Multi-modal data-driven project progress risk intelligent assessment method and multi-modal data-driven project progress risk intelligent assessment system

Abstract

The invention relates to the technical field of project progress management, and provides a multi-modal data driven project progress risk intelligent assessment method and system. Mapping the multi-modal engineering information into a third-order tensor, and generating a tensor slice of the third-order tensor as a space-time tensor based on a preset projection function; constructing a time-varying graph formed by engineering tasks and execution relations based on engineering information, extracting task features from the space-time tensor, generating edge weights based on the task features, adjusting node relations in the time-varying graph through the edge weights, and outputting a dynamic graph structure; based on the space-time tensor and the dynamic graph structure, generating a risk equation of the project progress in the space-time dimension, solving the risk equation through a differential equation, and outputting a risk item of the project progress; and generating a project progress risk assessment file according to the risk item and the actual progress data in the project information. Risk assessment errors are reduced, project progress risk assessment efficiency is improved, and project decision is supported to be converted from passive response to active intervention.

Description

Technical Field

[0001] The present application relates to the technical field of engineering progress management, and in particular to a multimodal data-driven intelligent assessment method and system for engineering progress risk. Background Art

[0002] Modern engineering projects, such as large-scale infrastructure and industrial manufacturing, involve multidisciplinary collaboration, multi-phase tasks, and dynamic environmental changes. Traditional single-data source analysis cannot capture multidimensional risk factors. With the widespread adoption of technologies such as the Industrial Internet of Things (IIoT), sensor networks, and drone inspections, data collection capabilities for project sites have significantly improved. Multimodal data, such as video, sensor data, and text logs, provides a richer source of information for risk assessment, but requires intelligent methods for efficient integration and analysis.

[0003] Existing technologies, integrating multi-source data, require large-scale computations and high computing resource requirements. This requires optimizing model structures or adopting edge and distributed computing technologies to improve efficiency. Furthermore, the separate processing of structured and unstructured data makes it difficult to capture cross-modal correlations. Consequently, it is unable to dynamically capture project progress changes and risks, resulting in low efficiency in project progress risk assessment. Summary of the Invention

[0004] The present application provides a multimodal data-driven intelligent assessment method and system for engineering progress risk, which can at least to some extent solve the problem of low efficiency in engineering progress risk assessment.

[0005] Other features and advantages of the present application will become apparent from the following detailed description, or may be learned in part by practice of the present application.

[0006] According to one aspect of the present application, a multimodal data-driven intelligent assessment method for project progress risk is provided, comprising: acquiring multimodal project information, wherein the project information comprises structured data, unstructured data and time series data; mapping the project information into a third-order tensor, and generating tensor slices of the third-order tensor as a spatiotemporal tensor based on a preset projection function; wherein the third-order tensor comprises a project entity dimension, a time dimension and a feature dimension; constructing a time-varying graph consisting of project tasks and execution relationships based on the project information, extracting task features from the spatiotemporal tensor, generating edge weights based on the task features, adjusting the node relationships in the time-varying graph by the edge weights, and outputting a dynamic graph structure; generating a risk equation for the project progress in the spatiotemporal dimension based on the spatiotemporal tensor and the dynamic graph structure, and solving the risk equation by differential equations to output risk items for the project progress; and generating a project progress risk assessment file based on the risk items and the actual progress data in the project information.

[0007] In the present application, based on the above-mentioned scheme, the engineering information is mapped into a third-order tensor, and a tensor slice of the third-order tensor is generated based on a preset projection function as a spatiotemporal tensor, including: extracting the first information belonging to the entity dimension, time dimension and feature dimension in the engineering information, and mapping the first information into a third-order tensor; determining the attention weight of each modality based on a learnable weight matrix; and generating a tensor slice of the third-order tensor based on the attention weight of each modality and a preset projection function as a spatiotemporal tensor.

[0008] In this application, based on the above scheme, the attention weight of each modality is determined based on the learnable weight matrix, including: the original input feature of the mth modality Perform spatial dimension compression to generate compressed data; based on the compressed data and the learnable weight matrix of the mth mode , determine the attention weight corresponding to the mth modality for:

[0009]

[0010] in, represents the original input features of the mth modality, represents compressed data, k and m represent the identifiers of the modes, and M represents the total number of modes. Represents the learnable weight matrix of the k-th modality.

[0011] In the present application, based on the aforementioned scheme, the tensor slice of the third-order tensor is generated based on the attention weights of the modalities and the preset projection function as a spatiotemporal tensor, including: mapping the original input features of the mth modal in the engineering entity dimension n and the time dimension t to a unified dimension to generate dimensional data; based on the attention weights of the modalities and the dimensional data, a tensor slice of the third-order tensor is generated as a spatiotemporal tensor. :

[0012]

[0013] in, represents the original input features of the mth mode in engineering entity n and time window t, Represents dimensional data, Represents the activation function.

[0014] In the present application, based on the aforementioned scheme, a time-varying graph consisting of engineering tasks and execution relationships is constructed based on engineering information, task features are extracted from the spatiotemporal tensor, edge weights are generated based on the task features, node relationships in the time-varying graph are adjusted by the edge weights, and a dynamic graph structure is output, including: obtaining engineering tasks and execution relationships in the engineering information, simulating nodes, edges and edge attributes based on the engineering tasks and execution relationships to form a time-varying graph; extracting task features from the spatiotemporal tensor, generating edge weights based on the task features; adjusting node relationships in the time-varying graph by the edge weights, and outputting a dynamic graph structure.

