Infrastructure resource optimization method, system and device based on big data and medium

Abstract

The invention is suitable for the field of infrastructure resource optimization and matching, and discloses an infrastructure resource optimization method, system, device and medium based on big data, and the method comprises the steps: obtaining infrastructure project parameters, and constructing a layered space-time resource map based on a directed acyclic graph; obtaining a feasible region through an iterative multi-dimensional constraint negotiation and allocation algorithm based on the layered space-time resource atlas in combination with global boundary constraint; performing structured curing through the feasible region to obtain a data structure; based on the data structure, geometric analysis is carried out by adopting a BIM model to obtain an atomic job package and construct a microscopic network diagram; and generating an operation instruction by adopting optimal matching based on a dynamic utility function through combination of the microscopic network diagram and the real-time Internet of Things data flow, and forming closed-loop control based on execution feedback. According to resource supply and demand matching and global optimal configuration of manpower, machinery and materials, resource waste is reduced, and the green and low-carbon level of a large capital construction project is improved.

Description

Technical Field

[0001] This invention relates to the field of infrastructure resource optimization and matching, and in particular to a method, system, equipment and medium for infrastructure resource optimization based on big data. Background Technology

[0002] Modern power infrastructure is a key area supporting energy strategy, and its project management faces unique challenges unlike traditional construction projects. These projects are typically wide-ranging, long-term, and technically demanding, classifying them as "system-level projects." Their geographically dispersed nature not only creates significant logistical challenges but also forms a fragile interconnected system. Within this system, even localized resource misallocation or supply disruptions can trigger a chain reaction through the complex supply chain network, posing systemic risks to the entire project's schedule and costs. Furthermore, these projects involve numerous stakeholders, diverse disciplines, and multiple management systems, easily leading to "data silos." Managers lack access to comprehensive, real-time information, hindering cross-regional and cross-disciplinary global resource optimization decisions, and exacerbating resource waste and inefficiency.

[0003] Currently, due to the promotion of energy transition, the core direction of power infrastructure construction has also become green and low-carbon. In this context, resource optimization is not only an economic issue of cost control, but also a key to achieving sustainable development. Idle equipment, excess materials, and inefficient logistics are essentially unnecessary carbon emissions. However, the paradox is that the construction process of green energy facilities is itself particularly resource-intensive and energy-intensive. This creates a dilemma: to achieve long-term green goals, resource consumption and waste during the construction phase must be well controlled, otherwise it will greatly reduce the environmental benefits of the project throughout its entire life cycle.

[0004] Therefore, there is a need for an intelligent, holistic resource optimization technology that can break through the limitations of traditional management models, integrate fragmented data, and take into account multiple economic and environmental objectives in order to address the aforementioned complex challenges. Summary of the Invention

[0005] In view of the aforementioned existing problems, the present invention is proposed.

[0006] Therefore, this invention provides a big data-based infrastructure resource optimization method, system, equipment, and medium to address the problems of how existing methods, which prioritize, allocate, and adjust all tasks at the same level, are prone to exceeding the overall macro-constraints of the project due to local dynamic adjustments; fail to anticipate and systematically resolve all potential conflicts in advance; and have relatively limited adjustment methods; and how to transform resource allocation and reconfiguration schemes into precise, real-time, and closed-loop work instructions for on-site personnel and equipment.

[0007] To solve the above-mentioned technical problems, the present invention provides the following technical solution: In a first aspect, the present invention provides a method for optimizing infrastructure resources based on big data, comprising: Obtain infrastructure project parameters and construct a hierarchical spatiotemporal resource map based on a directed acyclic graph; Based on a hierarchical spatiotemporal resource map combined with global boundary constraints, the feasible region is obtained through an iterative multidimensional constraint negotiation and allocation algorithm. The data structure is obtained by structuring and solidifying the feasible domain; Based on the data structure, geometric analysis is performed using a BIM model to obtain atomic work packages and construct a micro-network diagram. By combining micro-network diagrams with real-time IoT data streams, job instructions are generated using optimal matching based on dynamic utility functions, and closed-loop control is formed based on execution feedback.

[0008] As a preferred embodiment of the big data-based infrastructure resource optimization method described in this invention, the method includes: acquiring infrastructure project parameters and constructing a hierarchical spatiotemporal resource map based on a directed acyclic graph, comprising: Obtain infrastructure project parameters and construct a hierarchical spatiotemporal resource map based on a directed acyclic graph, including macroscopic node sets and edge sets; The node set represents aggregated power grid units, and the edge set represents the physical connection relationships between power grid units. By using resource demand functions and multidimensional trade-off functions, quantified spatiotemporal and business attributes are added to node sets and edge sets.

[0009] As a preferred embodiment of the infrastructure resource optimization method based on big data described in this invention, the feasible region is obtained by combining a hierarchical spatiotemporal resource map with global boundary constraints and using an iterative multidimensional constraint negotiation and allocation algorithm, including: Independent dimensional boundary calculations are performed using a hierarchical spatiotemporal resource map to obtain the initial feasible interval for each macroscopic node; Based on the initial feasible interval and resource demand function of each node, multi-dimensional constraint conflict detection is performed to obtain the constraint conflict. An iterative negotiation and redistribution based on an ordered rule set, combined with global boundary constraints, is used to handle constraint conflicts and obtain the feasible region.

