Building engineering intelligent decision-making and risk early warning method and system

Abstract

The invention relates to the technical field of constructional engineering, and discloses a constructional engineering intelligent decision-making and risk early warning method and system, and the method comprises the steps: constructing a construction environment influence model based on building related data, and determining a preliminary task plan scheme according to the influence result of the environment change predicted by the construction environment influence model on the constructional engineering; and in combination with the dynamic risk prediction model, potential risks in the construction process are pre-warned and predicted, so that scientific decision support is provided for engineering managers, smooth proceeding of construction tasks is ensured, and the risk of a project is remarkably reduced.

Description

Technical Field

[0001] The present invention relates to the technical field of construction engineering, and particularly relates to a method and system for intelligent decision-making and risk early warning in construction engineering. Background Art

[0002] In modern construction engineering, project management is a key link to ensure the smooth progress of the project. During the construction process, environmental factors such as meteorological changes and geological conditions have a significant impact on the construction progress and construction safety. Traditional construction environment management lacks real-time prediction and accurate simulation of environmental changes, resulting in a slow response from the construction party when facing sudden changes, making it difficult to adjust the construction plan or optimize resource allocation in a timely manner, thus increasing the time cost and risk of the project. In addition, complex construction task scheduling and resource allocation are also involved in construction projects. How to ensure construction safety while efficiently and reasonably scheduling tasks is an urgent problem to be solved. Summary of the Invention

[0003] In view of this, the purpose of the present invention is to provide a method and system for intelligent decision-making and risk early warning in construction engineering to solve the problem that it is difficult to efficiently and reasonably schedule construction tasks while ensuring construction safety during the construction of construction engineering.

[0004] The first aspect of the present invention discloses a method for intelligent decision-making and risk early warning in construction engineering, and the method includes the following steps:

[0005] S01. Collect construction-related data of the construction project, where the construction-related data includes environmental data, construction progress data, resource allocation data, and project quality data;

[0006] S02. Construct a construction environment impact model based on the collected historical construction-related data;

[0007] S03. Determine the construction characteristic parameters of the target construction project, input the construction characteristic parameters and the environmental data of the target construction project into the construction environment impact model, and predict the impact of environmental changes on the construction progress, project quality, and resource consumption of the target construction project through the construction environment impact model as the first prediction result;

[0008] S04. Determine the construction tasks of the target construction project, decompose the construction tasks of the target construction project, and perform priority ranking on the construction tasks based on the first prediction result to generate a preliminary task plan;

[0009] S05. Obtain the construction project data of the historical construction project, construct a dynamic risk prediction model based on the construction project data, input the construction-related data of the target construction project and the preliminary task plan into the dynamic risk prediction model, and predict the risk points during the construction process through the dynamic risk prediction model;

[0010] S06. Adjust the preliminary task plan based on the risk points and carry out construction based on the adjusted task plan.

[0011] Furthermore, the environmental data includes meteorological data, geological data, noise data, and vibration data; the construction progress data includes the completion status of construction processes; the resource allocation data includes the usage of construction machinery and equipment, the situation of staff, and material inventory; the project quality data includes construction quality inspection reports and non-conforming product detection data.

[0012] Furthermore, the training process of the construction environment impact model includes:

[0013] Determine the construction projects corresponding to the historical building-related data, determine the construction characteristic parameters of the historical building projects, extract the characteristics of the building-related data in units of construction projects, and combine the construction characteristic parameters as a comprehensive feature set; the construction projects corresponding to the historical building-related data are completed construction projects.

[0014] The construction characteristic parameters include the construction stage, construction method, construction period, and construction resources of the construction project.

[0015] Furthermore, the feature extraction operation of the building-related data includes:

[0016] Perform data cleaning, normalization, and missing value filling operations on the building-related data.

[0017] Align the environmental data, construction progress data, resource allocation data, and project quality data in time series to generate a first time series dataset.

[0018] Extract dynamic features from the environmental data based on the first time series dataset, including the time series trends of meteorological data and geological data, and the spectral characteristics of noise data and vibration data; extract progress deviation features from the construction progress data, consumption rate and utilization rate features from the resource allocation data, and quality deviation rate features from the project quality data.

[0019] Construct a feature vector based on the extracted features and fuse the data features of various types in time series.

