocation plan and the initial project schedule network.

[0047] Specifically, natural language processing technology and optical character recognition technology are used to convert text data in contract documents into a structured set of basic project parameters. Basic parameters may include the total budget, duration, key milestones, resource requirements, contract terms and breach of contract clauses of the project. Through automated text parsing algorithms, these parameters are extracted and classified and stored in a database. The database serves as the input data source for subsequent project management processes. The engineering project is preliminarily decomposed according to the project basic parameter set. Project management tools such as work breakdown structure are used to decompose the project into manageable subtasks. For example, if the project involves the construction of an office building, the work breakdown structure may include major tasks such as foundation engineering, main structure, interior decoration and facility installation. The project work breakdown structure is intelligently completed and optimized. Intelligent completion refers to the use of historical project data and industry standards to identify and fill in missing parts in the work breakdown structure. For example, in foundation engineering, if soil exploration is not explicitly mentioned in the project contract, the system will recommend adding this task based on the experience of previous similar projects. Optimization involves adjusting the order of tasks and improving resource allocation to improve the overall efficiency of the project. The optimized work breakdown structure forms a complete project structure tree, which describes all subtasks, timelines and resource requirements. Based on the complete project structure tree, the initial configuration of project resources is carried out to generate an initial resource allocation plan. This includes the allocation of resources such as manpower, materials, and equipment. In order to ensure the rationality of resource allocation, a multi-dimensional constraint analysis is carried out. This analysis involves multiple dimensions such as time, budget, and resource availability. For example, if a task needs to be completed within a specific time period, but the required equipment is occupied by another project in the same time period, this situation will be marked as a resource conflict. Through multi-dimensional constraint analysis, a resource conflict list is generated, listing all potential resource conflicts. According to the resource conflict list, intelligent resource allocation is adjusted to eliminate or minimize conflicts. This includes rescheduling task time, replacing resources, or adjusting task priorities. The adjusted plan forms an optimized resource allocation plan. Based on the optimized resource allocation plan, a preliminary arrangement of the project schedule is carried out to generate an initial project schedule network. According to the initial project schedule network, project risk factors are identified and classified. The project risk factor library includes various risks that the project may face, such as technical risks, financial risks, legal risks, etc. These risk factors can be collected through various methods such as historical data analysis, expert interviews, and market research. Conduct quantitative and qualitative analysis on the project risk factor library, using the formula:

[0048] ;

[0049] in, Indicates The risk score of each risk, Indicates The probability of a risk occurring, Indicates The potential impact of each risk. Combine the project risk assessment matrix, optimized resource allocation plan and initial project schedule network to build a digital model of the project. Use graphical interfaces and data analysis tools to help project teams monitor project progress in real time, predict potential problems, and adjust project plans and strategies in a timely manner.

[0050] In a specific embodiment, the process of executing step S2 may specifically include the following steps:

[0051] Extract features from the project digital model to obtain a project feature vector set, and calculate similarity of historical project data based on the project feature vector set to obtain a similar project set;

[0052] Normalize the cost data of similar project sets to obtain a standardized cost data set, and build a project cost prediction model based on the standardized cost data set;

[0053] Cross-validate and optimize the parameters of the project cost forecasting model to obtain an optimized cost forecasting model, and use the optimized cost forecasting model to make a preliminary forecast of the costs of each stage of the project to obtain the initial cost estimation results;

[0054] Perform Monte Carlo simulation on the initial cost estimation results to obtain the cost probability distribution, and generate multiple cost scenarios based on the cost probability distribution to obtain a cost scenario set;

[0055] Conduct sensitivity analysis on the cost scenario set to obtain key cost influencing factors, and generate a project budget plan through a dynamic programming algorithm based on the key cost influencing factors and cost scenario set.

[0056] Specifically, key features are extracted from various types of project data. These features may include project scale, type, geographic location, resource allocation, construction period, etc. This information is standardized and quantified to generate a feature vector set. Assume that the project feature vector is ,in Indicates The feature vector set is analyzed to obtain the unique feature identifier of each project. The feature vector set is used to calculate the similarity of historical project data to identify historical projects similar to the current project. Similarity calculation is usually performed using methods such as Euclidean distance or cosine similarity. Assume that the feature vectors of two projects are and , then the Euclidean distance It can be expressed as:

[0057] ;

[0058] in, and The projects are and No. By calculating the distance between each project, a set of similar projects is obtained. For the cost data in the similar project set, normalization is performed to eliminate the cost differences caused by factors such as project scale and location, and a standardized cost data set is obtained. Based on the standardized cost data set, a project cost prediction model is constructed. Assuming that the project cost prediction model is a linear regression model, it can be expressed as:

[0059] ;