[0010] As a preferred embodiment of the infrastructure resource optimization method based on big data described in this invention, the data structure obtained by structuring and solidifying the feasible domain includes: a macro node constraint set, a key node identifier set, and a macro scheme negotiation log.

[0011] As a preferred embodiment of the infrastructure resource optimization method based on big data described in this invention, the method involves: using a BIM model for geometric analysis based on the data structure to obtain atomic work packages and construct a micro-network diagram, including: Geometric components are obtained by performing geometric analysis on macroscopic nodes using a BIM building information model; By traversing the geometric components, the corresponding work process is obtained, and each step in the work process is used as an AWP atomic work package and assigned initial attributes. Based on the initial properties of the atomic job package, a directed acyclic AWP network graph is constructed and a full path analysis is performed to obtain the cumulative result; The micro-network graph is obtained by verifying the conformity of the accumulated results with the macro-node constraint set through iterative optimization.

[0012] As a preferred embodiment of the infrastructure resource optimization method based on big data described in this invention, the method includes: generating work instructions by combining a micro-network diagram with real-time IoT data streams, using optimal matching based on dynamic utility functions, and forming closed-loop control based on execution feedback, including: The scheduling engine receives and parses real-time IoT data streams, synchronizes on-site status and updates AWP progress, and filters out available resource pools. A candidate task queue for AWP was obtained by filtering through micro-network graphs; Based on the key node identifier set, combined with the available resource pool and the AWP candidate task queue, the optimal matching scheme is obtained by using the optimal matching solution based on the dynamic utility function. The system generates and executes job instructions based on the optimal matching scheme, and generates execution feedback through real-time IoT data streams to achieve closed-loop control.

[0013] As a preferred embodiment of the infrastructure resource optimization method based on big data described in this invention, the iterative negotiation and reallocation based on ordered rule sets includes: For each constraint conflict, a rule set method is applied sequentially to resolve it. The rule set includes: energy storage load shifting, unit output adjustment, standby resource activation, and output reduction rulings.

[0014] Secondly, the present invention provides an infrastructure resource optimization system based on big data, comprising: The graph construction module is used to obtain infrastructure project parameters and construct a hierarchical spatiotemporal resource graph based on a directed acyclic graph. The constraint negotiation module is used to obtain the feasible region based on a hierarchical spatiotemporal resource map combined with global boundary constraints, through an iterative multidimensional constraint negotiation and allocation algorithm. The data structure is solidified through feasible domains to obtain the data structure. The decomposition and verification module is used to perform geometric analysis based on the data structure and BIM model to obtain atomic work packages and construct micro-network diagrams; The execution module is used to generate job instructions by combining a micro-network diagram with real-time IoT data streams, using optimal matching based on dynamic utility functions, and forming closed-loop control based on execution feedback.

[0015] Thirdly, the present invention provides an electronic device, comprising: Memory and processor; The memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions, which, when executed by the processor, implement the steps of a big data-based infrastructure resource optimization method.

[0016] Fourthly, the present invention provides a computer-readable storage medium storing computer-executable instructions that, when executed by a processor, implement the steps of the big data-based infrastructure resource optimization method.

[0017] Compared with existing technologies, the beneficial effects of this invention are as follows: By constructing a hierarchical spatiotemporal resource map and an iterative multidimensional constraint negotiation algorithm, the problem of local adjustments breaking through the overall constraints of the project in the flat management approach is avoided; potential resource conflicts are resolved through negotiation using an ordered set of rules, thereby reducing resource waste; by adopting a conflict detection and resolution mechanism, various resource constraint conflicts can be identified in advance and systematically resolved, improving the overall project execution efficiency; based on the real-time data stream of IoT devices and a minute-level scheduling engine, seamless connection from macro planning to micro execution can be achieved, generating work instructions accurate to the minute and workstation, forming closed-loop control, and improving the stability and controllability of plan execution. Attached Figure Description

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

[0019] Figure 1 This is a schematic diagram of the overall process of a big data-based infrastructure resource optimization method according to an embodiment of the present invention. Detailed Implementation

[0020] To make the above-mentioned objects, features, and advantages of the present invention more apparent and understandable, specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of them. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the protection scope of the present invention.

[0021] Example 1, referring to Figure 1 As an embodiment of the present invention, a method for optimizing infrastructure resources based on big data is provided, comprising: S100: Obtain infrastructure project parameters and construct a hierarchical spatiotemporal resource map based on a directed acyclic graph; S200: Based on a hierarchical spatiotemporal resource map combined with global boundary constraints, the feasible region is obtained through an iterative multidimensional constraint negotiation and allocation algorithm; S300: Data structure is obtained by structuring and solidifying the feasible domain; S400: Based on the data structure, the BIM model is used for geometric analysis to obtain atomic work packages and construct micro-network diagrams; S500: By combining a micro-network diagram with real-time IoT data streams, it generates work instructions using optimal matching based on dynamic utility functions and forms closed-loop control based on execution feedback.