[0020] Furthermore, the training process of the construction environment impact model also includes:

[0021] Input the comprehensive feature set into the construction environment impact model, perform importance analysis on the input various features, and dynamically assign weights to various features.

[0022] The comprehensive feature set is mapped to the feature spaces of different subtasks through a feature mapping layer, and the time series features of the comprehensive feature set are analyzed in the feature spaces of different subtasks to extract the trend information of construction progress, resource consumption, and project quality.

[0023] Based on the extracted trend information of construction progress, resource consumption, and project quality, output the time nodes for completing each construction task, the project quality grade, the future resource utilization rate, and the information on changes in material inventory.

[0024] Furthermore, the mapping of the comprehensive feature set to the feature spaces of different subtasks includes:

[0025] Mapping based on construction stage, construction method, and meteorological data features to the construction progress prediction task;

[0026] Mapping based on the dynamic features of geological data, vibration data, and noise data to the project quality prediction task;

[0027] Mapping based on construction cycle and construction resource features to the resource consumption prediction task.

[0028] Furthermore, the step S04 includes the following sub-steps:

[0029] S041. Determine the construction tasks according to the overall construction plan of the target construction project, divide the construction tasks into multiple subtasks, and associate the corresponding resource requirements, construction time nodes, and quality requirements for each subtask to form a task list;

[0030] S042. Quantify the risk value and delay probability of the task affected by the environment based on the first prediction result;

[0031] S043. Calculate the priority of each subtask based on the process logic, task importance, risk value, and delay probability of the construction task;

[0032] S044. Sort the subtasks according to the priority;

[0033] S045. Match the required resources for each subtask based on the construction resource data to form a resource allocation plan; in case of resource conflicts, dynamically adjust the resource allocation according to the priority and the dependency relationship between tasks;

[0034] S046. Integrate the sorted construction tasks and the resource allocation plan to generate a preliminary task plan; the preliminary task plan includes the start and end times of each subtask, the detailed resource allocation, and the key nodes during the construction process.

[0035] Furthermore, the construction process of the dynamic risk prediction model includes:

[0036] Obtain the construction project data of completed historical construction projects, where the construction project data includes building-related data, task plan solutions, and data on safety risk records;

[0037] Perform time series alignment on all data according to the time axis of the task plan solution to generate a second time series data set;

[0038] Construct a dynamic risk prediction model based on the combination of deep learning and time series analysis through a recurrent neural network;

[0039] Perform feature extraction operations based on the second time series data set, construct a comprehensive feature vector based on the feature extraction results, and use it as the training set, validation set, and test set to train the dynamic risk prediction model, and configure the structural parameters and hyperparameters of the model.

[0040] Further, the performing feature extraction operations based on the second time series data set includes:

[0041] Extract dynamic features from environmental data, including the time series trends of meteorological data and geological data, as well as the spectral characteristics of noise data and vibration data; extract progress deviation features from construction progress data, consumption rate and utilization rate features from resource allocation data, and quality deviation rate features from project quality data; extract resource matching rate and task completion deviation features from the task plan solution; extract risk frequency, spatial distribution features, and prevention and control measure effectiveness features from safety risk records.

[0042] The second aspect of the present invention discloses a building project intelligent decision-making and risk warning system, which is implemented based on the method disclosed in the first aspect. The system includes a data acquisition module, a construction module, a task determination module, and a task adjustment module;

[0043] The data acquisition module is used to collect building-related data of the building project, and the building-related data includes environmental data, construction progress data, resource allocation data, and project quality data;

[0044] The construction module is used to construct a construction environment impact model based on the collected historical building-related data;

[0045] After constructing the construction environment impact model, determine the construction characteristic parameters of the target building project, input the construction characteristic parameters and the environmental data of the target building project into the construction environment impact model, and predict the impact of environmental changes on the construction progress, project quality, and resource consumption of the target building project through the construction environment impact model as the first prediction result;

[0046] The task determination module is used to determine the construction tasks of the target building project, decompose the construction tasks of the target building project, and prioritize the construction tasks based on the first prediction result to generate a preliminary task plan solution;

[0047] The data acquisition module is also used to obtain the construction project data of historical construction projects;

[0048] The construction module is also used to construct a dynamic risk prediction model based on the construction project data;

[0049] After constructing the dynamic risk prediction model, the building-related data and the preliminary task plan of the target construction project are input into the dynamic risk prediction model, and the risk points during the construction process are predicted through the dynamic risk prediction model;

[0050] The task adjustment module is used to adjust the preliminary task plan based on the risk points and perform construction based on the adjusted task plan.