[0015] In the present application, based on the aforementioned scheme, the risk equation of the project progress in the space-time dimension is generated based on the space-time tensor and the dynamic graph structure, and the risk equation is solved by a differential equation to output the risk item of the project progress, including: based on the dynamic graph structure, determining the diffusion item of the risk through the dynamic graph structure; based on the space-time tensor, determining the driving item of the risk caused by the change of multimodal features; based on the diffusion item and the driving item, determining the risk equation of the project progress risk in the space-time dimension; solving the risk equation by a differential equation to output the risk item of the project progress.

[0016] In the present application, based on the above solution, the determination of the diffusion item of risk through the dynamic graph structure based on the dynamic graph structure includes: determining the diffusion item of risk through the dynamic graph structure based on the risk probability of the task node in the dynamic graph structure. for:

[0017]

[0018] in, Represent the risk probabilities of task nodes i and j respectively, represents the edge weight between task node i and task node j, Represents the neighbor set of task node i.

[0019] According to one aspect of the present application, a multimodal data-driven intelligent project progress risk assessment system is provided, comprising:

[0020] an acquisition unit, configured to acquire multimodal engineering information, wherein the engineering information includes structured data, unstructured data, and time series data;

[0021] A tensor unit, configured to map the engineering information into a third-order tensor and generate tensor slices of the third-order tensor as a spatiotemporal tensor based on a preset projection function; wherein the third-order tensor includes an engineering entity dimension, a time dimension, and a feature dimension;

[0022] a structural unit for constructing a time-varying graph consisting of engineering tasks and execution relationships based on engineering information, extracting task features from the spatiotemporal tensor, generating edge weights based on the task features, adjusting node relationships in the time-varying graph using the edge weights, and outputting a dynamic graph structure;

[0023] A risk unit, configured to generate a risk equation for the project progress in the space-time dimension based on the space-time tensor and the dynamic graph structure, solve the risk equation by a differential equation, and output a risk term for the project progress;

[0024] The evaluation unit is used to generate a project progress risk evaluation file based on the risk items and actual progress data in the project information.

[0025] According to one aspect of the present application, a computer-readable medium is provided, on which a computer program is stored. When the computer program is executed by a processor, the multimodal data-driven intelligent assessment method for engineering progress risk as described in the above embodiments is implemented.

[0026] According to one aspect of the present application, an electronic device is provided, comprising: one or more processors; and a storage device for storing one or more programs, wherein when the one or more programs are executed by the one or more processors, the one or more processors implement the multimodal data-driven intelligent engineering progress risk assessment method as described in the above-mentioned embodiments.

[0027] According to one aspect of the present application, a computer program product or computer program is provided. The computer program product or computer program includes computer instructions stored in a computer-readable storage medium. A processor of a computer device reads the computer instructions from the computer-readable storage medium and executes the computer instructions, causing the computer device to perform the multimodal data-driven intelligent project progress risk assessment method provided in the various optional implementations described above.

[0028] In the technical solution of this application, multimodal data is uniformly encoded through third-order tensors, and the attention mechanism is used to dynamically weight and fuse structured, unstructured, and time-series features, thus breaking through the traditional data fragmentation problem. The dynamic graph structure adaptively updates edge weights based on spatiotemporal convolution, capturing changes in task dependencies and resource flows in real time and solving the lag of static models. The risk equation integrates graph diffusion terms (task network conduction), tensor gradient terms (multimodal feature drive), and nonlinear interaction terms (sudden event amplification effect) to achieve accurate modeling of the spatiotemporal evolution of risks. Compared with traditional methods, its dynamic fusion and mechanism-driven characteristics can reduce risk assessment errors, improve the efficiency of project progress risk assessment, and support engineering decision-making from passive response to active intervention.

[0029] It should be understood that the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the present application. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] The accompanying drawings are incorporated into and constitute a part of the specification, illustrate embodiments consistent with the present application, and together with the specification, are used to explain the principles of the present application. Obviously, the drawings described below are only some embodiments of the present application, and those skilled in the art can derive other drawings based on these drawings without inventive effort.

[0031] Figure 1 The flowchart of a multimodal data-driven intelligent assessment method for project progress risk in one embodiment of the present application is schematically shown.

[0032] Figure 2 The flowchart of generating a spatiotemporal tensor in one embodiment of the present application is schematically shown.

[0033] Figure 3 The following schematically illustrates a multimodal data-driven intelligent project progress risk assessment system in one embodiment of the present application.

[0034] Figure 4 A schematic diagram of the structure of a computer system suitable for implementing an electronic device according to an embodiment of the present application is shown. DETAILED DESCRIPTION

[0035] Example embodiments will now be described more fully with reference to the accompanying drawings. However, example embodiments can be implemented in many forms and should not be construed as limited to the examples set forth herein; rather, these embodiments are provided so that this application will be thorough and complete and will fully convey the concepts of the example embodiments to those skilled in the art.

[0036] In addition, described feature, structure or characteristic can be combined in one or more embodiments in any suitable manner.In the following description, many specific details are provided so as to provide a full understanding of the embodiments of the present application. However, it will be appreciated by those skilled in the art that the technical scheme of the present application can be put into practice without one or more of the specific details, or other methods, components, devices, steps etc. can be adopted. In other cases, known methods, devices, implementations or operations are not shown or described in detail to avoid blurring the various aspects of the application.

[0037] The block diagrams shown in the accompanying drawings are merely functional entities and do not necessarily correspond to physically separate entities. That is, these functional entities may be implemented in software, in one or more hardware modules or integrated circuits, or in different networks and / or processor devices and / or microcontroller devices.

[0038] The flowcharts shown in the accompanying drawings are for illustrative purposes only and do not necessarily include all contents and operations / steps, nor must they be executed in the order described. For example, some operations / steps may be decomposed, while others may be combined or partially combined. Therefore, the actual execution order may vary depending on the actual situation.