[0022] It should be noted that large-scale infrastructure projects are characterized by their massive scale, diverse resource types, and complex spatiotemporal constraints. Throughout the project's entire lifecycle, the supply and demand relationships of multi-dimensional resources, including personnel, machinery, materials, and space, undergo significant dynamic changes, posing a severe challenge to the spatiotemporal allocation and scheduling optimization of resources. The space for feasible resource allocation schemes is severely compressed due to multiple factors such as conflicts in on-site work areas, overlapping equipment scheduling paths, and material delivery delays. Furthermore, due to the complexity of on-site work processes and the close connection between procedures, resource allocation schemes require continuous iterative adjustments. Therefore, comprehensive resource perception, intelligent collaborative scheduling, and closed-loop dynamic optimization are crucial for infrastructure projects.

[0023] Therefore, to address the aforementioned issues of supply and demand matching and global optimization configuration, a hierarchical spatiotemporal resource map is constructed through steps S100-S500. This map is combined with global boundary constraints and iterative multidimensional constraint negotiation and allocation algorithms to ensure that local optimization conforms to global objectives and can proactively identify and resolve potential resource conflicts throughout the project. By performing geometric analysis on the BIM model, atomic work packages are obtained and a micro-network diagram is constructed. Combined with real-time IoT data streams, a minute-level scheduling engine enables closed-loop control of real-time dynamic scheduling.

[0024] Example 2, refer to Figure 1 This is one embodiment of the present invention. Based on the above embodiment, a method for optimizing infrastructure resources based on big data is provided.

[0025] In this embodiment of the application, step S100, which involves obtaining infrastructure project parameters and constructing a hierarchical spatiotemporal resource map based on a directed acyclic graph, includes the following steps A1-A3: A1: Obtain infrastructure project parameters and construct a hierarchical spatiotemporal resource map based on a directed acyclic graph, including macroscopic node sets and edge sets; Specifically, large-scale infrastructure projects can be abstracted at the macro level as a directed acyclic graph. As a macro model of the project, it contains a macro node set. With edge set .

[0026] It should be noted that the graph serves as the data foundation for all subsequent planning and scheduling algorithms. Its structure and attributes support dynamic updates. By decomposing complex projects into manageable macro-units, the graph quantifies the interrelationships and their respective resource consumption characteristics.

[0027] A2: Define the macroscopic node set and edge set in the graph structure; Specifically, each node in the macroscopic node set This represents a aggregated power grid unit, such as a regional power grid, a virtual power plant, or a key hub substation. Each macro-node All are mapped to a detailed microscopic operation diagram. .

[0028] Each edge in the edge set Representative node and The relationship between them refers to the physical connection between power grid units, such as tie lines or transmission channels.

[0029] A3: By using resource demand functions and multidimensional trade-off functions, quantified spatiotemporal and business attributes are added to node sets and edge sets; Specifically, the attribute set for the macroscopic node set includes: net load function, multidimensional trade-off function, electrical attributes, and grid security importance.

[0030] The net load function describes the projected net load of the power grid unit in the future, which is the actual electricity load minus the projected output of uncontrollable renewable energy sources within the unit. The multidimensional trade-off function describes the dispatch cost of power resources, including the generation cost function. and carbon emission function Electrical attributes are upper and lower voltage limits; the importance of power grid security is taken as a business priority.

[0031] The attribute set for an edge set includes: minimum time delay, resource transfer time, and spatial transport distance.

[0032] Wherein, minimum time delay represents the preceding node. After completion, proceed to the next node. The shortest necessary time interval before commencement, such as the concrete curing period; resource transfer time refers to the time from the node to the equipment, especially large equipment. The work area was transferred to the node. The average time required for the operation area; the spatial transportation distance is the physical distance calculated based on the node spatial data.

[0033] It should be noted that the various attribute parameters defined in the additional definitions can be generated by analyzing historical project data, industry quota standards, or by using machine learning models for prediction. It can be integrated into a large L1-level infrastructure domain model, for example, based on a large high-wattage infrastructure model and fine-tuned using an infrastructure domain knowledge base. By inputting parameters such as project type, scale, and geological conditions, it can automatically generate or recommend resource demand functions and multi-dimensional trade-off functions for each node.

[0034] In an optional implementation, the construction of a hierarchical spatiotemporal resource graph based on a directed acyclic graph in step S100 can also add ontological semantic tags to the nodes and edges in the graph, so that the nodes can be automatically associated with all relevant standard process specifications or safety procedures documents, thereby realizing the fusion of knowledge graph and resource graph.

[0035] In another optional implementation, the construction of a hierarchical spatiotemporal resource map based on a directed acyclic graph in step S100 can also perform correlation queries and comparisons between the hierarchical spatiotemporal resource map of the current project and the maps of other projects under construction or already built, so as to quickly reuse successful resource allocation schemes of similar nodes and realize cross-project migration and optimization of experience knowledge.