[0051] Compared with the prior art, the beneficial effects of the present invention are:

[0052] The present invention constructs a construction environment impact model based on building-related data, determines the preliminary task plan based on the impact results of the environmental changes predicted by the construction environment impact model on the construction project, and combines with the dynamic risk prediction model to warn and predict potential risks during the construction process, thereby providing scientific decision-making support for project managers, ensuring the smooth progress of construction tasks, and significantly reducing the risks of the project. BRIEF DESCRIPTION OF THE DRAWINGS

[0053] The drawings described herein are used to provide a further understanding of the embodiments of the present invention, and constitute a part of the economic application, but do not limit the embodiments of the present invention. In the drawings:

[0054] Figure 1 is a schematic flowchart of a method for intelligent decision-making and risk warning in construction projects disclosed in an embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0055] In order to enable those skilled in the art to better understand the solution of the present invention, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments.

[0056] Embodiment 1

[0057] The first aspect of the present invention discloses a method for intelligent decision-making and risk warning in construction projects. Please refer to Figure 1 , Figure 1 is a schematic flowchart of a method for intelligent decision-making and risk warning in construction projects disclosed in an embodiment of the present invention. The method includes the following steps:

[0058] S01. Collect construction-related data of a construction project; among them, the construction-related data includes, but is not limited to, environmental data, construction progress data, resource allocation data, and project quality data.

[0059] S02. Build a construction environment impact model based on the collected historical construction-related data;

[0060] S03. Determine the construction characteristic parameters of the target construction project, input the construction characteristic parameters and the environmental data of the target construction project into the construction environment impact model, and predict the impact of environmental changes on the construction progress, project quality, and resource consumption of the target construction project through the construction environment impact model as the first prediction result;

[0061] S04. Determine the construction tasks of the target construction project, decompose the construction tasks of the target construction project, and sort the construction tasks by priority based on the first prediction result to generate a preliminary task plan;

[0062] S05. Obtain the construction project data of historical construction projects, build a dynamic risk prediction model based on the construction project data, input the construction-related data of the target construction project and the preliminary task plan into the dynamic risk prediction model, and predict the risk points during the construction process through the dynamic risk prediction model;

[0063] S06. Adjust the preliminary task plan based on the risk points and perform construction based on the adjusted task plan.

[0064] In the embodiments of the present invention, the environmental data includes, but is not limited to, meteorological data, geological data, noise data, and vibration data. The construction progress data includes, but is not limited to, the completion status of construction processes. The resource allocation data includes, but is not limited to, the usage of construction machinery and equipment, the situation of staff, and material inventory. The project quality data includes, but is not limited to, construction quality inspection reports and unqualified product detection data.

[0065] Among them, the vibration data refers to the data information related to vibration monitored in a construction project, specifically, the vibration parameters in a building or the surrounding environment captured by vibration sensors (such as accelerometers or vibration monitors), including, but not limited to, vibration frequency, vibration amplitude, vibration waveform, and vibration direction. The sources of vibration data include the vibration of construction machinery: such as the vibration generated by drills, pile drivers, or other construction equipment during operation, geological vibration: the vibration of geological layers caused by construction or natural conditions (such as small earthquakes or uneven settlement of the foundation), and the vibration of the surrounding environment: such as the environmental vibration caused by traffic flow (vehicles or railways).

[0066] It is understandable that the reason for considering vibration data as environmental data in the embodiments of the present invention is that the structural stability of a building or a construction area can be evaluated through vibration data, whether there are dangerous resonances or excessive vibrations can be judged, whether there are faults in the equipment or maintenance is required can be determined by analyzing the vibration characteristics of mechanical equipment, and the impact of vibrations on the surrounding environment or other buildings can be monitored to avoid secondary hazards caused by construction.

[0067] Furthermore, the training process of the construction environment impact model includes:

[0068] Determine the construction project corresponding to the historical building-related data, determine the construction characteristic parameters of the historical construction project, extract the characteristics of the building-related data in units of construction projects and combine the construction characteristic parameters as a comprehensive feature set; wherein, the construction project corresponding to the historical building-related data is a completed construction project.