[0060] in, For the project cost forecast, is the intercept term, For the The regression coefficient of the feature, The project cost prediction model is cross-validated and parameter tuned. Cross-validation divides the data into multiple subsets, trains and tests the model multiple times to evaluate the stability and generalization ability of the model. Parameter tuning adjusts the parameters of the model. , find the parameter combination that minimizes the model error. The optimized cost prediction model can more accurately reflect the actual cost of the project. Through the optimized cost prediction model, the cost of each stage of the project is preliminarily predicted to obtain the initial cost estimation results. In order to evaluate the cost changes under different scenarios, the initial cost estimation results are simulated by Monte Carlo simulation. Monte Carlo simulation simulates multiple possibilities of cost through random sampling and generates cost probability distribution. Assume that the expected value of the cost estimate is and the standard deviation is , then the cost probability distribution can be expressed as:

[0061] ;

[0062] in, is the cost value, is the probability of the cost occurring. Through simulation, multiple cost scenarios are obtained to form a cost scenario set. Sensitivity analysis is performed on the cost scenario set to determine the key cost influencing factors. Sensitivity analysis identifies the variables most sensitive to cost fluctuations by evaluating the impact of each factor on the total cost. This can be achieved by calculating the partial derivative of each variable:

[0063] ;

[0064] in, is the total cost, For the Cost factors. Based on the key cost influencing factors, a dynamic programming algorithm is used to generate a project budget plan. The dynamic programming algorithm finds the path to minimize the total cost through staged decision-making.

[0065] In a specific embodiment, the process of executing step S3 may specifically include the following steps:

[0066] Conduct hierarchical analysis on the project budget plan to obtain the budget evaluation index system, and construct a fuzzy evaluation matrix based on the budget evaluation index system to obtain the fuzzy comprehensive evaluation results;

[0067] Calculate the membership degree of the fuzzy comprehensive evaluation results to obtain the feasibility score of the budget plan, and classify and cluster the project material requirements based on the feasibility score to obtain a classified list of material procurement;

[0068] Perform multi-objective optimization calculations on the material procurement classification list to obtain a preliminary procurement plan, and evaluate and screen suppliers based on the preliminary procurement plan to obtain a list of target suppliers;

[0069] Conduct game theory analysis on the target supplier list to obtain the optimal procurement strategy, and build a dynamic supply chain network model based on the optimal procurement strategy to generate a dynamic supply chain solution;

[0070] Discrete event simulation is performed on the dynamic supply chain solution to obtain the supply chain key node schedule, and based on the supply chain key node schedule, the constraint propagation algorithm is used to generate an adaptive schedule.

[0071] Specifically, a budget evaluation index system is constructed, including multiple levels of evaluation indicators, usually from the highest level goals (such as overall budget control) to secondary indicators (such as cost, quality, time, etc.) and then to more specific sub-indicators (such as material costs, labor costs, equipment leasing, etc.). Through the hierarchical analysis method, complex decision-making problems can be decomposed into simpler parts and the weights of each part can be calculated. The weight calculation formula of the hierarchical analysis method is as follows:

[0072] ;

[0073] in, It is The weight of the indicator, is the first Line The elements of the column represent The indicator is relative to The importance of indicators, is the number of indicators. A fuzzy evaluation matrix is ​​constructed based on these weights, and the fuzzy comprehensive evaluation method is used to comprehensively evaluate each budget plan. The fuzzy comprehensive evaluation method uses fuzzy set theory to combine the fuzzy membership of each evaluation indicator to obtain the fuzzy comprehensive evaluation results of each plan. Membership function The membership degree of each scheme on each indicator is defined to indicate the performance of the scheme on the indicator. The fuzzy comprehensive evaluation result is calculated by the following formula:

[0074] ;

[0075] in, is the fuzzy comprehensive evaluation result vector, is the weight vector, is the membership matrix. Each element of the membership matrix represents the membership of a budget plan on a certain indicator. According to the fuzzy comprehensive evaluation results, the feasibility score of each budget plan is calculated. The feasibility score is obtained by weighted average of the membership of each plan, and its formula is as follows:

[0076] ;

[0077] in, It is The feasibility rating of each plan, It is The weight of the indicator, It is The scheme is in The membership degree of each indicator. According to the feasibility score, the material requirements of the project are classified and clustered to form a classified list of material procurement. Multi-objective optimization calculations are performed on the classified list of material procurement. Multi-objective optimization problems usually require considering multiple objectives at the same time, such as minimizing cost, maximizing quality, and minimizing delivery time. Use the linear weighted sum method to combine multiple objective functions into a comprehensive objective function. The formula is as follows:

[0078] ;

[0079] in, is the comprehensive objective function, It is The weight of the target, It is Objective function. By optimizing the comprehensive objective function, a preliminary procurement plan is obtained. Based on the preliminary procurement plan, suppliers are evaluated and screened to form a target supplier list. Supplier evaluation usually includes a comprehensive evaluation of multiple factors such as price, delivery capability, and historical performance. In order to formulate the optimal procurement strategy, a game theory analysis is performed on the target supplier list. Game theory analysis takes into account the competition and cooperation relationship between suppliers and optimizes the procurement strategy. The formula is as follows:

