A wharf steel structure construction progress dynamic management and control and resource allocation system

Abstract

The application relates to the field of engineering management and discloses a wharf steel structure construction progress dynamic management and control and resource allocation system, which comprises the following modules: a real-time data sensing and fusion module, which collects field data and initial plans, generates fusion data flow, a dynamic construction digital twin construction and evolution module, which updates a digital twin and a dynamic experience risk layer in real time based on the fusion data flow, a multi-agent adaptive decision engine, which simulates and deduces based on the digital twin, generates an adaptive construction scheme, and feeds back a risk signal, and a man-machine cooperation and closed-loop execution module, which visualizes the scheme for manual decision and issues instructions to the field. The application fuses real-time sensed physical field data and a digital model, constructs a dynamic construction digital twin which can be updated synchronously with the physical construction field state, and achieves the technical effect of keeping the digital information and the physical entity state highly consistent.

Description

Technical Field

[0001] The present invention relates to the technical field of engineering management, and in particular to a system for dynamically controlling and allocating construction progress of steel structures at a wharf. Background Art

[0002] The construction of large-scale steel structures for docks is a crucial component of infrastructure development. The process involves the hoisting of massive components, complex welding techniques, and precise installation, placing extremely high demands on construction progress, resource allocation, and operational safety. Against this backdrop, achieving precise control over the entire construction process to ensure on-time, high-quality project completion while effectively mitigating potential risks has become a critical issue urgently needed within the industry.

[0003] Currently, the industry has developed a variety of technologies and methods for managing steel structure construction at docks. For example, the use of project management software enables digital management of construction plans, task allocation, and basic progress tracking. Building Information Modeling (BIM) technology is widely used for pre-construction design optimization and collision detection, improving the feasibility of solutions. Furthermore, some construction sites have deployed sensor equipment to collect equipment operating data or environmental parameters, assisting management personnel in monitoring specific information. These technologies have, to a certain extent, optimized the organization and presentation of construction information, improved the efficiency of planning, and provided data support for the construction process.

[0004] However, existing technologies still exhibit significant shortcomings in addressing the dynamic and complex nature of wharf steel structure construction. First, current construction plans are mostly static. When unexpected conditions occur at the construction site, such as logistics delays for key components or sudden severe weather, adjustments to the original plan are often delayed and rely on manual experience and judgment, resulting in the inability to efficiently and immediately reconfigure resources, which in turn affects the overall construction progress. Second, although some information collection methods have been adopted, the ability to integrate and synchronize multi-source heterogeneous data in real time is still insufficient, making it difficult to form a comprehensive, real-time mapping of the physical construction site status. This makes it difficult for managers to obtain unified, high-quality, and timely information when making decisions, which restricts the accuracy and response speed of decisions. Finally, existing risk management focuses on post-analysis or static prevention based on experience, lacking the ability to quantitatively predict and proactively avoid dynamic and coupled risks during the construction process. When multiple risk factors exist simultaneously or interact, traditional methods cannot effectively identify their potential combined impacts, making it difficult for the construction safety assurance system to achieve forward-looking defense. To this end, those skilled in the art have proposed a dynamic control and resource allocation system for wharf steel structure construction progress to address the above problems. Summary of the Invention

[0005] In view of the deficiencies of the prior art, the present application provides a wharf steel structure construction progress dynamic management and control and resource allocation system, which solves the problems of static planning, information lag and insufficient risk predictability in the existing wharf steel structure construction management method.

[0006] To achieve the above object, the present application is implemented by the following technical solutions:

[0007] A wharf steel structure construction progress dynamic management and control and resource allocation system is provided, which comprises:

[0008] A real-time data perception and fusion module is used to collect real-time state data of a physical construction site and an initial construction plan containing three-dimensional geometric information and task logical relationship, and to process the collected data and obtained plan to generate a unified fusion data stream;

[0009] A dynamic construction digital twin construction and evolution module is used to construct and real-time update a dynamic construction digital twin synchronized with the physical construction site based on the fusion data stream, wherein the dynamic construction digital twin contains a dynamic experience risk layer evolved according to the experience of intelligent agents;

[0010] A multi-agent adaptive decision engine is used to simulate training and real-time deduction based on the dynamic construction digital twin to generate an adaptive construction scheme containing resource scheduling instructions, and to feed back risk signals for updating the dynamic experience risk layer to the dynamic construction digital twin construction and evolution module during the training process;

[0011] A man-machine cooperation and closed-loop execution module is used to visually present the adaptive construction scheme for manual decision-making, and to generate executable work instructions according to the decided scheme and issue them to the physical construction site.