[0069] The construction characteristic parameters include the construction stage, construction method, construction period and construction resources of the construction project.

[0070] Furthermore, the feature extraction operation of the building-related data includes:

[0071] Perform data cleaning, normalization and missing value filling operations on the building-related data;

[0072] Align the environmental data, construction progress data, resource allocation data and project quality data in time series to generate a first time series dataset;

[0073] Extract dynamic features from the environmental data based on the first time series dataset, including the time series trends of meteorological data and geological data, and the spectral characteristics of noise data and vibration data; extract progress deviation features from the construction progress data, consumption rate and utilization rate features from the resource allocation data, and quality deviation rate features from the project quality data;

[0074] Construct feature vectors based on the extracted features and fuse the data features of various types in time series.

[0075] Furthermore, the training process of the construction environment impact model also includes:

[0076] Input the comprehensive feature set into the construction environment impact model, perform importance analysis on the input various features, and dynamically assign weights to various features;

[0077] Map the comprehensive feature set into the feature spaces of different subtasks through the feature mapping layer, analyze the time series features of the comprehensive feature set in the feature spaces of different subtasks, and extract the trend information of construction progress, resource consumption and project quality;

[0078] Output the time nodes for completing each construction task, the engineering quality grade, the future resource utilization rate, and the change information of material inventory based on the extracted trend information of construction progress, resource consumption, and engineering quality.

[0079] Furthermore, mapping the comprehensive feature set to the feature spaces of different subtasks includes:

[0080] Mapping based on construction stage, construction method, and meteorological data features to the construction progress prediction task;

[0081] Mapping based on the dynamic features of geological data, vibration data, and noise data to the engineering quality prediction task;

[0082] Mapping based on construction cycle and construction resource features to the resource consumption prediction task.

[0083] In the embodiment of the present invention, a construction environment impact model is constructed through an artificial neural network ANN. The construction environment impact model constructed using the ANN algorithm can make predictions in real time according to the input environmental data and construction progress, and dynamically adjust the model parameters according to the real-time feedback, so as to reflect the changes in the environment during the construction process, be able to respond to emergencies (such as bad weather, resource shortages, etc.) in a timely manner, and optimize the construction progress and resource allocation.

[0084] Based on the construction environment impact model, it is possible to automatically identify potential risks existing in the construction process (such as quality problems or schedule delays that may be caused by environmental changes), and generate an operable optimization plan. For example, when environmental factors change, the model can adjust task priorities, resource allocation, or adjust the construction plan to minimize risks to the greatest extent. Provide accurate progress and quality predictions for project staff, help them optimize task allocation, resource configuration, and construction processes, and improve the efficiency and accuracy of project management. The accurate prediction of risks by the model can also effectively reduce the emergency costs of on-site management and improve the refined management level of the construction process.

[0085] Furthermore, step S04 includes the following sub-steps:

[0086] S041. Determine the construction tasks according to the overall construction plan of the target construction project, divide the construction tasks into multiple subtasks, and associate the corresponding resource requirements, construction time nodes, and quality requirements with each subtask to form a task list;

[0087] S042. Quantify the risk value and delay probability of the task affected by the environment based on the first prediction result;

[0088] S043. Calculate the priority of each subtask based on the process logic, task importance, risk value, and delay probability of the construction task;

[0089] S044. Sort the subtasks according to the priority;

[0090] S045. Match the required resources for each subtask based on the construction resource data to form a resource allocation plan; in case of resource conflicts, dynamically adjust the resource allocation according to the priority and the dependency relationship between tasks;

[0091] S046. Integrate the sorted construction tasks and the resource allocation plan to generate a preliminary task plan; the preliminary task plan includes the start and end times of each subtask, the detailed resource allocation, and the key nodes during the construction process.

[0092] Preferably, the calculation of the task priority includes:

[0093]

[0094] where P task is the comprehensive priority of the task; R risk is the initial risk value of the task; R k is other risk factors related to the task (such as equipment failure risk, personnel health risk, etc.); ∝ k is the weight coefficient related to R k , which can all be set to 1, or can be set according to the actual situation, indicating the impact degree of various risks on the task; w1, w2, w3 are the weight parameters of the task priority, which can all be set to 1, or can be set according to the actual situation,; D delay is the delay probability of the task; I importance is the importance of the task; T remaining is the remaining time of the task; T total is the total time of the task; γ is the balance factor between time and resource allocation, which can be 0.8; C resource is the required resource quantity of the task; T required is the total required resource quantity of the task.