[0036] In this embodiment of the application, step S200, based on a hierarchical spatiotemporal resource map combined with global boundary constraints, obtains the feasible region through an iterative multidimensional constraint negotiation and allocation algorithm, including the following steps B1-B4: B1: Calculate the independent dimensional boundaries using a hierarchical spatiotemporal resource map to obtain the initial feasible interval for each macroscopic node; Specifically, perform forward propagation: starting from the starting node of the graph, traverse all nodes in topological order, and calculate the earliest start time of the current node based on the completion value of the predecessor node and the minimum time delay in the edge set attributes. and the earliest completed Value, as well as minimum cumulative cost and carbon emission increment.

[0037] Perform backpropagation: Starting from the terminal node of the graph, calculate the latest start time of all nodes in the above dimensions in reverse topological order. and the latest to be completed Values ​​and maximum permissible costs and carbon emission increments.

[0038] Furthermore, the initial feasible interval for each node is generated by performing forward and backward propagation.

[0039] B2: Based on the initial feasible interval and resource demand function of each node, perform multi-dimensional constraint conflict detection to obtain constraint conflicts; Specifically, based on the initial feasible interval of each node and its resource requirement function Scan from the earliest start time to the latest finish time throughout the entire project cycle: at each time step, such as a day or an hour, accumulate the demand of all nodes that are active at that moment for the same limited or mutually exclusive resource, such as a specific model of crane or a team of highly skilled technicians.

[0040] Furthermore, identifying and recording all time points or time periods that do not meet constraints primarily involves proactively detecting, through power flow calculations, whether grid safety issues such as line power exceeding transmission capacity or node voltage exceeding safe ranges will occur within future scheduling cycles. At a certain moment... For a certain type of resource Total demand Exceeding the total supply of the resource At that time, a constraint conflict event is identified and recorded, including the start and end time of the conflict, the nodes involved, the resource type, and the number of gaps.

[0041] B3: Iterative negotiation and redistribution based on ordered rule sets, combined with global boundary constraints, are used to handle constraint conflicts and obtain the feasible region.

[0042] Specifically, for each conflict, an attempt is made to resolve it according to a pre-defined, ordered set of rules oriented towards optimal economic efficiency: Energy storage load time shift: First, try the most economical solution and check the available time fluctuation of conflicting nodes. It attempts to adjust its execution time within this floating range to avoid peak resource usage without incurring additional costs.

[0043] Unit output adjustment: If time shifting cannot resolve the issue or no float is available, the multidimensional trade-off function of the conflicting node or its immediate predecessor node is invoked. The calculations show that increasing costs can shorten the project time, thereby releasing resources in advance to resolve conflicts, and verifying whether the increased costs will exceed the global total budget limit.

[0044] Activating backup resources: If increasing costs to gain time is not economical or feasible, assess the cost of temporarily increasing resource supply, such as querying a market database to obtain the cost of renting a piece of equipment and calculating its total cost over the entire conflict period. Also, check whether the global total budget limit has been exceeded.

[0045] Output reduction ruling: When the above rules are ineffective or too costly, this will be used as a last resort. The business priority attribute values ​​of all conflicting participating nodes will be compared, and the low-priority nodes will be forced to delay execution to give up resources to the high-priority nodes. At the same time, the time constraints of the low-priority nodes and all their successor nodes will be compressed and updated accordingly.

[0046] Furthermore, each time a rule is applied and a conflict is successfully resolved, the impact of the change is immediately re-propagated locally, updating the feasible range of the relevant nodes to prepare for the next round of conflict detection.

[0047] Furthermore, steps B2 and B3 are repeated to form an iterative loop. The algorithm terminates when no further unresolved conflicts are found in a complete conflict detection round, or when the number of remaining conflicts converges to below a preset acceptable threshold, such as 0, yielding the feasible region. To ensure the algorithm's robustness, a maximum iteration limit is set. If unresolved conflicts remain after reaching the limit, they are marked and submitted for manual decision-making.

[0048] In an alternative implementation, instead of using an ordered rule set to resolve constraint conflicts in step S200, a reinforcement learning model can be used to dynamically adjust and optimize the calling order and parameters of the rule set based on indicators such as the success rate of historical negotiations and resource utilization.

[0049] In another optional implementation, the feasible domain in step S200 can also be presented to the planner through multi-dimensional data visualization, allowing them to manually adjust the priority of certain constraints. The system can recalculate and render the changes in the feasible domain in real time, realizing interactive planning through human-machine collaboration.

[0050] In this embodiment of the application, the data structure obtained by structuring and solidifying the feasible domain in step S300 includes: a macro-node constraint set, a key node identifier set, and a macro-solution negotiation log, including: Specifically, the data structure of the macro node constraint set is Key-Value, where the Key is the unique ID of the node and the Value is a sub-object containing all the final constraint parameters of that node.

[0051] The data structure of the key node identifier set is an array containing node IDs. The inclusion criteria are node IDs whose final feasible range margin (such as time fluctuation or cost margin) is less than a preset threshold.