[0080] ;

[0081] in, It is The utility value of a supplier, It is Suppliers and The intensity of cooperation in each project It is The procurement probability of a project. Find the optimal procurement strategy by optimizing the utility value. Construct a dynamic supply chain network model based on the optimal procurement strategy. The model includes the selection of suppliers, the flow path of materials and the time arrangement. Generate a dynamic supply chain solution based on the dynamic supply chain network model. Perform discrete event simulation on the dynamic supply chain solution to simulate various events and changes in actual operation and obtain the key node schedule of the supply chain. Discrete event simulation takes into account the uncertainty and randomness in the supply chain and uses an event-driven approach to simulate the behavior and interaction of each node. According to the key node schedule of the supply chain, an adaptive schedule is generated using the constraint propagation algorithm. The constraint propagation algorithm optimizes the overall schedule of the project by considering the interdependence and time constraints between tasks. The formula is as follows:

[0082] ;

[0083] in, It's a task The start time of is the completion time of all predecessor tasks, It's a task By adjusting the start and end times of each task, adaptive scheduling can flexibly respond to changes and uncertainties in the supply chain and ensure smooth project execution.

[0084] In a specific embodiment, the execution step performs discrete event simulation on the dynamic supply chain solution to obtain a supply chain key node schedule, and the process of generating an adaptive schedule using a constraint propagation algorithm based on the supply chain key node schedule can specifically include the following steps:

[0085] Decompose the dynamic supply chain solution into events to obtain a discrete event sequence, and build an event-driven state transition model based on the discrete event sequence to obtain a supply chain simulation model;

[0086] Generate multiple sets of simulation results based on the supply chain simulation model, and perform statistical analysis based on the multiple sets of simulation results to obtain the time distribution of key nodes in the supply chain;

[0087] Perform cluster analysis on the time distribution of key nodes in the supply chain to obtain a set of time windows, and build a time constraint network based on the set of time windows to obtain an initial constraint graph;

[0088] Perform arc consistency check on the initial constraint graph to obtain the target constraint graph, and perform variable sorting and value selection based on the target constraint graph to obtain the constraint satisfaction problem;

[0089] The constraint satisfaction problem is backtracked to obtain a feasible solution set, and optimization is performed based on the feasible solution set to obtain an adaptive schedule.

[0090] Specifically, each activity and process in the supply chain is decomposed into discrete events. These events may include the procurement, transportation, storage, production and processing, and distribution of materials. By decomposing these events, a discrete event sequence is formed. Each event is independent in time and space, and its time, place, participating resources, and completion conditions can be described separately. For example, in a manufacturing project, a discrete event sequence may include the issuance of orders for raw materials, the arrival of goods, warehousing processing, the start of production, and the delivery of finished products. Based on the discrete event sequence, an event-driven state transition model is constructed. This model describes the relationship between events in the supply chain and the state transition mechanism. The state transition model is usually represented by a directed graph, where nodes represent the state of the system and edges represent the state changes caused by events. The conditions and probabilities of setting state transitions can be expressed as follows:

[0091] ;

[0092] in, Indicates that the system is at time status, Indicates at time The events that occurred, Is the status from Transfer to The probability of . Generate multiple sets of simulation results based on the supply chain simulation model. Generate different sets of simulation results by simulating different event sequences and state transition paths. Each set of simulation results represents the performance of the supply chain under different circumstances, such as different supplier delivery times, transportation delays, production failures, etc. Perform statistical analysis on the simulation results to obtain the time distribution of key nodes in the supply chain. Key nodes are the points in the supply chain that have the greatest impact on the overall operation, such as material arrival time, production start time, finished product delivery time, etc. Perform cluster analysis on the time distribution of key nodes to obtain a set of time windows. The set of time windows represents the possible time range for each key event to occur, and the window helps to consider possible time fluctuations and uncertainties in planning. Based on the set of time windows, a time constraint network is constructed. The time constraint network uses nodes to represent events and edges to represent time constraints, such as the earliest start time and the latest end time. In the time constraint network, the initial constraint graph represents all possible time arrangements and constraints. To ensure the consistency between all time constraints, an arc consistency check is performed. The arc consistency check ensures that all possible time arrangements are feasible by verifying whether there are conflicts between constraints. For example, if a time window requires event A to occur before event B, but another constraint requires event B to occur before event A, this is a contradiction that needs to be resolved by adjusting the constraints. The result of the arc consistency check is the target constraint graph, which represents the state where all constraints have been adjusted to be consistent. According to the target constraint graph, each variable (event time) is sorted and value selected to form a constraint satisfaction problem. Constraint satisfaction problems are a type of mathematical problem that involves finding a combination of variables that satisfies a set of constraints. Backtracking search is performed on the constraint satisfaction problem to find all feasible solution sets by systematically trying and excluding different time arrangement combinations. These solution sets represent the possible supply chain operation schedules under all constraints. Optimization is performed based on the feasible solution set to obtain an adaptive schedule. The adaptive schedule not only takes into account the current supply chain conditions and constraints, but also can be dynamically adjusted according to changes in simulation results and actual conditions.