[0095] In the embodiment of the present invention, by sorting the construction tasks according to the prediction results of the construction environment impact model and subsequent resource allocation, it is ensured that each task can be reasonably arranged according to environmental changes, resource availability, and risk prediction, thereby improving the construction progress, project quality, and resource utilization rate. At the same time, by quantifying the risk value and delay probability of the task, the task plan can be flexibly adjusted, key tasks can be solved first, and the probability of delays and resource conflicts during the construction process can be reduced. In addition, the dynamic adjustment ensures that in case of limited resources or emergencies, resources can be reallocated according to the dependency relationship and priority between tasks, ensuring the smooth progress of the construction. By generating a preliminary task plan, the construction management becomes more transparent and controllable, improving the decision-making efficiency and the refinement degree of project management.

[0096] Furthermore, the construction process of the dynamic risk prediction model includes:

[0097] Obtain the construction project data of completed historical construction projects, where the construction project data includes building-related data, task plan schemes, and data on safety risk records;

[0098] Align all data according to the time axis of the task plan scheme to generate a second time series dataset;

[0099] Construct a dynamic risk prediction model based on the combination of deep learning and time series analysis through a recurrent neural network;

[0100] Perform feature extraction operations based on the second time series dataset, construct a comprehensive feature vector based on the feature extraction results, and use it as the training set, validation set, and test set to train the dynamic risk prediction model and configure the structural parameters and hyperparameters of the model.

[0101] Furthermore, the performing feature extraction operations based on the second time series dataset includes:

[0102] Extract dynamic features from environmental data, including the time series trends of meteorological data and geological data, as well as the spectral characteristics of noise data and vibration data; extract progress deviation features from construction progress data, consumption rate and utilization rate features from resource allocation data, and quality deviation rate features from project quality data; extract resource matching rate and task completion deviation features from the task plan scheme; extract risk frequency, spatial distribution features, and effectiveness features of prevention and control measures from safety risk records.

[0103] Specifically, the safety risk record includes the accident type, occurrence time, occurrence location, influence range, and accident cause analysis of historical safety events; the record of potential risk points of risk events, the results of hidden danger investigation, risk prevention and control measures, and their effectiveness evaluation; the marking of extreme weather and geologically unstable areas of environmental risks and their relevance to construction behaviors.

[0104] In the embodiments of the present invention, the risk points output by the dynamic risk prediction model include, but are not limited to, construction progress risks, project quality risks, resource consumption risks, safety risks, resource matching risks, and environmental risks.

[0105] Preferably, the dynamic risk prediction model includes:

[0106]

[0107] Among them, is the predicted risk value at time t; α iis the weighted coefficient of feature i in the first group of features. For example, they can all be set to 1, or the values can be set according to the actual situation, and it is used to measure the contribution of different features to the risk; f i (X t ) is the prediction function related to feature i category, and estimates the risk contribution based on features such as environmental data and construction progress; X t is the total feature data set; λ i is the attenuation factor. It can all be set to 0.16, or it can be set according to the actual situation, and it is used to represent the attenuation effect of each feature over time; β j is the weighted coefficient of feature j category in the second group of features; for example, they can all be set to 1, or other values can be set according to the actual situation. It is the feature function of the second group, which combines time t and feature data X t ; is the sine function that simulates periodic changes and is used to reflect the periodic behavior of some features (such as environmental fluctuations, equipment aging, etc.); θ j is the frequency parameter related to time; is the phase offset.

[0108] In the embodiment of the present invention, by setting up a dynamic risk prediction model, it helps to take preventive measures in advance, optimize the construction plan and resource allocation, thereby reducing the occurrence of accidental risks, improving the safety and quality control level of engineering construction, and finally minimizing the risks in the construction process and efficiently achieving the engineering goals.