[0052] The macro-level solution negotiation log has the following data structure: an array of structured entries sorted by time. Each entry records one instance of conflict detection and resolution action between B2 and B3.

[0053] In one optional implementation, the data structure obtained through structured solidification in step S300 can also support versioned management. Each key negotiation decision and constraint adjustment will generate a new version node, fully recording the decision context, changes and responsible persons, and achieving full lifecycle traceability.

[0054] In another optional implementation, the data structure obtained by the structured solidification in step S300 can also enable the system to automatically compare the differences between any two versions of the data structure, highlight the affected macro nodes, resource allocation and critical path, and automatically assess the impact of version changes on project schedule and cost.

[0055] In this embodiment of the application, step S400, based on the data structure, uses a BIM model for geometric analysis to obtain atomic work packages and construct a micro-network diagram, including the following steps C1-C3: C1: Geometric components are obtained by performing geometric analysis on macro nodes using BIM building information model; Specifically, the first step is to locate the BIM model portion corresponding to the current macro-node. Then, the 3D model of this portion is deconstructed into independent, constructible minimum geometric components.

[0056] C2: By traversing the geometric components, obtain the corresponding work process, and use each step in the work process as an AWP atomic work package and assign it initial attributes; Specifically, each extracted geometric component is traversed, and its type, material, specifications, and other attributes are used to query and match it from an extensible standard construction process library. Upon successful matching, the required standard operating procedure (SOP) for that component is retrieved, and each step in the SOP is instantiated as an atomic work package (AWP).

[0057] Each generated AWP is assigned initial attributes, including logical pre- and post-requirements, baseline duration, type and quantity of required resources, and estimated cost.

[0058] C3: Based on the initial properties of the atomic job package, construct a directed acyclic AWP network graph and perform full path analysis to obtain the cumulative result; Specifically, based on the logical precedence relationships assigned to all AWPs, a directed acyclic AWP network graph is constructed, which fully describes the micro-execution paths for completing all jobs within this macro-node.

[0059] A full path analysis was performed on the AWP network diagram, and the estimated total duration, total cost, total carbon emissions, and total resource requirements were cumulatively calculated to obtain the cumulative result. C4: The micro-network graph is obtained by verifying the conformity of the accumulated results with the macro-node constraint set through iterative optimization.

[0060] Specifically, if the accumulated result exceeds the constraint set of the macro-node, an optimization procedure will be initiated. Optimization methods include, but are not limited to, selecting alternative construction processes with lower costs or shorter durations from the process library for some AWPs, or adjusting the logical relationships of some AWPs, such as changing from "complete-start" to "start-start" to improve parallelism, while meeting technical specifications. This process is repeated until all indicators of the entire AWP network graph meet the constraint set of the macro-node, resulting in the micro-network graph.

[0061] In an optional implementation, the atomic work package obtained by geometric analysis of the BIM model in step S400 can not only rely on its geometric information, but also combine computer vision algorithms to analyze the BIM model, automatically identify key areas such as high-precision and complex assembly areas, and preferentially decompose them into more granular atomic work packages.

[0062] In another optional implementation, when assigning initial attributes to the atomic work package in step S400, the system can automatically match and recommend the required special types of work, safety risk warnings, recommended tool lists, etc. from the associated knowledge base or historical data, making the work package content more comprehensive and instructive.

[0063] In this embodiment of the application, step S500 uses a micro-network diagram combined with real-time IoT data streams to generate job instructions based on optimal matching of dynamic utility functions, and forms closed-loop control based on execution feedback, including the following steps D1-D4: D1: Receives and parses real-time IoT data streams through the scheduling engine, synchronizes on-site status and updates AWP progress, and filters to obtain available resource pools; Specifically, the scheduling engine continuously receives and parses real-time IoT data streams to achieve on-site status synchronization: updating the location, status (such as idle, working, faulty) and availability of every human, mechanical, and material resource in the system.

[0064] AWP Progress Update: Based on the on-site physical conditions fed back by AI cameras and sensors, the system automatically determines whether an ongoing AWP is complete. When an AWP is confirmed to be complete, its status is updated to "Completed".

[0065] Furthermore, all resources in an idle state are selected to form an available resource pool.

[0066] The real-time IoT data stream comes from the power grid data acquisition and monitoring control system and phasor measurement unit, including but not limited to electrical measurement data such as system frequency, node voltage, and actual power flow of the line.

[0067] D2: The AWP candidate task queue is obtained by filtering through the micro-network diagram; Specifically, AWPs that have completed all their prerequisite tasks are selected from the micro-network graph to form the current AWP candidate task queue.

[0068] D3: Based on the key node identifier set, combined with the available resource pool and the AWP candidate task queue, the optimal matching scheme is obtained by using the optimal matching solution based on the dynamic utility function. Specifically, a short-time domain optimization solver is started to perform optimal matching calculations for each available resource in the available resource pool and the executable subset of tasks in the AWP candidate task queue, so as to obtain the optimal matching scheme.