[0012] Preferably, the specific working steps of the real-time data perception and fusion module include:

[0013] Dynamic physical information of construction equipment, materials and environment is collected through an Internet of Things sensor network deployed in the physical construction site;

[0014] The initial construction plan is accessed from a project management system;

[0015] The collected dynamic physical information and obtained initial construction log are data cleaned and standardized pretreated;

[0016] And the pretreated data is aligned and associated to generate the unified fusion data stream.

[0017] Preferably, the specific working steps of the dynamic construction digital twin construction and evolution module include:

[0018] initializing a static skeleton of the dynamic construction digital twin based on the initial construction plan;

[0019] continuously receiving the fused data stream to drive the state of the digital model in the dynamic construction digital twin to be synchronously updated with the physical construction site;

[0020] constructing and maintaining the dynamic empirical risk map layer, which is a four-dimensional tensor, for quantifying the potential risk of performing a construction task at a specific time and spatial coordinate;

[0021] and in response to receiving the risk signal fed back by the multi-agent adaptive decision engine, performing evolutionary update on the dynamic empirical risk map layer.

[0022] Preferably, the evolutionary update of the dynamic empirical risk map layer is implemented by the following formula:

[0023] ;

[0024] wherein, is the risk map layer after the n-th update, is the risk map layer at the n-th update, is the spatiotemporal coordinate of an event, is a preset risk map layer learning rate, is the observation information of the agent at time t, is the action performed by the agent at time t, is an event evaluation function. The specific working mode of the event evaluation function is as follows: When the action performed by the agent under the observation information results in a preset safety penalty or conflict penalty, the event evaluation function returns a preset positive risk value;

[0025] When the action performed by the agent under the observation information does not result in a preset safety penalty or conflict penalty, the event evaluation function returns zero.

[0026] Preferably, the specific working steps of the multi-agent adaptive decision engine include:

[0027] abstracting key resources in the construction process as a group of independent agents, and defining the observation space, action space and a composite reward function for each agent;

[0028] Preferably, the specific working steps of the multi-agent adaptive decision engine include:

[0029] abstracting key resources in the construction process as a group of independent agents, and defining the observation space, action space and a composite reward function for each agent;

[0030] ​​In the dynamic construction digital twin environment, the group of independent agents is centrally trained to learn to generate an optimal collaborative decision-making strategy;

[0031] And in actual operation, based on the learned optimal collaborative decision-making strategy and the real-time state of the dynamic construction digital twin, distributed decision-making is performed to generate the adaptive construction scheme.

[0032] The composite reward function is constructed in the following way:

[0033] By weighted sum of global progress reward, resource cost penalty, safety penalty and collaborative reward;

[0034] Wherein, the global progress reward is related to the completion state of the critical path task; the resource cost penalty is related to the idle time or energy consumption of unnecessary movement of resources; the safety penalty is related to whether the action of the agent enters a high-risk area or violates safety rules; the collaborative reward is related to the cooperation efficiency between the actions of multiple agents.

[0035] Preferably, the multi-agent adaptive decision engine also includes an adaptive task generation step when generating the adaptive construction scheme, which includes:

[0036] Monitoring the deviation metric between the actual state and the planned state of the dynamic construction digital twin; when the deviation metric exceeds a preset threshold, triggering a task generation mechanism;

[0037] The task generation mechanism explores and evaluates the expected return of the new task list formed by dynamically decomposing or merging the original construction task through forward reasoning;

[0038] And select the new task list with the maximum expected return as part of the adaptive construction scheme.

[0039] The deviation metric is determined by calculating the weighted norm distance between the actual state vector and the planned state vector of the dynamic construction digital twin.

[0040] Preferably, the multi-agent adaptive decision engine feeds back the specific way of updating the risk signal of the dynamic experience risk map layer during the training process, which includes:

[0041] In each simulation training step, monitor the safety penalty component and the collaborative reward component in the composite reward function, which are preset to represent negative results;

[0042] When it is determined that the value of the safety penalty component triggers a safety penalty condition, or the value of the collaborative reward component triggers a conflict penalty condition, it is determined that a negative risk event has occurred;

[0043] and generating, based on the negative risk event, the risk signal containing spatio-temporal coordinate information of the event.

[0044] Preferably, the specific working steps of the man-machine cooperation and closed-loop execution module include:

[0045] receiving the adaptive construction scheme and converting it into a visual form including a four-dimensional construction animation and a resource scheduling path diagram;

[0046] providing an interactive interface for the project manager to audit, fine-tune or approve the scheme in the visual form;

[0047] analyzing the approved scheme into explicit work instructions for specific construction resources;

[0048] and issuing the explicit work instructions through a mobile terminal or a field command system.