[0093] In a specific embodiment, the process of executing step S4 may specifically include the following steps:

[0094] Decompose the adaptive schedule into tasks to obtain a set of key tasks, build a Bayesian network model based on the set of key tasks, and generate an initial risk dependency graph;

[0095] Perform parameter learning on the initial risk dependency graph to obtain a conditional probability table, and calculate the risk occurrence probability distribution based on the conditional probability table;

[0096] Perform threshold analysis on the probability distribution of risk occurrence to obtain a list of high-risk events, and build a transaction database for association rule mining based on the list of high-risk events to obtain a risk item set;

[0097] Based on the Apriori algorithm, the risk item set is processed to obtain the frequent item set, and the confidence and lift are calculated based on the frequent item set to obtain the strong association rule set;

[0098] Semantic analysis is performed on the strong association rule set to obtain the risk association network. Based on the risk association network and the list of high-risk events, a dynamic risk list of the project is generated through a graph traversal algorithm.

[0099] Specifically, the overall schedule of the project is broken down into smaller tasks, which usually include the main activities and milestones of the project. Through task decomposition, a set of key tasks that are important to the success of the project is identified. Key tasks may involve multiple aspects, such as resource allocation, time management, technology development, etc. By analyzing key tasks, potential risk factors are preliminarily identified. A Bayesian network model is constructed based on the set of key tasks. A Bayesian network is a probabilistic graph model used to represent conditional dependencies between a set of variables. In this model, nodes represent key tasks or events, and directed edges represent dependencies between these tasks or events. For example, if the completion of a task depends on the completion of another task, there is a directed edge from the former to the latter in the network. The initial risk dependency graph is a diagram of the Bayesian network constructed based on the set of key tasks. Parameter learning is performed on the initial risk dependency graph to determine the conditional probabilities between nodes. The conditional probability table describes the probability of an event occurring under the condition that another event occurs. For each node , the conditional probability can be expressed as ,in is a preceding event or condition. For example, if the task Delays may cause tasks The delay is Representation Task Delayed task Probability of delay. Parameter learning is performed through historical data and expert knowledge to fill in the values ​​in the conditional probability table. According to the conditional probability table, the probability distribution of each risk is calculated. Through Bayesian reasoning, the overall risk probability of each key task or event under various conditions is calculated. For example, if a supplier's delay (event B) causes a production line shutdown (event A), and historical data shows that the supplier's on-time rate is 80%, the risk probability of the production line shutdown can be calculated as:

[0100] ;

[0101] in, is the probability of supplier delay, is the probability that the supplier delivers on time, is the probability that the production line may still be shut down when the supplier delivers on time. Threshold analysis is performed on the probability distribution of these risks to identify high-risk events. High-risk events are those events whose probability of occurrence exceeds the preset threshold and may have a significant impact on the project. Based on the list of high-risk events, a transaction database for association rule mining is constructed. The transaction database records each event and other events associated with it to form a risk item set. The Apriori algorithm is used to process the risk item set to identify frequent item sets. Frequent item sets refer to combinations of events that often appear at the same time in the risk item set, and these combinations may represent potential risk associations. For example, if supplier delays (event B) and equipment failures (event C) often occur at the same time, they constitute a frequent item set. Based on the frequent item sets, the confidence and lift are calculated to obtain a strong association rule set. Confidence It represents the probability of event Y occurring when event X occurs. The calculation formula is:

[0102] ;

[0103] Lift The influence of event X on event Y is measured by the following calculation formula:

[0104] ;

[0105] The calculation formula helps identify which event combinations have strong correlations. Semantic analysis of the strong correlation rule set results in a risk correlation network that shows the interrelationships between different risk events and reveals potential risk propagation paths. For example, if the analysis shows that delays from certain suppliers often lead to production line shutdowns, this information can help management take preventive measures in advance. Based on the risk correlation network and the list of high-risk events, a dynamic risk list for the project is generated through a graph traversal algorithm. The graph traversal algorithm traverses the entire risk correlation network to identify and record all potential risk paths and key nodes.