[0069] The objective utility function is dynamically adjusted based on task attributes: If the macroscopic node to which the candidate task belongs is in the critical node identifier set, the utility function takes "maximizing the guarantee that the task starts on time" as its primary objective. In this case, the solver will prioritize allocating resources to the task, even if this may result in a longer resource movement distance and slightly higher costs.

[0070] Otherwise, the utility function takes "minimizing the overall execution cost" as its primary objective, and the cost model can comprehensively consider multiple factors such as the distance the resource travels, estimated fuel consumption, waiting time, and the usage cost of different resources.

[0071] D4: Generate and execute job instructions based on the optimal matching scheme, and achieve closed-loop control by forming execution feedback through real-time IoT data stream.

[0072] Specifically, based on the optimal matching scheme, precise real-time work instructions are generated for each successfully matched resource-task pair. The instructions specify the personnel, equipment, the precise GIS coordinates of the workstation to be reached, the AWP number to be executed, and the content.

[0073] Furthermore, the execution status of instructions is captured by the system through the real-time IoT data stream of the next cycle, forming a micro-operational closed loop.

[0074] It should be noted that step S500 runs continuously at a frequency of minutes. This enables the system to respond to on-site IoT data streams in near real-time, quickly capturing changes in resource status and task progress, and promptly eliminating resource idleness and bottlenecks. Simultaneously, based on the latest on-site conditions, it can continuously and dynamically optimize the resource and task matching scheme with extremely high timeliness, ensuring that the scheme always closely matches the actual on-site situation and improving resource utilization efficiency.

[0075] Example 3 is an embodiment of the present invention. Based on the above embodiments, a project application of a big data-based infrastructure resource optimization method is provided to verify its feasibility and effectiveness. Applied to the N3 section of the XX UHVDC transmission project of the power grid: The N3 section is defined as three macro-level nodes: v1 basic engineering, v2 tower erection, and v3 overhead line construction, along with two dependent edges e12 and e23. Specific attribute values ​​are assigned to these elements. For node v1 basic engineering, Based on L1 model predictions, the peak demand for Sany Heavy Industry SY215C excavators will be 15 units per day during days 30-60 of the project, and the peak demand for C30 commercial concrete will be 800 cubic meters per day. The function is C(d) = 15000 × d² + 80000 × d. The carbon emission function is E(d) = 0.5 × d. Spatial data entry includes GIS polygon fence data for all 98 tower foundations in this section, with a business priority set to the highest level of 0.9. For edge e12, i.e., the connection from the foundation engineering to the tower erection, the minimum time delay is set to 7 days to meet the process requirements of concrete curing, the resource transfer time is set to 0.5 days, and the spatial transportation distance is calculated based on GIS data to be an average of 2.5 kilometers.

[0076] The initial time fluctuation for tower erection at node v2 was calculated to be 10 days. A conflict CF-001 was detected: on day 75 of the project, node v2 required 8 XCMG QY50K truck cranes, while the supply was only 6, resulting in a shortage of 2 cranes. To resolve this conflict, the energy storage load time-shifting rule was first applied to attempt to shift v2's work plan within the 10-day time fluctuation. However, calculations showed that even after shifting, the resource peak still existed, leaving a shortage of 1 crane, thus the energy storage load time-shifting rule partially failed. Next, the unit output adjustment rule was applied to the preceding node v1, calling its function C(d) = 15000 × d² + 80000 × d. Calculations showed that spending 180,000 yuan could shorten its construction period by 2 days, thus releasing 1 crane earlier for v2. This cost was within the total budget, and this solution was adopted. At this point, only one crane remained to resolve conflict CF-001. The backup resource activation rule was applied again, and rental market data showed that temporarily renting this crane model for 15 days would cost 60,000 yuan. This solution was also adopted. Ultimately, the combined solution of adjusting the unit output rule and activating the backup resource rule completely resolved conflict CF-001. After this solution was implemented, the total project cost and time constraints for relevant nodes were updated, and a new round of global conflict detection was performed. After three rounds of such iterations, confirming that there were no remaining conflicts, the algorithm converged and terminated.

[0077] For node v2 in the XX UHV power grid project, the generated constraint set includes node ID v2, node name "Tower Erection", final start time of 8:00 AM on February 18, 2026, final end time of 5:00 PM on April 30, 2026, final cost budget of 18.24 million yuan, and final carbon budget of 400 units. The final time fluctuation for the node v3 line erection project is only 2 days, which is less than the 3-day threshold. Therefore, the generated key node identifier set is [v3]. The log entry for resolving conflict CF-001 records the following: timestamp: 10:35:12 AM, September 2, 2025; conflict ID: CF-001; conflict type: resource constraint; resource name: XCMG QY50K truck crane; conflict nodes include v2 and other project nodes; resolution rules adopted: unit output adjustment rule and spare resource activation rule; impacts include an increase in cost of RMB 240,000; project schedule change: 0 days; affected node: v1; detailed description: node v1 cost increase of RMB 180,000, shortening the project schedule by 2 days; and v2 cost of renting one QY50K crane for RMB 60,000.