[0049] The present application provides a wharf steel structure construction progress dynamic management and resource allocation system. It has the following beneficial effects:

[0050] 1. The present application fuses real-time physical field data with digital models to construct a dynamic construction digital twin that can be updated synchronously with the physical construction site state, achieving a high degree of consistency between digital information and physical entity state. Compared with the existing technology that relies on static plans or preset scripts for simulation, the present application solves the inherent defects of serious disconnection between plans and reality and significant lag in information transmission.

[0051] 2. The present application proposes a unique risk management mechanism, the core of which is a dynamic experience risk layer generated by multi-agent through self-evolution based on massive simulation training, achieving quantitative and forward-looking identification of potential construction risks. It effectively addresses the shortcomings of existing technologies that rely on manual experience for post-mortem or static safety checks, solving the problem of being unable to predict sudden and complex risks caused by the coupling of multiple dynamic factors.

[0052] 3. The present application provides a dynamic planning method based on a multi-agent adaptive decision engine. This method has the ability to adaptively generate tasks and can actively decompose and recombine original tasks to generate new and better solutions when significant construction deviations are detected. This technical solution breaks through the framework of existing optimization algorithms that can only sort fixed task lists, solving the bottleneck of rigid strategies and the need for manual intervention for creative problem solving when facing major emergencies. BRIEF DESCRIPTION OF DRAWINGS

[0053] Figure 1A functional module diagram of the system of the present application;

[0054] Figure 2 A workflow diagram of the real-time data perception and fusion module of the present application;

[0055] Figure 3 A data structure and interaction diagram of the dynamic construction digital twin construction and evolution module of the present application;

[0056] Figure 4 A workflow diagram of the multi-agent adaptive decision engine of the present application;

[0057] Figure 5 A workflow diagram of the man-machine cooperation and closed-loop execution module of the present application.

[0058] Among them, 10, real-time data perception and fusion module; 20, dynamic construction digital twin construction and evolution module; 30, multi-agent adaptive decision engine; 40, man-machine cooperation and closed-loop execution module. DETAILED DESCRIPTION

[0059] The technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the drawings in the specification of the present application. Obviously, the described embodiments are only some of the embodiments of the present application, not all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor fall within the scope of protection of the present application.

[0060] Referring to the drawings Figure 1 , Figure 1 A functional module diagram of the wharf steel structure construction progress dynamic control and resource allocation system according to an embodiment of the present application. The system provided by the present application can include: a real-time data perception and fusion module 10, a dynamic construction digital twin construction and evolution module 20, a multi-agent adaptive decision engine 30, and a man-machine cooperation and closed-loop execution module 40.

[0061] The real-time data perception and fusion module 10 is used to collect real-time state data from the physical construction site and initial construction plans from the project management system. The real-time data perception and fusion module 10 performs preprocessing and standardization operations on the collected and obtained heterogeneous data, finally generates a unified, structured fusion data stream, and transmits it to the dynamic construction digital twin construction and evolution module 20.

[0062] A dynamic construction digital twin construction and evolution module 20, which is configured to receive the fused data stream from the real-time data perception and fusion module 10, and construct and maintain a dynamic construction digital twin based on the data stream. The dynamic construction digital twin is a high-fidelity digital replica of the physical construction site, and the state of the internal model is updated in real time as the state of the physical site changes. Another function of the dynamic construction digital twin construction and evolution module 20 is to construct and maintain a dynamic experience risk layer, which is used to quantify potential risks in the construction process, and its value will be evolved and updated according to the risk signal from the multi-agent adaptive decision engine 30. The dynamic construction digital twin construction and evolution module 20 provides the dynamic construction digital twin carrying the complete site state to the multi-agent adaptive decision engine 30 as the basis for its operation and deduction.

[0063] The multi-agent adaptive decision engine 30 is configured to perform large-scale simulation training and real-time forward deduction based on the dynamic construction digital twin provided by the dynamic construction digital twin construction and evolution module 20. Through these operations, the engine generates an adaptive construction scheme containing specific resource scheduling instructions. During the simulation training operation process, when a preset negative event occurs, the engine generates a risk signal and feeds it back to the dynamic construction digital twin construction and evolution module 20. The adaptive construction scheme generated by the engine is finally transmitted to the human-machine collaboration and closed-loop execution module 40.

[0064] The human-machine collaboration and closed-loop execution module 40 is configured to receive the adaptive construction scheme from the multi-agent adaptive decision engine 30. The human-machine collaboration and closed-loop execution module 40 converts the scheme data into a visual form for presentation to the project management personnel for final review and decision. After the scheme is approved, the human-machine collaboration and closed-loop execution module 40 parses it into explicit and executable work instructions, and issues these instructions to the corresponding execution units in the physical construction site.