[0039] The implementation details of the technical solution of this application are described in detail below:

[0040] Figure 1 The flowchart of the multimodal data driven engineering progress risk intelligent assessment method according to one embodiment of the present application is shown. Figure 1 As shown, the multimodal data-driven intelligent assessment method for project progress risk includes at least steps S110 to S150, which are described in detail as follows:

[0041] S110, acquiring multimodal engineering information, where the engineering information includes structured data, unstructured data, and time series data.

[0042] In one embodiment of the present application, multimodal engineering information is acquired, including structured data, unstructured data, and time series data. Based on the acquired engineering industrial big data, an industrial cloud platform based on industrial cloud computing is constructed. Simultaneously, combined with the industrial Internet platform architecture, an industrial network-based engineering progress risk identification system is generated.

[0043] Specifically, in this embodiment, the structured data includes schedule information and resource allocation information. Schedule information includes tabular data such as the time arrangement and dependency relationships of engineering tasks, such as a Gantt chart. Resource allocation information includes quantitative data on the allocation of resources such as manpower, equipment, and materials, such as a matrix or database table.

[0044] Specifically, in this embodiment, the unstructured data includes on-site images and text reports. The on-site images include visual data of the construction scene, such as pictures or video frames captured by a camera. The text reports include natural language descriptions such as construction logs and inspection reports.

[0045] Specifically, in this embodiment, the time series data includes physical quantity data collected in real time and monitored by sensors, such as time series signals such as temperature, vibration, and displacement.

[0046] S120, mapping the engineering information into a third-order tensor, and generating tensor slices of the third-order tensor based on a preset projection function as a spatiotemporal tensor; wherein the third-order tensor includes an engineering entity dimension, a time dimension, and a feature dimension.

[0047] like Figure 2 As shown, in one embodiment of the present application, the engineering information is mapped into a third-order tensor, and a tensor slice of the third-order tensor is generated based on a preset projection function as a spatiotemporal tensor, including:

[0048] S210, extracting first information belonging to entity dimension, time dimension, and feature dimension from the engineering information, and mapping the first information into a third-order tensor;

[0049] S220, determining the attention weight of each modality based on the learnable weight matrix;

[0050] S230, based on the attention weights of the modalities and a preset projection function, generating tensor slices of the third-order tensor as a spatiotemporal tensor.

[0051] In one embodiment of the present application, multimodal data is mapped into a unified mathematical representation to form a third-order tensor, which includes an engineering entity dimension, a time dimension, and a feature dimension. First information belonging to the entity dimension, the time dimension, and the feature dimension is extracted from the engineering information and mapped into the third-order tensor.

[0052] Specifically, in the engineering entity dimension, tasks or process nodes are represented, such as concrete pouring or steel structure installation. The time dimension is divided into fixed intervals (e.g., days / weeks) to capture dynamic changes. The multimodal feature dimension integrates features from various data sources, such as progress features, image features, and text semantics. Through tensorization, heterogeneous data, such as tables, images, text, and time series, is encoded into a unified three-dimensional structure as a third-order tensor, addressing the difficulty of aligning multimodal data in traditional methods.

[0053] In one embodiment of the present application, the attention weight of each modality is determined based on a learnable weight matrix, including:

[0054] The original input features of the mth modality Perform spatial dimension compression to generate compressed data;

[0055] A learnable weight matrix based on the compressed data and the mth modality , determine the attention weight corresponding to the mth modality for:

[0056]

[0057] in, represents the original input features of the mth modality, represents compressed data, k and m represent the identifiers of the modalities, M represents the total number of modalities, and T represents the transpose operation of the learnable weight matrix. Represents the learnable weight matrix of the k-th modality.

[0058] The importance of each modality is automatically learned through the attention mechanism. For example, in the early stages of construction, the weight of structured data in the schedule planning modality is higher; in emergencies, the weight of time series data in the sensor modality is significantly increased.

[0059] In one embodiment of the present application, based on the attention weights of the modalities and a preset projection function, a tensor slice of the third-order tensor is generated as a spatiotemporal tensor, including:

[0060] Map the original input features of the mth mode in the engineering entity dimension n and time dimension t to a unified dimension to generate dimensional data;

[0061] Based on the attention weights of each modality and the dimensional data, a tensor slice of the third-order tensor is generated as a spatiotemporal tensor :

[0062]

[0063] in, represents the original input features of the mth mode in engineering entity n and time window t, Represents dimensional data, Represents the activation function.

[0064] In this process, structured data (such as schedules), unstructured data (such as on-site images and text reports), and time-series data (such as sensor monitoring) are mapped into third-order tensors. A projection function is used to map heterogeneous data into a unified space, resolving the mismatch between multimodal feature dimensions and semantics and achieving a unified representation of heterogeneous data. Activation functions are also used to enhance the model's ability to fit complex features, while attention weights are used to dynamically adjust modal contributions. For example, to capture the synergistic effects of schedule delays and resource shortages, the schedule modality is given a higher weight in the early stages of construction; during emergencies, the sensor data modality is given an increased weight. This unified encoding of multimodal data into a spatiotemporal tensor addresses the difficulties of multimodal data alignment and information fragmentation in traditional methods, significantly improving the accuracy and flexibility of data fusion and laying the foundation for subsequent dynamic graph construction and risk propagation modeling.

[0065] S130, constructing a time-varying graph consisting of engineering tasks and execution relationships based on engineering information, extracting task features from the spatiotemporal tensor, generating edge weights based on the task features, adjusting the node relationships in the time-varying graph through the edge weights, and outputting a dynamic graph structure.