[0078] When assembling the ZJ-035 tower at node v2, all independent components such as main legs, crossarms, and bolts are extracted from the BIM model. For one main leg component of the ZJ-035 tower, the ground splicing process for a single main member of the angle steel tower is matched from the process library, and a series of AWPs are generated, including AWP-2-035-01 material inventory and AWP-2-035-02 main leg ground bolt connection. After constructing the AWP network diagram for node v2, its estimated total cost is calculated to be 18.5 million yuan. This result is compared with the final cost budget of 18.24 million yuan for v2, and it is found to exceed the budget by 260,000 yuan, at which point the compliance verification fails. To address the cost overrun of 260,000 yuan, the program is optimized to focus on the installation of connectors, which has a higher cost proportion. An alternative electroplated galvanized bolt construction process is found from the process library, which has lower material costs but slightly longer installation and inspection time. Applying the alternative process and re-performing the full path analysis, the new total cost was calculated to be 18.2 million yuan, and the total project duration increased by 8 hours. Since both the new cost and duration were within constraints, the verification was successful, and the optimization loop terminated.

[0079] In the scheduling loop example at 10:00 AM on March 15, 2026, data from the vehicle terminal was received, updating the status of crane-05 to idle and its location to tower base ZJ-040. Simultaneously, the AI ​​camera detected that the crossarm of tower base ZJ-041 had been hoisted into place, automatically updating the status of AWP-2-041-08 to completed. After the completion of AWP-2-041-08, its successor task AWP-2-041-09, tower head hoisting, entered the candidate task queue. Querying the macro node v2 to which AWP-2-041-09 belongs confirmed that it is not in the critical node set [v3]. Therefore, minimizing the overall cost was chosen as the objective utility function. After calculation, the solver found that assigning the nearest crane-05 to perform the task would have the lowest overall cost. An instruction was immediately generated and sent to the operator's PDA via the App, requesting that crane-05 be driven to the ZJ-041 foundation pit to perform the tower head hoisting operation AWP-2-041-09. In the next cycle at 10:05, the IoT data confirms that the crane-05 has reached the designated position and started working, and the status is updated to "working," completing the closed loop.

[0080] In summary, this invention constructs a hierarchical spatiotemporal resource map, combines it with global boundary constraints, and employs an iterative multidimensional constraint negotiation and allocation algorithm to ensure that global constraints such as overall project cost, schedule, and green low-carbon development are met while performing local and dynamic resource scheduling and optimization. Specifically, it uses ordered rule sets to handle constraint conflicts, proactively and systematically anticipating and resolving potential temporal and spatial conflicts of various resources throughout the project lifecycle, and providing conflict solutions to minimize resource waste. Furthermore, it uses a BIM model for atomized decomposition of tasks, obtaining a micro-network diagram, and integrates real-time IoT data to decompose macro-level resource planning schemes into real-time dynamic work instructions for frontline work units, achieving a closed loop through IoT data feedback. This effectively solves the problem of resource supply and demand matching and global optimization in large-scale infrastructure projects.

[0081] Example 4 illustrates a schematic scheme for a big data-based infrastructure resource optimization method. It should be noted that the technical solution of this big data-based infrastructure resource optimization system belongs to the same concept as the technical solution of the big data-based infrastructure resource optimization method described above. Details not described in detail in the technical solution of the big data-based infrastructure resource optimization system in this embodiment can be found in the description of the technical solution of the big data-based infrastructure resource optimization method described above.

[0082] This embodiment also provides an infrastructure resource optimization system based on big data, including: The graph construction module is used to obtain infrastructure project parameters and construct a hierarchical spatiotemporal resource graph based on a directed acyclic graph. The constraint negotiation module is used to obtain the feasible region based on a hierarchical spatiotemporal resource map combined with global boundary constraints, through an iterative multidimensional constraint negotiation and allocation algorithm. The data structure is solidified through feasible domains to obtain the data structure. The decomposition and verification module is used to perform geometric analysis based on the data structure and BIM model to obtain atomic work packages and construct micro-network diagrams; The execution module is used to generate job instructions by combining a micro-network diagram with real-time IoT data streams, using optimal matching based on dynamic utility functions, and forming closed-loop control based on execution feedback.

[0083] This embodiment also provides an electronic device suitable for infrastructure resource optimization based on big data, including: a memory and a processor; the memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions to implement the infrastructure resource optimization method based on big data as proposed in the above embodiment.

[0084] This embodiment also provides a storage medium on which a computer program is stored. When the program is executed by a processor, it implements the infrastructure resource optimization method based on big data as proposed in the above embodiments.

[0085] The storage medium proposed in this embodiment and the infrastructure resource optimization method based on big data proposed in the above embodiments belong to the same inventive concept. Technical details not described in detail in this embodiment can be found in the above embodiments, and this embodiment has the same beneficial effects as the above embodiments.

[0086] Based on the above description of the implementation methods, those skilled in the art can clearly understand that the present invention can be implemented using software and necessary general-purpose hardware, and of course, it can also be implemented using hardware. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as a computer floppy disk, read-only memory (ROM), random access memory (RAM), flash memory, hard disk, or optical disk, etc., including several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods of the various embodiments of the present invention.