[0065] The system provided by the present application constitutes a closed-loop control process in its overall workflow. The process begins with the continuous collection of the state of the physical construction site by the real-time data perception and fusion module 10. The generated fused data stream drives the dynamic construction digital twin construction and evolution module 20 to synchronize the state. The multi-agent adaptive decision engine 30 makes decisions based on the updated dynamic construction digital twin and generates an adaptive construction scheme. The scheme is manually reviewed and the instructions are issued through the human-machine collaboration and closed-loop execution module 40 to guide the work of the physical construction site. The state changes produced by the physical construction site after executing the new instructions will be captured again by the real-time data perception and fusion module 10, starting a new cycle of perception, modeling, decision-making and execution, and achieving rolling optimization and control of the entire construction process.

[0066] The internal structure and specific implementation of the real-time data perception and fusion module 10, the dynamic construction digital twin construction and evolution module 20, the multi-agent adaptive decision engine 30, and the man-machine cooperation and closed-loop execution module 40 in the embodiment of the present application will be described in detail below with reference to the accompanying drawings.

[0067] Referring to the accompanying drawings Figure 2 , Figure 2 is a workflow schematic diagram of the real-time data perception and fusion module 10 according to an embodiment of the present application. The specific implementation of the real-time data perception and fusion module 10 is as follows:

[0068] First, the real-time data perception and fusion module 10 performs collection of multi-source heterogeneous data. This step is performed from two dimensions of physical and digital. In the physical dimension, dynamic physical information is collected by deploying an Internet of Things sensor network on key entities in the construction site.

[0069] Specifically, a global positioning system (GPS) module and an inertial measurement unit (IMU) are installed on large construction equipment (such as a crawler crane, a transport vehicle) to obtain its three-dimensional spatial coordinates and attitude data (pitch, roll, yaw), respectively. Radio frequency identification (RFID) tags or two-dimensional codes are attached to the steel structure prefabricated components to be installed, and the logistics state information is collected through the deployment of readers at key path nodes (such as factory gates, construction site entrances, temporary storage yards). In the key operation area (such as the high-altitude hoisting area), meteorological sensors are installed to collect environmental parameters such as wind speed, wind direction, temperature, and humidity.

[0070] In the digital dimension, the real-time data perception and fusion module 10 accesses the project management system through a standard data interface to obtain the building information model (BIM) file containing all component three-dimensional geometric information, attributes, and spatial relationships, as well as the initial construction plan file defining the task list, logical dependency relationship, and estimated duration.

[0071] Second, the real-time data perception and fusion module 10 performs data preprocessing on the collected multi-source heterogeneous data. This step aims to eliminate data noise and unify data formats. For continuous signal data collected from sensors (such as GPS coordinates), Kalman filtering algorithm is used for processing to smooth the data and eliminate abnormal jump points. For all sources of data, standardized conversion is performed.

[0072] This conversion includes: converting all spatial position data (derived from GPS, BIM model) into a preset local coordinate system of the construction site; synchronizing the time of all devices through the network time protocol (NTP) and adding a uniform format timestamp to each data record; converting all physical quantity measurements (such as wind speed, temperature) into international standard units (SI).

[0073] Finally, the real-time data perception and fusion module 10 performs data fusion and distribution. This step aligns the pre-processed multiple independent data streams according to their unified timestamps, and integrates them into a structured fusion data stream. At any time point , the fusion data stream can be represented as a data set:

[0074] ;

[0075] where, is the state set of all devices at this moment, and the state information of each device in the set includes its three-dimensional coordinates, attitude parameters and a state flag bit representing its working state (such as idle, moving, working); is the state set of all key materials at this moment, and the state information of each material in the set includes its current position and a state flag bit representing its logistics state (such as in transit, arrived, installed); is the environmental parameter set at this moment, including wind speed, temperature and other numerical values.

[0076] The real-time data perception and fusion module 10 generates a structured fusion data stream and transmits it to the dynamic construction digital twin construction and evolution module 20 in real time and continuously through the message queue transmission protocol (such as MQTT), as the data source for updating the internal state of the dynamic construction digital twin.

[0077] Referring to the accompanying Figure 3 , Figure 3 is a data structure and interaction schematic diagram of the dynamic construction digital twin construction and evolution module 20 according to an embodiment of the present application. The specific implementation of the dynamic construction digital twin construction and evolution module 20 is as follows:

[0078] The dynamic construction digital twin construction and evolution module 20 first performs initialization of the dynamic construction digital twin (CDT). Based on the initial construction plan and BIM model file obtained from the real-time data perception and fusion module 10, this step constructs a static geometric model containing all components, devices and sites in the three-dimensional space, and loads the task list, resource library and logical constraint relationship between tasks, forming the static skeleton of the CDT.