[0109] Embodiment Two

[0110] The second aspect of the present invention discloses an intelligent decision-making and risk warning system for construction projects. The system includes a data acquisition module, a construction module, a task determination module, and a task adjustment module;

[0111] The data acquisition module is used to collect construction-related data of construction projects. The construction-related data includes environmental data, construction progress data, resource allocation data, and project quality data;

[0112] The construction module is used to construct a construction environment impact model based on the collected historical construction-related data;

[0113] After constructing the construction environment impact model, determine the construction characteristic parameters of the target construction project, input the construction characteristic parameters and the environmental data of the target construction project into the construction environment impact model, and predict the impact of environmental changes on the construction progress, project quality, and resource consumption of the target construction project through the construction environment impact model as the first prediction result;

[0114] The task determination module is used to determine the construction tasks of the target construction project, decompose the construction tasks of the target construction project, and prioritize the construction tasks based on the first prediction result to generate a preliminary task plan;

[0115] The data acquisition module is also used to obtain the construction project data of historical construction projects;

[0116] The construction module is also used to construct a dynamic risk prediction model based on the construction project data;

[0117] After constructing the dynamic risk prediction model, input the building-related data of the target construction project and the preliminary task plan into the dynamic risk prediction model, and predict the risk points during the construction process through the dynamic risk prediction model;

[0118] The task adjustment module is used to adjust the preliminary task plan based on the risk points and carry out construction based on the adjusted task plan.

[0119] It should be noted that the specific implementation process of the second embodiment is similar to that of the first embodiment and will not be elaborated in the second embodiment.

[0120] Finally, it should be noted that: the disclosed intelligent decision-making and risk warning method and system for construction projects in the embodiments of the present invention are only the preferred embodiments of the present invention, and are only used to illustrate the technical solutions of the present invention, rather than to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions recorded in the foregoing embodiments, or perform equivalent replacements for some of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. An intelligent decision-making and risk early warning method for construction projects, characterized in that, The method includes the following steps: S01. Collect construction-related data of the construction project, where the construction-related data includes environmental data, construction progress data, resource allocation data, and project quality data; S02. Construct a construction environment impact model based on the collected historical construction-related data; S03. Determine the construction characteristic parameters of the target construction project, input the construction characteristic parameters and the environmental data of the target construction project into the construction environment impact model, and predict the impacts of environmental changes on the construction progress, project quality, and resource consumption of the target construction project through the construction environment impact model as the first prediction result; S04. Determine the construction tasks of the target construction project, decompose the construction tasks of the target construction project, and perform priority ranking on the construction tasks based on the first prediction result to generate a preliminary task plan; S05. Obtain the construction project data of the historical construction project, construct a dynamic risk prediction model based on the construction project data, input the construction-related data and the preliminary task plan of the target construction project into the dynamic risk prediction model, and predict the risk points during the construction process through the dynamic risk prediction model; S06. Adjust the preliminary task plan based on the risk points, and perform construction based on the adjusted task plan.

2. The intelligent decision-making and risk warning method for construction projects according to claim 1, wherein The environmental data includes meteorological data, geological data, noise data, and vibration data; the construction progress data includes the completion status of construction processes; the resource allocation data includes the usage of construction machinery and equipment, the situation of staff, and material inventory; the project quality data includes construction quality inspection reports and non-conforming product detection data.

3. The intelligent decision-making and risk warning method for construction projects according to claim 1, characterized in that The training process of the construction environment impact model includes: Determine the construction project corresponding to the historical construction-related data, determine the construction characteristic parameters of the historical construction project, extract the characteristics of the construction-related data with the construction project as the unit and combine the construction characteristic parameters as the comprehensive feature set; the construction project corresponding to the historical construction-related data is a completed construction project; The construction characteristic parameters include the construction stage, construction method, construction period, and construction resources of the construction project.

4. The intelligent decision-making and risk warning method for construction projects according to claim 3, wherein The feature extraction operation of the construction-related data includes: Perform data cleaning, normalization, and missing value filling operations on the construction-related data; Align the environmental data, construction progress data, resource allocation data, and project quality data in time series to generate a first time series data set; Extract dynamic features from the environmental data based on the first time series data set, including the time series trends of meteorological data and geological data, and the spectral characteristics of noise data and vibration data; extract progress deviation features from the construction progress data, consumption rate and utilization rate features from the resource allocation data, and quality deviation rate features from the project quality data; Construct a feature vector based on the extracted features, and fuse the data features of each type in time series.