[0106] In a specific embodiment, the process of executing step S5 may specifically include the following steps:

[0107] Process the dynamic risk list of the project to obtain risk entities and relationship sets, and build an initial knowledge graph based on the risk entities and relationship sets to obtain a risk knowledge network;

[0108] Perform ontology mapping and alignment on the risk knowledge network to obtain a standardized risk ontology, and semantically annotate historical project cases based on the standardized risk ontology to obtain a structured case library;

[0109] Feature extraction and vectorization are performed on the structured case library to obtain the case feature space, and a k-nearest neighbor index is constructed based on the case feature space to obtain a fast retrieval model;

[0110] Calculate the similarity of the fast retrieval model to obtain a similar case set, and perform case adaptability analysis based on the similar case set to obtain candidate solutions;

[0111] A multi-criteria decision-making evaluation is performed on candidate solutions to obtain the optimal risk response strategy. Based on the optimal risk response strategy and standardized risk ontology, an engineering project management knowledge base is constructed through graph database technology.

[0112] Specifically, risk entities and relationship sets are extracted from the project dynamic risk list. Risk entities can be specific events, situations or conditions, such as "supply chain disruption", "cost overrun", etc., while relationship sets describe the interactions and dependencies between these entities, such as "cause", "trigger", etc. By analyzing the project dynamic risk list, all potential risk entities and their relationships are identified. Based on the identified risk entities and relationship sets, an initial knowledge graph is constructed. A knowledge graph is a graph structure that represents entities and their relationships, which systematically organizes risk information in a project. The nodes in the graph represent risk entities, and the edges represent the relationships between entities. For example, if "supply chain disruption" may lead to "project delay", there will be an edge of "cause" relationship between the two entities. The initial knowledge graph is ontology mapped and aligned. Ontology mapping refers to mapping concepts and relationships in the knowledge graph to a standardized risk ontology to ensure information consistency and interoperability. The standardized risk ontology is a normative framework that defines basic concepts and relationships in the field of risk management. Through ontology alignment, specific risk instances in the project are matched with standardized risk categories. For example, "material shortage" can be mapped to the "supply chain risk" category in the standardized risk ontology. According to the standardized risk ontology, historical project cases are semantically annotated to obtain a structured case library. The information in the historical cases is classified and labeled according to the standardized ontology so that the information conforms to the unified standards in structure and content. Feature extraction and vectorization are performed on the structured case library to obtain a case feature space. Key features are extracted from the cases, such as project type, scale, risk type, etc., and these features are digitized to form a vector representation. Assume that the feature vector of each case is ,in Indicates The numerical value of each feature. Based on the feature vector, a k-nearest neighbor index (k-NN) is constructed to form a fast retrieval model. The k-nearest neighbor index is a method for finding the case that is most similar to a given input. The principle is to find the k cases closest to the input vector. Using the fast retrieval model, similarity calculation is performed to obtain a set of similar cases. Similarity calculations usually use metrics such as Euclidean distance or cosine similarity. For example, the Euclidean distance can be calculated using the following formula:

[0113] ;

[0114] in, and are the eigenvectors for the two cases, Represents the similarity between the two cases. Through this process, the historical case that is most similar to the current project risk situation is identified. Case adaptability analysis is performed based on the set of similar cases to evaluate the applicability of the response strategies in similar cases in the current project. Factors considered include project conditions, risk types, and effectiveness of response measures in the case. The result of the case adaptability analysis is a set of candidate solutions, which are potentially effective risk response strategies borrowed from historical cases. A multi-criteria decision-making evaluation is performed on the candidate solution set to determine the best risk response strategy. The evaluation process comprehensively considers multiple decision criteria, such as cost-effectiveness, time efficiency, risk reduction effect, etc. The comprehensive score of each candidate solution is evaluated through weighted scoring or other multi-criteria decision-making methods to select the best solution. Assume that the weight of each criterion is , the scheme's score on each criterion is , the total score can be expressed as:

[0115] ;

[0116] Based on the optimal risk response strategy and standardized risk ontology, the project management knowledge base is constructed through graph database technology. Graph database is a tool for efficient storage and query of graph structure data, which is suitable for the storage of knowledge graphs. The knowledge base not only stores standardized risk ontology and specific risk instances, but also includes the response strategy and effect evaluation of each risk instance.

[0117] The digital management method of the engineering project in the embodiment of the present invention is described above. The digital management system of the engineering project in the embodiment of the present invention is described below. Figure 2 In one embodiment of the present invention, a digital management system for an engineering project includes:

[0118] The collection unit 1 is used to collect multi-dimensional project information of the engineering project, and perform digital modeling and real-time synchronization processing based on the multi-dimensional project information to obtain a digital model of the project;

[0119] Simulation unit 2, used to perform deep learning analysis and multi-scenario simulation on the project digital model to generate a project budget plan;

[0120] Planning unit 3 is used to perform fuzzy comprehensive evaluation and procurement planning on the project budget plan to obtain a dynamic supply chain plan, and to perform progress management on the dynamic supply chain plan to obtain an adaptive progress plan;

[0121] Prediction unit 4, used to perform risk prediction and association rule mining on the adaptive schedule to obtain a dynamic risk list of the project;

[0122] Construction unit 5 is used to construct a knowledge graph and conduct case reasoning analysis on the project dynamic risk list to build an engineering project management knowledge base.