[0066] In one embodiment of the present application, a time-varying graph consisting of engineering tasks and execution relationships is constructed based on engineering information, task features are extracted from the spatiotemporal tensor, edge weights are generated based on the task features, node relationships in the time-varying graph are adjusted using the edge weights, and a dynamic graph structure is output, including:

[0067] Acquire engineering tasks and execution relationships in the engineering information, and simulate nodes, edges, and edge attributes based on the engineering tasks and execution relationships to form a time-varying graph;

[0068] extracting task features from the spatiotemporal tensor and generating edge weights based on the task features;

[0069] The node relationships in the time-varying graph are adjusted using the edge weights to output a dynamic graph structure.

[0070] In one embodiment of the present application, engineering tasks and execution relationships are obtained from engineering information. The engineering task execution relationships describe the logical order and dependencies between tasks, such as Task A must be started after Task B is completed. These relationships are typically represented as directed edges, such as A→B. For example, foundation pouring in a construction project depends on foundation pit excavation, forming a dependency edge.

[0071] Furthermore, resource flow networks can be derived from engineering information to describe the transfer paths of resources (manpower, equipment, and materials) between different tasks. For example, a crane moves from task X to task Y, or a concrete pump truck schedules between tasks like pouring floor slabs and pouring beams and columns.

[0072] Based on this acquired information, a time-varying graph, or industrial knowledge graph, is constructed. This graph includes a node set, a time-dependent edge set, and edge attributes. Specifically, each node in the node set represents an engineering task, such as installing a pipeline. The time-dependent edge set includes two types of edges: dependency edges based on task logical relationships, and resource flow edges based on resource transfer paths. Edge attributes include information such as dependency strength and resource flow volume.

[0073] Task features are extracted from the spatiotemporal tensor and mapped to a unified space using a projection matrix. Cross-task similarities are calculated and normalized to weight values ​​to generate edge weights. Simultaneously, incremental changes in edge weights are predicted based on the current graph state and spatiotemporal features. The adjustment range is controlled by combining an evolution rate parameter. The edge weights are used to adjust node relationships in the time-varying graph, outputting a dynamic graph structure and enabling progressive updates of the graph structure.

[0074] This process outputs a dynamic graph structure, with each time window corresponding to an updated graph structure, dynamically reflecting changes in task dependencies and resource flows. Technically, the adaptive adjustment of edge weights accurately captures anomalies in project progress, such as resource conflicts that weaken the weight of flowing edges. It also integrates task logic and physical constraints to reveal the transmission paths of risks along high-weight edges, providing structural support for subsequent risk item modeling.

[0075] The above process constructs a time-varying graph based on the project task dependencies and resource flow network. Task feature similarity is calculated through spatiotemporal convolution, and edge weights are dynamically updated. For example, if a device failure causes a decrease in resource flow edge weights, the risk transmission path for dependent tasks will be adjusted accordingly. This dynamic graph structure can capture topological changes (such as new urgent tasks and resource conflicts) during project progress in real time, overcoming the limitations of traditional static graph models that cannot respond to real-time changes, making risk assessment more timely and scenario-adaptive.

[0076] S140 , generating a risk equation of the project progress in the space-time dimension based on the space-time tensor and the dynamic graph structure, solving the risk equation through a differential equation, and outputting a risk item of the project progress.

[0077] In one embodiment of the present application, based on the spatiotemporal tensor and the dynamic graph structure, a risk equation for the project progress in the spatiotemporal dimension is generated, and the risk equation is solved by a differential equation to output a risk term for the project progress, including:

[0078] Based on the dynamic graph structure, determining a diffusion item of risk through the dynamic graph structure;

[0079] Determining a driving factor of the multimodal feature change on risk based on the spatiotemporal tensor;

[0080] Determining a risk equation for project progress risk in time and space dimensions based on the diffusion term and the driving term;

[0081] The risk equation is solved by a differential equation to output a risk term for the project progress.

[0082] In one embodiment of the present application, determining, based on the dynamic graph structure, a risk diffusion item through the dynamic graph structure includes:

[0083] In one embodiment of the present application, based on the risk probability of the task nodes in the dynamic graph structure, the diffusion item of the risk through the dynamic graph structure is determined. for:

[0084]

[0085] in, Represent the risk probabilities of task nodes i and j respectively, represents the edge weight between task node i and task node j, Represents the neighbor set of task node i.

[0086] In this example, the diffusion term describes the degree of risk diffusion along task dependency and resource flow edges. For example, the risk of a critical path task is transmitted to adjacent tasks via high-weight edges. High risk differences between tasks are amplified by edge weights, such as when a delay in a critical path task triggers a chain reaction.

[0087] In one embodiment of the present application, based on the spatiotemporal tensor, the driving factor of the multimodal feature change on the risk is determined. for:

[0088]

[0089] in, Represents risk versus characteristics sensitivity, Representation characteristics The time derivative of The rate of change over time, D represents the total number of feature dimensions, and d represents the identity of the feature dimension.

[0090] In this embodiment, the driving effect of multimodal features (such as resource shortages and schedule deviations) on risk over time is quantified by driving items. Worsening over time ( ), such as a resource shortage indicator, and the risk is sensitive to this characteristic ( ), the risk accumulates rapidly.

[0091] In one embodiment of the present application, based on the diffusion term and the driving term, the risk equation for determining the project progress risk in the time and space dimensions is:

[0092]

[0093] in, Represents the preset hyperparameters that control the contribution of graph diffusion and tensor gradient respectively.