[0087] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A method for optimizing infrastructure resources based on big data, characterized in that, include: Obtain infrastructure project parameters and construct a hierarchical spatiotemporal resource map based on a directed acyclic graph; Based on a hierarchical spatiotemporal resource map combined with global boundary constraints, the feasible region is obtained through an iterative multidimensional constraint negotiation and allocation algorithm. The data structure is obtained by structuring and solidifying the feasible domain; Based on the data structure, geometric analysis is performed using a BIM model to obtain atomic work packages and construct a micro-network diagram. By combining micro-network diagrams with real-time IoT data streams, job instructions are generated using optimal matching based on dynamic utility functions, and closed-loop control is formed based on execution feedback.

2. The infrastructure resource optimization method based on big data as described in claim 1, characterized in that, Obtain infrastructure project parameters and construct a hierarchical spatiotemporal resource map based on a directed acyclic graph, including: Obtain infrastructure project parameters and construct a hierarchical spatiotemporal resource map based on a directed acyclic graph, including macroscopic node sets and edge sets; The node set represents aggregated power grid units, and the edge set represents the physical connection relationships between power grid units. By using resource demand functions and multidimensional trade-off functions, quantified spatiotemporal and business attributes are added to node sets and edge sets.

3. The infrastructure resource optimization method based on big data as described in claim 2, characterized in that, Based on a hierarchical spatiotemporal resource map combined with global boundary constraints, an iterative multidimensional constraint negotiation and allocation algorithm is used to obtain the feasible region, which includes: Independent dimensional boundary calculations are performed using a hierarchical spatiotemporal resource map to obtain the initial feasible interval for each macroscopic node; Based on the initial feasible interval and resource demand function of each node, multi-dimensional constraint conflict detection is performed to obtain the constraint conflict. An iterative negotiation and redistribution based on an ordered rule set, combined with global boundary constraints, is used to handle constraint conflicts and obtain the feasible region.

4. The infrastructure resource optimization method based on big data as described in claim 3, further comprising: The data structure obtained by structuring and solidifying the feasible domain includes: macro node constraint set, key node identifier set, and macro solution negotiation log.

5. The infrastructure resource optimization method based on big data as described in claim 4, characterized in that, Based on the data structure, geometric analysis is performed using a BIM model to obtain atomic work packages and construct micro-network diagrams, including: Geometric components are obtained by performing geometric analysis on macroscopic nodes using a BIM building information model; By traversing the geometric components, the corresponding work process is obtained, and each step in the work process is used as an AWP atomic work package and assigned initial attributes. Based on the initial properties of the atomic job package, a directed acyclic AWP network graph is constructed and a full path analysis is performed to obtain the cumulative result; The micro-network graph is obtained by verifying the conformity of the accumulated results with the macro-node constraint set through iterative optimization.

6. The infrastructure resource optimization method based on big data as described in claim 5, characterized in that, By combining micro-network diagrams with real-time IoT data streams, optimal matching based on dynamic utility functions is used to generate work instructions, and closed-loop control is formed based on execution feedback, including: The scheduling engine receives and parses real-time IoT data streams, synchronizes on-site status and updates AWP progress, and filters out available resource pools. A candidate task queue for AWP was obtained by filtering through micro-network graphs; Based on the key node identifier set, combined with the available resource pool and the AWP candidate task queue, the optimal matching scheme is obtained by using the optimal matching solution based on the dynamic utility function. The system generates and executes job instructions based on the optimal matching scheme, and generates execution feedback through real-time IoT data streams to achieve closed-loop control.

7. The infrastructure resource optimization method based on big data as described in claim 2, characterized in that, The iterative negotiation and reallocation based on ordered rule sets includes: For each constraint conflict, a rule set method is applied sequentially to resolve it. The rule set includes: energy storage load shifting, unit output adjustment, standby resource activation, and output reduction rulings.

8. A big data-based infrastructure resource optimization system, employing the method described in any one of claims 1-7, characterized in that, include: The graph construction module is used to obtain infrastructure project parameters and construct a hierarchical spatiotemporal resource graph based on a directed acyclic graph. The constraint negotiation module is used to obtain the feasible region based on a hierarchical spatiotemporal resource map combined with global boundary constraints, through an iterative multidimensional constraint negotiation and allocation algorithm. The data structure is solidified through feasible domains to obtain the data structure. The decomposition and verification module is used to perform geometric analysis based on the data structure and BIM model to obtain atomic work packages and construct micro-network diagrams; The execution module is used to generate job instructions by combining a micro-network diagram with real-time IoT data streams, using optimal matching based on dynamic utility functions, and forming closed-loop control based on execution feedback.

9. An electronic device, characterized in that, include: Memory and processor; The memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions. When the computer-executable instructions are executed by the processor, they implement the steps of the infrastructure resource optimization method based on big data as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, It includes storing computer-executable instructions that, when executed by a processor, implement the steps of the big data-based infrastructure resource optimization method as described in any one of claims 1 to 7.

Images