[0123] In this embodiment, for the specific implementation of each unit in the above device embodiment, please refer to the above method embodiment, which will not be repeated here.

[0124] Reference Figure 3 In an embodiment of the present invention, a computer device is also provided. The computer device may be a server, and its internal structure may be as follows: Figure 3 As shown. The computer device includes a processor, a memory, a display screen, an input device, a network interface and a database connected through a system bus. Among them, the processor designed by the computer is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The database of the computer device is used to store the corresponding data in this embodiment. The network interface of the computer device is used to communicate with an external terminal through a network connection. When the computer program is executed by the processor, the above method is implemented.

[0125] Those skilled in the art will understand that Figure 3 The structure shown in the figure is merely a block diagram of a portion of the structure related to the solution of the present invention, and does not constitute a limitation on the computer device to which the solution of the present invention is applied.

[0126] An embodiment of the present invention further provides a computer-readable storage medium on which a computer program is stored, and when the computer program is executed by a processor, the above method is implemented. It can be understood that the computer-readable storage medium in this embodiment can be a volatile readable storage medium or a non-volatile readable storage medium.

[0127] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, storage, database or other media provided by the present invention and used in the embodiments may include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. As an illustration and not limitation, RAM is available in a variety of forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double-speed data rate SDRAM (SSRSDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM.

[0128] It should be noted that, in this article, the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, device, article or method including a series of elements includes not only those elements, but also includes other elements not explicitly listed, or also includes elements inherent to such process, device, article or method. In the absence of further restrictions, an element defined by the sentence "includes a ..." does not exclude the presence of other identical elements in the process, device, article or method including the element.

[0129] The above description is only a preferred embodiment of the present invention, and does not limit the patent scope of the present invention. Any equivalent structure or equivalent process transformation made by using the contents of the present invention specification and drawings, or directly or indirectly applied in other related technical fields, are also included in the patent protection scope of the present invention.