[0094] in, , represents a nonlinear interaction term. It is used to capture the amplification effect of sudden events through deep neural networks, such as the surge in risk caused by the superposition of severe weather and equipment failure.

[0095] Among them, W represents the learnable weight matrix, R represents the risk term, and T represents the feature term. When a high-risk task (high R value) is combined with a specific feature (such as a high temperature environment), ) can trigger a sharp increase in nonlinear risks. For example, abnormal concrete solidification in a high-temperature environment can lead to an exponential increase in schedule risk.

[0096] In one embodiment of the present application, the risk equation is solved by differential equations to output the risk term of the project progress. The risk equation is converted into a numerical integration problem by a neural differential equation solver, and the risk term is calculated by adaptive iteration. In the specific implementation, the current risk term of each time window t is Predicting rate of change through neural networks , and update it based on the historical status points .

[0097] The output risk term R is in matrix form, with each element represents the risk value of task n at time t. By integrating physical mechanisms with data-driven approaches, this approach not only explains the transmission path of risk along the task network, such as chain reactions triggered by changes in dependency edge weights, but also dynamically responds to the nonlinear effects of multimodal features, such as risk correction triggered by abnormal sensor signals. In practical applications, risk items can be visualized using heat maps, helping managers quickly identify high-risk tasks and optimize resource allocation. For example, when risk values ​​exceed a threshold, an alert can be automatically triggered, reducing the probability of project delays.

[0098] S150: Generate a project progress risk assessment file based on the risk items and actual progress data in the project information.

[0099] In one embodiment of the present application, the system analyzes risk items, identifies the risk value of each task, and categorizes it into three levels: low, medium, and high based on thresholds. For example, high-risk tasks are marked as progress delays exceeding 7 days or resource consumption exceeding the budget by 20%. The system then combines the edge weights in the dynamic graph to identify the source of risk transmission, such as a delay in a critical task triggering a chain reaction risk for subsequent tasks.

[0100] Translate the abstract risk values ​​output by the mathematical model into engineering language. For example, Task A (steel structure installation) currently has a high risk level, primarily due to equipment failure, which has delayed the project by 15%. This has impacted Task B (roofing construction) through dependent edges, and the recommendation is to prioritize the deployment of spare equipment and extend daily working hours. The report is also linked to actual progress data, such as Gantt chart deviations and resource inventory levels, to verify the credibility of the risk calculation.

[0101] Based on analysis of risk transmission pathways and nonlinear effects, targeted measures are proposed. For example, for high-risk tasks affected by resource flow, increased buffer inventory is recommended. For risks driven by multimodal characteristics (such as rainy season sensor data), adjustments to the construction plan or the implementation of contingency plans are recommended. The report presents recommendations in a prioritized manner, ensuring a rapid response from the engineering team. By integrating theoretical models with field data, complex mathematical results are transformed into actionable decision support, enabling managers to proactively intervene before risks escalate and improving overall project controllability.

[0102] This technical solution acquires multimodal project information, including structured data, unstructured data, and time series data; maps the project information into a third-order tensor, and generates tensor slices of the third-order tensor based on a preset projection function, as a spatiotemporal tensor; constructs a time-varying graph consisting of project tasks and execution relationships based on the project information, extracts task features from the spatiotemporal tensor, generates edge weights based on the task features, and adjusts node relationships in the time-varying graph using the edge weights to output a dynamic graph structure; generates a risk equation for project progress in the spatiotemporal dimension based on the spatiotemporal tensor and the dynamic graph structure, and solves the risk equation using differential equations to output risk items for the project progress; and generates a project progress risk assessment document based on the risk items and actual progress data in the project information. This process uniformly encodes multimodal data using a third-order tensor and utilizes an attention mechanism to dynamically weight and fuse structured, unstructured, and time series features, thus overcoming the traditional data fragmentation problem. The dynamic graph structure adaptively updates edge weights based on spatiotemporal convolution, capturing changes in task dependencies and resource flows in real time and addressing the lag inherent in static models. The risk equation integrates graph diffusion terms (task network conduction), tensor gradient terms (driven by multimodal features), and nonlinear interaction terms (the amplification effect of sudden events), combined with neural differential equation solving to accurately model the spatiotemporal evolution of risk. Compared to traditional methods, its dynamic fusion and mechanism-driven nature can reduce risk assessment errors, improve the efficiency of project schedule risk assessment, and support the shift from passive response to proactive intervention in project decision-making.

[0103] The following describes an apparatus embodiment of the present application, which can be used to implement the multimodal data-driven intelligent project progress risk assessment method described in the aforementioned embodiments of the present application. It is understood that the apparatus can be a computer program (including program code) running on a computer device, such as application software; the apparatus can be used to perform the corresponding steps of the method provided in the embodiments of the present application. For details not disclosed in the apparatus embodiments of the present application, please refer to the aforementioned embodiments of the multimodal data-driven intelligent project progress risk assessment method.

[0104] Figure 3 A block diagram of a multimodal data-driven engineering progress risk intelligent assessment system according to an embodiment of the present application is shown.

[0105] Reference Figure 3 As shown, a multimodal data-driven engineering progress risk intelligent assessment system according to one embodiment of the present application includes:

[0106] An acquisition unit 310 is configured to acquire multimodal engineering information, wherein the engineering information includes structured data, unstructured data, and time series data;

[0107] A tensor unit 320 is configured to map the engineering information into a third-order tensor and generate tensor slices of the third-order tensor as a spatiotemporal tensor based on a preset projection function; wherein the third-order tensor includes an engineering entity dimension, a time dimension, and a feature dimension;

[0108] A structure unit 330 is configured to construct a time-varying graph consisting of project tasks and execution relationships based on the project information, extract task features from the spatiotemporal tensor, generate edge weights based on the task features, adjust node relationships in the time-varying graph using the edge weights, and output a dynamic graph structure;

[0109] A risk unit 340 is configured to generate a risk equation for the project progress in the space-time dimension based on the space-time tensor and the dynamic graph structure, solve the risk equation by using a differential equation, and output a risk term for the project progress;

[0110] The evaluation unit 350 is configured to generate a project progress risk evaluation file based on the risk items and the actual progress data in the project information.