[0079] After initialization, the dynamic construction digital twin construction and evolution module 20 enters the continuous state synchronization update phase. It continuously receives the fusion data stream The dynamic construction digital twin construction and evolution module 20 traverses each piece of state information in the data stream, retrieves the corresponding digital model in the CDT through the entity unique identifier (such as equipment number, component ID), and updates its state parameters (such as three-dimensional coordinates, pose, logistics state flag bit) to the latest value in the data stream. In this way, it is ensured that the state of the CDT is synchronized with the state of the physical construction site in the time dimension.

[0080] One of the core functions of the dynamic construction digital twin construction and evolution module 20 is to construct and maintain a dynamic empirical risk map layer. This layer is constructed as a four-dimensional tensor data structure, of which the three dimensions correspond to the spatial coordinates in the local coordinate system of the construction site , and the fourth dimension corresponds to time . The value of this tensor at any spatiotemporal coordinate point quantifies the potential risk of performing a construction task near this point. At system initialization, all values of this risk map layer are set to zero.

[0081] This dynamic empirical risk map layer is not static, but is evolved and updated according to the risk signals from the multi-agent adaptive decision engine 30. During system operation, when the multi-agent adaptive decision engine 30 determines that a negative risk event has occurred in its internal simulation training, it will generate and send a risk signal to this module. After receiving this signal, this module will trigger an update operation on the risk map layer. This evolutionary update is achieved through the following formula:

[0082] ;

[0083] wherein is the risk map layer after the th update, is the risk map layer at the th update, is the spatiotemporal coordinate of the event, is the preset risk map layer learning rate, is the observation information of the agent at time , is the action performed by the agent at time , is an event evaluation function.

[0084] The specific working mode is as follows: when it is determined that the agent performs action under observation information When the result triggers a preset safety penalty (such as entering a restricted area) or a conflict penalty (such as physical interference with other resources), the function returns a preset positive risk value; in all other normal cases where no penalty is triggered, the function returns zero.

[0085] Ultimately, the dynamic construction digital twin construction and evolution module 20 will provide the complete CDT status data of the updated equipment, materials, environmental status and evolved dynamic experience risk layer to the multi-agent adaptive decision engine 30 as the environmental basis for its subsequent simulation training and real-time decision-making.

[0086] Refer to the attached Figure 4 , Figure 4 FIG. 1 is a schematic diagram of the workflow of a multi-agent adaptive decision engine 30 according to an embodiment of the present invention. The specific implementation of the multi-agent adaptive decision engine 30 is as follows:

[0087] The engine first performs multi-agent modeling. This step abstracts the key dispatchable resources in the dock steel structure construction process, such as crawler cranes, flatbed trucks, and welding teams, into a group of independent agents. Let the agent set be For each agent , defining its three elements:

[0088] Observation Space: Agent In time Obtained local observation information , which includes the state of the agent itself (three-dimensional coordinates, posture, load), its local environment information (the state of other agents within a certain range, information on components to be processed), the risk value of its location and planned path on the dynamic experience risk layer, and information on its associated tasks in the initial construction plan.

[0089] Action Space: Agent A set of discrete actions that can be performed For example, for a crane modeled as an intelligent agent, its action space may include: standby, move to the target point, grab the component, rotate to the target angle, and release the component.

[0090] Composite reward function: used to evaluate the agent during training In time Execute an action Rewards received This function is composed of a weighted sum of multiple components to guide the agents to learn the desired collaborative behavior.

[0091] In a specific embodiment, the compound reward function can be expressed as:

[0092] ;

[0093] wherein, , , , is a preset weight coefficient for balancing different optimization objectives; is a global progress reward, which is a positive value when any agent completes a task on a critical path; is a cost penalty related to the agent , which is a negative value proportional to its idle time and energy consumption of unnecessary movement; is a safety penalty, which is a negative value when the agent 's action leads it to enter a high-risk area defined by the dynamic experience risk map, or violates preset safety rules; is a synergy reward, which is a positive value when the agent 's behavior forms efficient cooperation with other agents, and a negative value when its behavior conflicts or interferes with other agents.

[0094] The engine then performs offline centralized policy training for the group of agents in the dynamic construction digital twin environment. The goal of the training is to find a set of optimal policy network parameters that maximizes the cumulative expected return of all agents. The objective function is represented as:

[0095] ;

[0096] wherein, represents the joint policy of all agents determined by the parameters , is the expected operator, is the maximum time step of a simulation training, is a discount factor with a value range between 0 and 1, used to balance short-term and long-term rewards, is the reward function defined above at time is the reward value calculated by the agent .