5. The intelligent decision-making and risk warning method for construction projects according to claim 4, characterized in that The training process of the construction environment impact model further includes: Input the comprehensive feature set into the construction environment impact model, perform importance analysis on the input various features, and dynamically assign weights to various features; The comprehensive feature set is mapped to the feature spaces of different subtasks through a feature mapping layer, and the time series features of the comprehensive feature set are analyzed in the feature spaces of different subtasks to extract the trend information of construction progress, resource consumption, and project quality; Based on the extracted trend information of construction progress, resource consumption, and project quality, output the time nodes for completing each construction task, the project quality grade, the future resource utilization rate, and the change information of material inventory.

6. The intelligent decision-making and risk early warning method for construction projects according to claim 5, characterized in that, The mapping of the comprehensive feature set to the feature spaces of different subtasks includes: Mapping based on construction stage, construction method, and meteorological data features to the construction progress prediction task; Mapping based on the dynamic features of geological data, vibration data, and noise data to the project quality prediction task; Mapping based on the construction cycle and construction resource features to the resource consumption prediction task.

7. The intelligent decision-making and risk early warning method for construction projects according to claim 1, characterized in that Step S04 includes the following sub-steps: S041. Determine the construction tasks according to the overall construction plan of the target construction project, divide the construction tasks into multiple subtasks, and associate corresponding resource requirements, construction time nodes, and quality requirements with each subtask to form a task list; S042. Quantify the risk value and delay probability of the task affected by the environment based on the first prediction result; S043. Calculate the priority of each subtask based on the process logic, task importance, risk value, and delay probability of the construction task; S044. Sort the subtasks according to the priority; S045. Match the required resources for each subtask based on the construction resource data to form a resource allocation plan; in case of resource conflicts, dynamically adjust the resource allocation according to the priority and the dependency relationship between tasks; S046. Integrate the sorted construction tasks and the resource allocation plan to generate a preliminary task plan; the preliminary task plan includes the start and end times of each subtask, the detailed resource allocation, and the key nodes during the construction process.

8. The intelligent decision-making and risk early warning method for construction projects according to any one of claims 1-7, characterized in that, The construction process of the dynamic risk prediction model includes: Obtain the construction project data of the completed historical construction projects, and the construction project data includes the data of building-related data, task plan, and safety risk records; Perform time series alignment on all data according to the time axis of the task plan to generate a second time series dataset; Construct a dynamic risk prediction model based on the combination of deep learning and time series analysis through a recurrent neural network; Perform feature extraction operations based on the second time series dataset, construct a comprehensive feature vector based on the feature extraction results, and use it as the training set, validation set, and test set to train the dynamic risk prediction model and configure the structure parameters and hyperparameters of the model.

9. The intelligent decision-making and risk early warning method for construction projects according to claim 8, characterized in that The performing feature extraction operations based on the second time series dataset includes: Extract dynamic features from the environmental data, including the time series trends of meteorological data and geological data, and the spectral characteristics of noise data and vibration data; extract progress deviation features from the construction progress data, consumption rate and utilization rate features from the resource allocation data, and quality deviation rate features from the project quality data; extract resource matching rate and task completion deviation features from the task plan; extract risk frequency, spatial distribution features, and the effectiveness features of prevention and control measures from the safety risk records.

10. An intelligent decision-making and risk early warning system for construction projects, implemented based on the method described in any one of claims 1-9, characterized in that, The system includes a data acquisition module, a construction module, a task determination module, and a task adjustment module; The data acquisition module is used to collect building-related data of a construction project, and the building-related data includes environmental data, construction progress data, resource allocation data, and project quality data; The construction module is used to construct a construction environment impact model based on the collected historical building-related data; After constructing the construction environment impact model, determine the construction characteristic parameters of the target construction project, input the construction characteristic parameters and the environmental data of the target construction project into the construction environment impact model, and predict the impact of environmental changes on the construction progress, project quality, and resource consumption of the target construction project through the construction environment impact model as the first prediction result; The task determination module is used to determine the construction tasks of the target construction project, decompose the construction tasks of the target construction project, and prioritize the construction tasks based on the first prediction result to generate a preliminary task plan; The data acquisition module is also used to obtain the building project data of historical construction projects; The construction module is also used to construct a dynamic risk prediction model based on the building project data; After constructing the dynamic risk prediction model, input the building-related data and the preliminary task plan of the target construction project into the dynamic risk prediction model, and predict the risk points during the construction process through the dynamic risk prediction model; The task adjustment module is used to adjust the preliminary task plan based on the risk points and perform construction based on the adjusted task plan.

Images