[0111] In the present application, based on the above-mentioned scheme, the engineering information is mapped into a third-order tensor, and a tensor slice of the third-order tensor is generated based on a preset projection function as a spatiotemporal tensor, including: extracting the first information belonging to the entity dimension, time dimension and feature dimension in the engineering information, and mapping the first information into a third-order tensor; determining the attention weight of each modality based on a learnable weight matrix; and generating a tensor slice of the third-order tensor based on the attention weight of each modality and a preset projection function as a spatiotemporal tensor.

[0112] In this application, based on the above scheme, the attention weight of each modality is determined based on the learnable weight matrix, including: the original input feature of the mth modality Perform spatial dimension compression to generate compressed data; based on the compressed data and the learnable weight matrix of the mth mode , determine the attention weight corresponding to the mth modality for:

[0113]

[0114] in, represents the original input features of the mth modality, represents compressed data, k and m represent the identifiers of the modes, and M represents the total number of modes.

[0115] In the present application, based on the aforementioned scheme, the tensor slice of the third-order tensor is generated based on the attention weights of the modalities and the preset projection function as a spatiotemporal tensor, including: mapping the original input features of the mth modal in the engineering entity dimension n and the time dimension t to a unified dimension to generate dimensional data; based on the attention weights of the modalities and the dimensional data, a tensor slice of the third-order tensor is generated as a spatiotemporal tensor. :

[0116]

[0117] in, represents the original input features of the mth mode in engineering entity n and time window t, Represents dimensional data, Represents the activation function.

[0118] In the present application, based on the aforementioned scheme, a time-varying graph consisting of engineering tasks and execution relationships is constructed based on engineering information, task features are extracted from the spatiotemporal tensor, edge weights are generated based on the task features, node relationships in the time-varying graph are adjusted by the edge weights, and a dynamic graph structure is output, including: obtaining engineering tasks and execution relationships in the engineering information, simulating nodes, edges and edge attributes based on the engineering tasks and execution relationships to form a time-varying graph; extracting task features from the spatiotemporal tensor, generating edge weights based on the task features; adjusting node relationships in the time-varying graph by the edge weights, and outputting a dynamic graph structure.

[0119] In the present application, based on the aforementioned scheme, the risk equation of the project progress in the space-time dimension is generated based on the space-time tensor and the dynamic graph structure, and the risk equation is solved by a differential equation to output the risk item of the project progress, including: based on the dynamic graph structure, determining the diffusion item of the risk through the dynamic graph structure; based on the space-time tensor, determining the driving item of the risk caused by the change of multimodal features; based on the diffusion item and the driving item, determining the risk equation of the project progress risk in the space-time dimension; solving the risk equation by a differential equation to output the risk item of the project progress.

[0120] In the present application, based on the above solution, the determination of the diffusion item of risk through the dynamic graph structure based on the dynamic graph structure includes: determining the diffusion item of risk through the dynamic graph structure based on the risk probability of the task node in the dynamic graph structure. for:

[0121]

[0122] in, Represent the risk probabilities of task nodes i and j respectively, represents the edge weight between task node i and task node j, Represents the neighbor set of task node i.

[0123] This technical solution acquires multimodal project information, including structured data, unstructured data, and time series data; maps the project information into a third-order tensor, and generates tensor slices of the third-order tensor based on a preset projection function, as a spatiotemporal tensor; constructs a time-varying graph consisting of project tasks and execution relationships based on the project information, extracts task features from the spatiotemporal tensor, generates edge weights based on the task features, and adjusts node relationships in the time-varying graph using the edge weights to output a dynamic graph structure; generates a risk equation for project progress in the spatiotemporal dimension based on the spatiotemporal tensor and the dynamic graph structure, and solves the risk equation using differential equations to output risk items for the project progress; and generates a project progress risk assessment document based on the risk items and actual progress data in the project information. This process uniformly encodes multimodal data using a third-order tensor and utilizes an attention mechanism to dynamically weight and fuse structured, unstructured, and time series features, thus overcoming the traditional data fragmentation problem. The dynamic graph structure adaptively updates edge weights based on spatiotemporal convolution, capturing changes in task dependencies and resource flows in real time and addressing the lag inherent in static models. The risk equation integrates graph diffusion terms (task network conduction), tensor gradient terms (driven by multimodal features), and nonlinear interaction terms (the amplification effect of sudden events), combined with neural differential equation solving to accurately model the spatiotemporal evolution of risk. Compared to traditional methods, its dynamic fusion and mechanism-driven nature can reduce risk assessment errors, improve the efficiency of project schedule risk assessment, and support the shift from passive response to proactive intervention in project decision-making.

[0124] Figure 4 A schematic diagram of the structure of a computer system suitable for implementing an electronic device according to an embodiment of the present application is shown.

[0125] It should be noted that the computer system of the electronic device in this embodiment is only an example and should not bring any limitation to the functions and scope of use of the embodiments of the present application.