Claims

1. A digital management method for an engineering project, characterized in that: The method comprises: Collect multi-dimensional project information of the engineering project, and perform digital modeling and real-time synchronization processing based on the multi-dimensional project information to obtain a digital model of the project; specifically, the method includes: intelligently scanning and parsing the project contract documents of the engineering project to obtain a basic project parameter set, and performing preliminary structural decomposition of the engineering project according to the basic project parameter set to obtain a project work breakdown structure; intelligently completing and optimizing the project work breakdown structure to obtain a complete project structure tree, and performing preliminary configuration of project resources according to the complete project structure tree to obtain an initial resource allocation plan; performing multi-dimensional constraint analysis on the initial resource allocation plan to obtain a resource conflict list, and performing intelligent adjustment of resource allocation according to the resource conflict list to obtain an optimized resource allocation plan; performing preliminary project schedule arrangement on the optimized resource allocation plan to obtain an initial project schedule network, and identifying and classifying project risk factors according to the initial project schedule network to obtain a project risk factor library; performing quantitative and qualitative analysis on the project risk factor library to obtain a project risk assessment matrix, and constructing a digital model of the project according to the project risk assessment matrix, the optimized resource allocation plan and the initial project schedule network; Perform deep learning analysis and multi-scenario simulation on the digital model of the project to generate a project budget plan; specifically including: extracting features from the digital model of the project to obtain a project feature vector set, and performing similarity calculation on historical project data based on the project feature vector set to obtain a similar project set; normalizing the cost data of the similar project set to obtain a standardized cost data set, and constructing a project cost prediction model based on the standardized cost data set; cross-validating and parameter tuning the project cost prediction model to obtain an optimized cost prediction model, and using the optimized cost prediction model to perform a preliminary prediction of the costs of each stage of the project to obtain an initial cost estimation result; performing Monte Carlo simulation on the initial cost estimation result to obtain a cost probability distribution, and generating multiple cost scenarios based on the cost probability distribution to obtain a cost scenario set; performing sensitivity analysis on the cost scenario set to obtain key cost influencing factors, and generating a project budget plan through a dynamic programming algorithm based on the key cost influencing factors and the cost scenario set; Perform fuzzy comprehensive evaluation and procurement planning on the project budget plan to obtain a dynamic supply chain plan, and perform progress management on the dynamic supply chain plan to obtain an adaptive progress plan; specifically, the method includes: performing hierarchical analysis on the project budget plan to obtain a budget evaluation index system, and constructing a fuzzy evaluation matrix based on the budget evaluation index system to obtain a fuzzy comprehensive evaluation result; performing membership calculation on the fuzzy comprehensive evaluation result to obtain a feasibility score of the budget plan, and classifying and clustering the project material requirements according to the feasibility score to obtain a material procurement classification list; performing multi-objective optimization calculation on the material procurement classification list to obtain a preliminary procurement plan, and evaluating and screening suppliers according to the preliminary procurement plan to obtain a target supplier list; performing game theory analysis on the target supplier list to obtain an optimal procurement strategy, and clustering the material requirements according to the optimal procurement strategy Construct a dynamic supply chain network model and generate a dynamic supply chain solution; perform event decomposition on the dynamic supply chain solution to obtain a discrete event sequence, and construct an event-driven state transition model based on the discrete event sequence to obtain a supply chain simulation model; generate multiple groups of simulation results based on the supply chain simulation model, and perform statistical analysis based on the multiple groups of simulation results to obtain the time distribution of key nodes in the supply chain; perform cluster analysis on the time distribution of key nodes in the supply chain to obtain a time window set, and construct a time constraint network based on the time window set to obtain an initial constraint graph; perform arc consistency check on the initial constraint graph to obtain a target constraint graph, and perform variable sorting and value selection based on the target constraint graph to obtain a constraint satisfaction problem; perform backtracking search on the constraint satisfaction problem to obtain a feasible solution set, and optimize based on the feasible solution set to obtain an adaptive schedule; Perform risk prediction and association rule mining on the adaptive schedule to obtain a dynamic risk list of the project; specifically, the process includes: performing task decomposition on the adaptive schedule to obtain a set of key tasks, constructing a Bayesian network model based on the set of key tasks and generating an initial risk dependency graph; performing parameter learning on the initial risk dependency graph to obtain a conditional probability table, and calculating a risk occurrence probability distribution based on the conditional probability table; performing threshold analysis on the risk occurrence probability distribution to obtain a list of high-risk events, and constructing a transaction database for association rule mining based on the list of high-risk events to obtain a risk item set; processing the risk item set based on the Apriori algorithm to obtain a frequent item set, and calculating confidence and lift based on the frequent item set to obtain a strong association rule set; performing semantic analysis on the strong association rule set to obtain a risk association network, and generating a dynamic risk list of the project through a graph traversal algorithm based on the risk association network and the list of high-risk events; The knowledge graph and case reasoning analysis are performed on the dynamic risk list of the project to construct an engineering project management knowledge base; specifically, the process includes: processing the dynamic risk list of the project to obtain risk entities and relationship sets, and constructing an initial knowledge graph based on the risk entities and the relationship sets to obtain a risk knowledge network; ontology mapping and alignment of the risk knowledge network to obtain a standardized risk ontology, and semantically annotating historical project cases based on the standardized risk ontology to obtain a structured case library; feature extraction and vectorization are performed on the structured case library to obtain a case feature space, and a k-nearest neighbor index is constructed based on the case feature space to obtain a rapid retrieval model; similarity calculation is performed on the rapid retrieval model to obtain a similar case set, and case adaptability analysis is performed based on the similar case set to obtain candidate solutions; multi-criteria decision evaluation is performed on the candidate solutions to obtain an optimal risk response strategy, and based on the optimal risk response strategy and the standardized risk ontology, an engineering project management knowledge base is constructed through graph database technology.