[0097] During the training process, when it is determined that or the negative part is triggered, i.e., a negative risk event occurs, the engine generates a risk signal containing the spatiotemporal coordinate information of the event and feeds it back to the dynamic construction digital twin construction and evolution module 20.

[0098] In actual operation, the engine utilizes the trained policy to perform online real-time decision making. The engine continuously monitors the deviation metric between the actual state and the planned state of the dynamic construction digital twin When the deviation metric exceeds a pre-set threshold, indicating a significant construction deviation, an adaptive task generation mechanism is triggered. The mechanism evaluates the expected return of adhering to the original plan and executing a list of alternative new tasks by performing a high-speed forward inference in the dynamic construction digital twin, and selects the list of new tasks that maximizes the expected return as part of the adaptive construction plan. The adaptive construction plan contains specific scheduling instructions for all relevant resources in the future time period.

[0099] Finally, the multi-agent adaptive decision engine 30 transmits the generated adaptive construction plan containing resource scheduling instructions to the human-machine collaboration and closed-loop execution module 40.

[0100] Referring to the accompanying drawings Figure 5 , Figure 5 is a workflow schematic diagram of the human-machine collaboration and closed-loop execution module 40 according to an embodiment of the present application. The specific implementation of the human-machine collaboration and closed-loop execution module 40 is as follows:

[0101] The human-machine collaboration and closed-loop execution module 40 first receives the adaptive construction plan transmitted from the multi-agent adaptive decision engine 30. The plan is a data structure containing an updated task list, resource allocation in the future time period, and specific action sequences. After receiving the plan, the module performs plan visualization processing to convert it into one or more visual forms that are easy for humans to understand. For example, based on the action sequences and time information in the plan, a four-dimensional construction animation containing the time dimension is generated to dynamically show the subsequent construction process; on the three-dimensional model of the dynamic construction digital twin, the scheduling path and work points of each resource in the future time period are displayed in the form of highlighted paths or icons; at the same time, an updated construction Gantt chart is generated to clearly show the start and end times and logical relationships of tasks.

[0102] The human-machine collaboration and closed-loop execution module 40 then provides a human-machine interaction interface. The interface can be deployed on the display device of the project command center or the mobile terminal held by the project manager. The project manager can browse the aforementioned four-dimensional construction animation, scheduling path diagram, and Gantt chart through the interaction interface to conduct a comprehensive review of the plan generated by the multi-agent adaptive decision engine 30. The interface provides interactive functions to allow the manager to make a final decision on the plan, including direct approval or fine-tuning before approval. Fine-tuning operations can include small-range drag adjustments to the movement path of individual resources or changes to the execution order of non-critical tasks.

[0103] After the adaptive construction plan is approved (whether directly approved or approved after fine-tuning), the man-machine collaboration and closed-loop execution module 40 performs the instruction analysis and generation step. This step analyzes the abstract action instructions for agents in the plan (such as "move to target", "grab component") into specific and executable work instructions for specific construction resources and operating personnel. For example, a "move to target" instruction is analyzed into a navigation instruction containing specific target three-dimensional coordinates and a recommended path; a "grab component" instruction is analyzed into a hoisting task instruction containing a unique component identifier, hoisting weight, and target installation location.

[0104] Finally, the man-machine collaboration and closed-loop execution module 40 issues specific work instructions through a preset communication link. These instructions are sent to mobile terminal applications held by field operating personnel through a wireless network, or directly to the control units of semi-automated construction equipment with instruction receiving and execution capabilities. The execution results of these instructions in the physical construction site, and the resulting physical construction site state changes, are then collected by the real-time data perception and fusion module 10, thereby forming a complete, continuously running closed-loop management and control process for the system.

[0105] In order to further illustrate the collaborative working process of the technical solutions of the present application, a specific working scenario will be described below.

[0106] In a steel structure construction project at a wharf, a key prefabricated component A that was originally scheduled to be hoisted in the afternoon of the day was delayed in its arrival at the construction site due to a failure in the upstream transportation link. At the same time, real-time monitoring data from the weather sensor installed in the work area showed that the wind speed in the afternoon of the day was expected to exceed the safety threshold for hoisting operations. The system provided by the present application will perceive, decide, and provide a response plan for the above-mentioned sudden situation.

[0107] First, the real-time data perception and fusion module 10 continuously collects dynamic information from the construction site. In this scenario, its sensor network captures the following information:

[0108] The RFID reader did not detect the entry signal of component A at the scheduled arrival time, and received a delay notification from the logistics system.

[0109] The weather sensor continuously measures and predicts that the wind speed in the work area will continue to exceed the safety upper limit of 12 meters per second during the time period when component A is scheduled to be hoisted.