[0126] In this embodiment, the computer system includes a central processing unit (CPU) 401, which can execute various appropriate actions and processes based on programs stored in a read-only memory (ROM) 402 or programs loaded from a storage unit 408 into a random access memory (RAM) 403, such as executing the multimodal data-driven intelligent project progress risk assessment method described in the above embodiment. RAM 403 also stores various programs and data required for system operation. CPU 401, ROM 402, and RAM 403 are interconnected via a bus 404. An input / output interface 405 is also connected to bus 404.

[0127] The following components are connected to the input / output interface 405: an input section 406 including a keyboard, mouse, and the like; an output section 407 including devices such as a cathode ray tube (CRT), a liquid crystal display (LCD), and speakers; a storage section 408 including devices such as a hard disk; and a communication section 409 including a network interface card such as a LAN (Local Area Network) card or a modem. The communication section 409 performs communication processing via a network such as the Internet. A drive 410 is also connected to the input / output interface 405 as needed. Removable media 411, such as a magnetic disk, an optical disk, a magneto-optical disk, or a semiconductor memory, is installed in the drive 410 as needed, so that computer programs read from the removable media can be installed in the storage section 408 as needed.

[0128] In particular, according to an embodiment of the present application, the process described above with reference to the flowchart can be implemented as a computer software program. For example, an embodiment of the present application includes a computer program product that includes a computer program carried on a computer-readable medium, the computer program including a computer program for executing the method shown in the flowchart. In such an embodiment, the computer program can be downloaded and installed from a network via the communication section 409 and / or installed from a removable medium 411. When the computer program is executed by the central processing unit 401, the various functions defined in the system of the present application are performed.

[0129] It should be noted that the computer-readable medium described in the embodiments of the present application may be a computer-readable signal medium or a computer-readable storage medium, or any combination thereof. A computer-readable storage medium may be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination thereof. More specific examples of computer-readable storage media may include, but are not limited to, an electrical connection having one or more conductors, a portable computer disk, a hard disk, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM), flash memory, optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination thereof. In the present application, a computer-readable storage medium may be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, apparatus, or device. In the present application, a computer-readable signal medium may include a data signal transmitted in baseband or as part of a carrier wave, which carries a computer-readable computer program. Such propagated data signals may take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination thereof. A computer-readable signal medium may also be any computer-readable medium other than a computer-readable storage medium that can transmit, propagate, or transport a program for use by or in connection with an instruction execution system, apparatus, or device. A computer program embodied on a computer-readable medium may be transmitted using any suitable medium, including but not limited to wireless, wired, or any suitable combination thereof.

[0130] The flowcharts and block diagrams in the accompanying drawings illustrate the possible implementation architecture, functions and operations of the systems, methods and computer program products according to various embodiments of the present application. Among them, each box in the flowchart or block diagram can represent a module, program segment, or part of the code, and the above-mentioned module, program segment, or part of the code contains one or more executable instructions for implementing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the box can also occur in an order different from that marked in the accompanying drawings. For example, two boxes represented in succession can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each box in the block diagram or flowchart, and the combination of boxes in the block diagram or flowchart, can be implemented with a dedicated hardware-based system that performs the specified function or operation, or can be implemented with a combination of dedicated hardware and computer instructions.

[0131] The units involved in the embodiments described in this application may be implemented by software or hardware, and the units described may also be set in a processor. In some cases, the names of these units do not constitute limitations on the units themselves.

[0132] According to one aspect of the present application, a computer program product or computer program is provided, the computer program product or computer program including computer instructions stored in a computer-readable storage medium. A processor of a computer device reads the computer instructions from the computer-readable storage medium and executes the computer instructions, causing the computer device to perform the methods provided in the various optional implementations described above.

[0133] As another aspect, the present application also provides a computer-readable medium, which may be included in the electronic device described in the above embodiments, or may exist independently without being incorporated into the electronic device. The computer-readable medium carries one or more programs, and when the one or more programs are executed by the electronic device, the electronic device implements the multimodal data-driven intelligent project progress risk assessment method described in the above embodiments.

[0134] It should be noted that, although several modules or units of the device for action execution are mentioned in the above detailed description, this division is not mandatory. In fact, according to the embodiment of the application, the features and functions of two or more modules or units described above can be concretized in one module or unit. On the contrary, the features and functions of one module or unit described above can be further divided into multiple modules or units to be concretized.

[0135] Through the description of the above embodiments, it is easy for those skilled in the art to understand that the example embodiments described herein can be implemented by software or by combining software with necessary hardware. Therefore, the technical solution according to the embodiments of the present application can be embodied in the form of a software product, which can be stored in a non-volatile storage medium (which can be a CD-ROM, a USB flash drive, a mobile hard disk, etc.) or on a network, and includes a number of instructions to enable a computing device (which can be a personal computer, a server, a touch terminal, or a network device, etc.) to execute the method according to the embodiments of the present application.

[0136] Those skilled in the art will readily conceive of other embodiments of the present application after considering the specification and practicing the embodiments disclosed herein. This application is intended to cover any variations, uses, or adaptations of the present application that follow the general principles of this application and include common knowledge or customary techniques in the art that are not disclosed herein.

[0137] It should be understood that the present application is not limited to the exact structures described above and shown in the drawings, and that various modifications and changes may be made without departing from the scope thereof. The scope of the present application is limited only by the appended claims.

Claims

1. A multimodal data-driven intelligent assessment method for project progress risk, characterized by: include: Acquire multimodal engineering information, the engineering information including structured data, unstructured data, and time series data; Mapping the engineering information into a third-order tensor, and generating tensor slices of the third-order tensor as a spatiotemporal tensor based on a preset projection function; wherein the third-order tensor includes an engineering entity dimension, a time dimension, and a feature dimension; Constructing a time-varying graph consisting of engineering tasks and execution relationships based on engineering information, extracting task features from the spatiotemporal tensor, generating edge weights based on the task features, adjusting node relationships in the time-varying graph using the edge weights, and outputting a dynamic graph structure; Based on the space-time tensor and the dynamic graph structure, a risk equation of the project progress in the space-time dimension is generated, and the risk equation is solved by a differential equation to output a risk term of the project progress; A project progress risk assessment document is generated based on the risk items and actual progress data in the project information.