2. A digital management system for engineering projects, characterized in that: For executing the digital management method of the engineering project as claimed in claim 1, the system comprises: A collection unit is used to collect multi-dimensional project information of the engineering project, and perform digital modeling and real-time synchronization processing based on the multi-dimensional project information to obtain a digital model of the project; specifically, it includes: intelligently scanning and parsing the project contract documents of the engineering project to obtain a project basic parameter set, and performing preliminary structural decomposition of the engineering project according to the project basic parameter set to obtain a project work breakdown structure; intelligently completing and optimizing the project work breakdown structure to obtain a complete project structure tree, and performing preliminary configuration of project resources according to the complete project structure tree to obtain an initial resource allocation plan; performing multi-dimensional constraint analysis on the initial resource allocation plan to obtain a resource conflict list, and performing intelligent adjustment of resource allocation according to the resource conflict list to obtain an optimized resource allocation plan; performing preliminary project schedule arrangement on the optimized resource allocation plan to obtain an initial project schedule network, and identifying and classifying project risk factors according to the initial project schedule network to obtain a project risk factor library; performing quantitative and qualitative analysis on the project risk factor library to obtain a project risk assessment matrix, and constructing a digital model of the project according to the project risk assessment matrix, the optimized resource allocation plan and the initial project schedule network; A simulation unit, used for performing deep learning analysis and multi-scenario simulation on the digital model of the project to generate a project budget plan; specifically comprising: extracting features from the digital model of the project to obtain a project feature vector set, and performing similarity calculation on historical project data based on the project feature vector set to obtain a similar project set; normalizing the cost data of the similar project set to obtain a standardized cost data set, and constructing a project cost prediction model based on the standardized cost data set; cross-validating and parameter tuning the project cost prediction model to obtain an optimized cost prediction model, and using the optimized cost prediction model to perform a preliminary prediction of the costs of each stage of the project to obtain an initial cost estimation result; performing Monte Carlo simulation on the initial cost estimation result to obtain a cost probability distribution, and generating multiple cost scenarios based on the cost probability distribution to obtain a cost scenario set; performing sensitivity analysis on the cost scenario set to obtain key cost influencing factors, and generating a project budget plan through a dynamic programming algorithm based on the key cost influencing factors and the cost scenario set; The planning unit is used to perform fuzzy comprehensive evaluation and procurement planning on the project budget plan to obtain a dynamic supply chain plan, and to perform progress management on the dynamic supply chain plan to obtain an adaptive progress plan; specifically, the planning unit includes: performing hierarchical analysis on the project budget plan to obtain a budget evaluation index system, and constructing a fuzzy evaluation matrix based on the budget evaluation index system to obtain a fuzzy comprehensive evaluation result; performing membership calculation on the fuzzy comprehensive evaluation result to obtain a feasibility score of the budget plan, and classifying and clustering the project material requirements according to the feasibility score to obtain a material procurement classification list; performing multi-objective optimization calculation on the material procurement classification list to obtain a preliminary procurement plan, and evaluating and screening suppliers according to the preliminary procurement plan to obtain a target supplier list; performing game theory analysis on the target supplier list to obtain an optimal procurement strategy, and clustering the material requirements according to the optimal strategy; The procurement strategy constructs a dynamic supply chain network model and generates a dynamic supply chain solution; the dynamic supply chain solution is decomposed into events to obtain a discrete event sequence, and an event-driven state transition model is constructed based on the discrete event sequence to obtain a supply chain simulation model; multiple groups of simulation results are generated based on the supply chain simulation model, and statistical analysis is performed based on the multiple groups of simulation results to obtain the time distribution of key nodes in the supply chain; cluster analysis is performed on the time distribution of key nodes in the supply chain to obtain a time window set, and a time constraint network is constructed based on the time window set to obtain an initial constraint graph; arc consistency check is performed on the initial constraint graph to obtain a target constraint graph, and variable sorting and value selection are performed based on the target constraint graph to obtain a constraint satisfaction problem; backtracking search is performed on the constraint satisfaction problem to obtain a feasible solution set, and optimization is performed based on the feasible solution set to obtain an adaptive schedule; A prediction unit is used to perform risk prediction and association rule mining on the adaptive schedule to obtain a dynamic risk list of the project; specifically comprising: performing task decomposition on the adaptive schedule to obtain a set of key tasks, constructing a Bayesian network model based on the set of key tasks and generating an initial risk dependency graph; performing parameter learning on the initial risk dependency graph to obtain a conditional probability table, and calculating a risk occurrence probability distribution based on the conditional probability table; performing threshold analysis on the risk occurrence probability distribution to obtain a list of high-risk events, and constructing a transaction database for association rule mining based on the list of high-risk events to obtain a risk item set; processing the risk item set based on the Apriori algorithm to obtain a frequent item set, and calculating confidence and lift based on the frequent item set to obtain a strong association rule set; performing semantic analysis on the strong association rule set to obtain a risk association network, and generating a dynamic risk list of the project through a graph traversal algorithm based on the risk association network and the high-risk event list; A construction unit is used to construct a knowledge graph and perform case reasoning analysis on the dynamic risk list of the project to construct an engineering project management knowledge base; specifically comprising: processing the dynamic risk list of the project to obtain risk entities and relationship sets, and constructing an initial knowledge graph based on the risk entities and the relationship sets to obtain a risk knowledge network; ontology mapping and alignment of the risk knowledge network to obtain a standardized risk ontology, and semantically annotating historical project cases based on the standardized risk ontology to obtain a structured case library; feature extraction and vectorization of the structured case library to obtain a case feature space, and constructing a k-nearest neighbor index based on the case feature space to obtain a rapid retrieval model; similarity calculation of the rapid retrieval model to obtain a similar case set, and case adaptability analysis based on the similar case set to obtain candidate solutions; multi-criteria decision evaluation of the candidate solutions to obtain an optimal risk response strategy, and based on the optimal risk response strategy and the standardized risk ontology, constructing an engineering project management knowledge base through graph database technology.

3. A computer device, characterized in that: The computer device comprises: a memory and at least one processor, wherein instructions are stored in the memory; The at least one processor calls the instructions in the memory to enable the computer device to execute the digital management method for the engineering project as claimed in claim 1.

4. A computer-readable storage medium having instructions stored thereon, characterized in that: When the instructions are executed by the processor, the digital management method for the engineering project as claimed in claim 1 is implemented.

Images