[0110] The real-time data perception and fusion module 10 pre-processes, normalizes and fuses the above information, generating a fused data stream containing the logistics state anomaly (delay) of component A, the environmental risk (high wind) of the hoisting area, and the state of all resources in the current site. This fused data stream is transmitted to the dynamic construction digital twin construction and evolution module 20.

[0111] Secondly, the dynamic construction digital twin construction and evolution module 20 receives the fused data stream from the real-time data perception and fusion module 10. The dynamic construction digital twin construction and evolution module 20 updates the state inside the dynamic construction digital twin according to the data stream.

[0112] Specifically, the state of component A in the digital twin is updated to "transportation delay", and its location in the site model remains "not in place". At the same time, due to the excessive wind predicted by the weather forecast, the multi-agent adaptive decision engine 30 does not generate a safety penalty signal, but will evaluate the potential risks in combination with the current environmental data in the simulation and deduction phase.

[0113] Then, the multi-agent adaptive decision engine 30 continuously monitors the deviation between the real-time state of the dynamic construction digital twin and the initial construction plan. When the delay of component A and the excessive predicted wind speed cause the system to identify a significant deviation between the current actual progress and the plan, the adaptive task generation mechanism of the multi-agent adaptive decision engine 30 is triggered.

[0114] The multi-agent adaptive decision engine 30 then performs high-speed forward deduction in the dynamic construction digital twin.

[0115] Firstly, it evaluates the expected return of performing the task according to the original plan (i.e. hoisting component A in the afternoon), and the deduction result shows that this scheme will result in high risk and low efficiency due to the high wind.

[0116] Subsequently, the engine explores and evaluates a new list of candidate tasks. For example, it may deduce the following alternative scheme:

[0117] Postpone the task of hoisting component A in the afternoon to the next day when the wind speed is expected to be safe.

[0118] Reschedule the idle hoisting resources and installation teams in the afternoon to perform the preparatory tasks with lower priority but not affected by wind in the original plan, such as steel material transportation in other areas, equipment maintenance or ground welding work.

[0119] Plan a temporary storage area and transportation task for component A after its arrival to avoid occupying space on the critical path.

[0120] The multi-agent adaptive decision engine 30 performs expected return calculation on all candidate task lists and finally selects the dynamic task list that can maximize the expected return as the optimal adaptive construction scheme. This scheme will contain a new task sequence, resource allocation and specific execution instructions. For example, it can contain the instruction "allocate No. 2 crane and No. 3 welder team to B area for ground pre-welding work". The adaptive construction scheme is transmitted to the man-machine collaborative and closed-loop execution module 40.

[0121] Finally, the man-machine collaborative and closed-loop execution module 40 receives the adaptive construction scheme from the multi-agent adaptive decision engine 30. This module converts the scheme into a visual form, for example, dynamically demonstrates the adjusted construction progress and resource scheduling path in a four-dimensional construction animation, and updates the Gantt chart to display the new task start and end time.

[0122] The project manager reviews these visual schemes through the interactive interface. After confirming the rationality of the scheme, the project manager approves the scheme.

[0123] After approval, the module parses the scheme into specific and executable work instructions. For example, the instruction "immediately go to B area to perform the pre-welding component transfer task" is issued to the mobile terminal of the No. 2 crane operator; the instruction "carry out ground component pre-welding work in B area" is issued to the No. 3 welder team. These instructions are issued to the corresponding execution units in the physical construction site.

[0124] The physical construction site performs work according to the new instructions. The real-time data generated during the work execution process, such as equipment position, component state, environmental parameters, etc. are collected again by the real-time data perception and fusion module 10, forming a continuous and rolling closed loop, ensuring the continuous and dynamic control of the system on the construction site.

[0125] Although embodiments of the present application have been shown and described, it is to be understood that various modifications, substitutions, replacements and changes can be made to these embodiments without departing from the principles and spirit of the present application, and the scope of the present application is defined by the appended claims and their equivalents.