2. The multimodal data-driven intelligent assessment method for project progress risk according to claim 1, characterized in that: Mapping the engineering information into a third-order tensor, and generating a tensor slice of the third-order tensor as a spatiotemporal tensor based on a preset projection function, including: Extracting first information belonging to entity dimension, time dimension, and feature dimension from the engineering information, and mapping the first information into a third-order tensor; Determine the attention weight of each modality based on the learnable weight matrix; Based on the attention weights of the modalities and a preset projection function, tensor slices of the third-order tensor are generated as spatiotemporal tensors.

3. The multimodal data-driven intelligent assessment method for project progress risk according to claim 2, characterized in that: Based on the learnable weight matrix, the attention weight of each modality is determined, including: The original input features of the mth modality Perform spatial dimension compression to generate compressed data; A learnable weight matrix based on the compressed data and the mth modality , determine the attention weight corresponding to the m-th modality.

4. The multimodal data-driven intelligent assessment method for project progress risk according to claim 3 is characterized in that: Based on the attention weights of the modalities and a preset projection function, a tensor slice of the third-order tensor is generated as a spatiotemporal tensor, including: Map the original input features of the mth mode in the engineering entity dimension n and time dimension t to a unified dimension to generate dimensional data; Based on the attention weights of the modalities and the dimensional data, tensor slices of the third-order tensor are generated as spatiotemporal tensors.

5. The multimodal data-driven intelligent assessment method for project progress risk according to claim 1, characterized in that: A time-varying graph consisting of engineering tasks and execution relationships is constructed based on engineering information, task features are extracted from the spatiotemporal tensor, edge weights are generated based on the task features, node relationships in the time-varying graph are adjusted using the edge weights, and a dynamic graph structure is output, including: Acquire engineering tasks and execution relationships in the engineering information, and simulate nodes, edges, and edge attributes based on the engineering tasks and execution relationships to form a time-varying graph; extracting task features from the spatiotemporal tensor and generating edge weights based on the task features; The node relationships in the time-varying graph are adjusted using the edge weights to output a dynamic graph structure.

6. The multimodal data-driven intelligent assessment method for project progress risk according to claim 1, characterized in that: Based on the space-time tensor and the dynamic graph structure, a risk equation of the project progress in the space-time dimension is generated, and the risk equation is solved by a differential equation to output the risk item of the project progress, including: Based on the dynamic graph structure, determining a diffusion item of risk through the dynamic graph structure; Determining a driving factor of the multimodal feature change on risk based on the spatiotemporal tensor; Determining a risk equation for project progress risk in time and space dimensions based on the diffusion term and the driving term; The risk equation is solved by a differential equation to output a risk term for the project progress.

7. The multimodal data-driven intelligent assessment method for project progress risk according to claim 6, characterized in that: Determining, based on the dynamic graph structure, a diffusion item of risk through the dynamic graph structure, including: Based on the risk probability of the task nodes in the dynamic graph structure, determine the diffusion item of the risk through the dynamic graph structure for: ； in, Represent the risk probabilities of task nodes i and j respectively, represents the edge weight between task node i and task node j, Represents the neighbor set of task node i.

8. A multimodal data-driven intelligent assessment system for project progress risk, characterized by: include: an acquisition unit, configured to acquire multimodal engineering information, wherein the engineering information includes structured data, unstructured data, and time series data; A tensor unit, configured to map the engineering information into a third-order tensor and generate tensor slices of the third-order tensor as a spatiotemporal tensor based on a preset projection function; wherein the third-order tensor includes an engineering entity dimension, a time dimension, and a feature dimension; a structural unit for constructing a time-varying graph consisting of engineering tasks and execution relationships based on engineering information, extracting task features from the spatiotemporal tensor, generating edge weights based on the task features, adjusting node relationships in the time-varying graph using the edge weights, and outputting a dynamic graph structure; A risk unit, configured to generate a risk equation for the project progress in the space-time dimension based on the space-time tensor and the dynamic graph structure, solve the risk equation by a differential equation, and output a risk term for the project progress; The evaluation unit is used to generate a project progress risk evaluation file based on the risk items and actual progress data in the project information.

9. The multimodal data-driven engineering progress risk intelligent assessment system according to claim 1, characterized in that: Mapping the engineering information into a third-order tensor, and generating a tensor slice of the third-order tensor as a spatiotemporal tensor based on a preset projection function, including: Extracting first information belonging to entity dimension, time dimension, and feature dimension from the engineering information, and mapping the first information into a third-order tensor; Determine the attention weight of each modality based on the learnable weight matrix; Based on the attention weights of the modalities and a preset projection function, tensor slices of the third-order tensor are generated as spatiotemporal tensors.

10. The multimodal data-driven engineering progress risk intelligent assessment system according to claim 9, characterized in that: Based on the learnable weight matrix, the attention weight of each modality is determined, including: The original input features of the mth modality Perform spatial dimension compression to generate compressed data; A learnable weight matrix based on the compressed data and the mth modality , determine the attention weight corresponding to the mth modality for: ； in, represents the original input features of the mth modality, represents compressed data, k and m represent the identifiers of the modes, and M represents the total number of modes. Represents the learnable weight matrix of the k-th modality.

Images