Claims

1. A dynamic control and resource allocation system for the construction progress of a wharf steel structure, characterized in that: include: A real-time data perception and fusion module, which collects real-time status data of the physical construction site and the initial construction plan, and generates a unified fused data stream; A dynamic construction digital twin construction and evolution module, configured to construct and update in real time a dynamic construction digital twin synchronized with the physical construction site based on the fused data stream, wherein the dynamic construction digital twin includes a dynamic empirical risk layer that evolves based on the agent's experience; The specific working steps of the dynamic construction digital twin construction and evolution module include: Initializing a static skeleton of the dynamic construction digital twin based on the initial construction plan; Continuously receiving the fused data stream to drive the state of the digital model in the dynamic construction digital twin to be updated synchronously with the physical construction site; Constructing and maintaining the dynamic empirical risk layer, wherein the dynamic empirical risk layer is a four-dimensional tensor used to quantify the potential risk of performing a construction task at a specific time and space coordinate; In response to receiving a risk signal fed back by the multi-agent adaptive decision engine, evolving and updating the dynamic empirical risk layer; The evolution and update of the dynamic empirical risk layer is achieved through the following formula: ； in, For the After the updated risk layer, For the The risk layer at the time of the update, is the space-time coordinate of the event, is the default risk layer learning rate, For the agent at time Observation information, For the agent at time The action performed, Evaluate a function for an event; The event evaluation function The specific working method is: When the agent observes the information Execute the above action When a preset safety penalty or conflict penalty is caused, the event evaluation function returns a preset positive risk value; When the agent observes the information Execute the above action When no preset safety penalty or conflict penalty is caused, the event evaluation function returns zero; A multi-agent adaptive decision engine is used to perform simulation training and real-time deduction based on the dynamic construction digital twin, generate an adaptive construction plan including resource scheduling instructions, and provide feedback during the training process on risk signals used to update the dynamic empirical risk layer; The human-machine collaboration and closed-loop execution module is used to visualize the adaptive construction plan for manual review and decision-making, and generate executable work instructions based on the decided plan and send them to the physical construction site.

2. A dynamic control and resource allocation system for the construction progress of a wharf steel structure according to claim 1, characterized in that: The specific working steps of the real-time data perception and fusion module include: Collecting dynamic physical information of construction equipment, materials, and environment at the physical construction site through an Internet of Things sensor network; Access the project management system to obtain the initial construction plan including 3D geometry information and task logic relationships; Performing data cleaning and standardization preprocessing on the collected dynamic physical information and the obtained initial construction plan; The pre-processed data are aligned and associated to generate the unified fusion data stream.

3. A system for dynamic control and resource allocation of dock steel structure construction progress according to claim 1, characterized in that: The specific working steps of the multi-agent adaptive decision engine include: Abstract the key resources in the construction process into a group of independent agents, and define the observation space, action space and compound reward function for each agent; In the dynamic construction digital twin environment, the group of independent intelligent agents are centrally trained to learn and generate an optimal collaborative decision-making strategy; In actual operation, based on the learned optimal collaborative decision-making strategy and the real-time status of the dynamic construction digital twin, distributed decision-making is performed to generate the adaptive construction plan.

4. A system for dynamic control and resource allocation of dock steel structure construction progress according to claim 3, characterized in that: The composite reward function is constructed as follows: Determined by taking the weighted sum of global progress rewards, resource cost penalties, safety penalties, and coordination rewards; Among them, the global progress reward is related to the completion status of the critical path task; the resource cost penalty is related to the idle time of resources or the energy consumption of unnecessary movement; the safety penalty is related to whether the action of the intelligent agent enters a high-risk area or violates safety rules; the collaborative reward is related to the coordination efficiency between the actions of multiple intelligent agents.

5. A system for dynamic control and resource allocation of dock steel structure construction progress according to claim 3, characterized in that: The multi-agent adaptive decision engine further includes an adaptive task generation step when generating the adaptive construction plan, which step includes: monitoring a deviation metric between an actual state and a planned state of the dynamic construction digital twin; When the deviation metric exceeds a preset threshold, a task generation mechanism is triggered; The task generation mechanism explores and evaluates the expected returns of the new task list formed by dynamically decomposing or merging the original construction tasks through forward deduction; The new task list with the greatest expected reward is selected as part of the adaptive construction solution.

6. A dock steel structure construction progress dynamic control and resource allocation system according to claim 3, characterized in that: The specific manner in which the multi-agent adaptive decision engine feeds back risk signals for updating the dynamic empirical risk layer during the training process includes: In each simulation training step, monitoring the safety penalty component and the collaborative reward component preset in the composite reward function for representing negative results; When it is determined that the value of the safety penalty component triggers a safety penalty condition, or the value of the collaboration reward component triggers a conflict penalty condition, it is determined that a negative risk event has occurred; Based on the negative risk event, the risk signal including the spatiotemporal coordinate information of the event is generated.

7. A system for dynamic control and resource allocation of dock steel structure construction progress according to claim 1, characterized in that: The specific working steps of the human-machine collaboration and closed-loop execution module include: receiving the adaptive construction plan and converting it into a visual form including a four-dimensional construction animation and a resource scheduling path diagram; Providing an interactive interface for project managers to review, fine-tune or approve the visual plan; Parse approved plans into clear work instructions for specific construction resources; The clear work instructions are issued through a mobile terminal or an on-site command system.

Images