BIM-based highway engineering solid waste generation prediction and dispatching processing system

Abstract

The application relates to the technical field of highway engineering construction management, and discloses a highway engineering solid waste generation amount prediction and scheduling treatment system based on BIM, which comprises the following modules: a multi-source data access and BIM fusion module, which is used for analyzing a BIM model and a 4D construction plan; a construction process entropy dynamic calculation module, which is used for receiving activity instance data packets and is based on a preset entropy factor model; a material state migration network construction and evolution module, which is used according to the construction process entropy value; a solid waste form and resource portrait prediction module, which is used for carrying out probability simulation deduction on the evolved material state migration network; and an event-driven and adaptive scheduling decision module, which is used for listening to scheduling events triggered by the resource demand portrait or real-time field data. Through quantification of construction uncertainty and driving of a probability prediction model, a change from passive cleaning and removal to active pre-scheduling is realized, and the efficiency and precision of solid waste management are significantly improved.

Description

Technical Field

[0001] This invention relates to the field of highway engineering construction management technology, specifically a BIM-based system for predicting and scheduling the generation of solid waste in highway engineering projects. Background Technology

[0002] Highway construction is a complex process involving various materials such as earthwork, concrete, and steel reinforcement, inevitably generating a large amount of solid waste in various forms. Effective management, forecasting, and timely removal of this solid waste are crucial for ensuring a civilized construction site, reducing environmental pollution, and controlling project costs.

[0003] Existing technologies have attempted to utilize BIM and 4D construction planning to statically estimate the amount of construction solid waste generated through model-based quantity calculations and schedule linkage. However, these methods are highly dependent on the initial design model and construction plan, resulting in rigid predictions. This approach completely ignores the high degree of uncertainty prevalent in actual construction processes, such as changes in work procedures due to severe weather, sudden malfunctions of construction machinery, and operational errors by on-site workers. These dynamic factors significantly affect the actual conversion paths of materials and the final form and quantity of solid waste, leading to a large discrepancy between the predictions based on static plans and the actual on-site conditions.

[0004] In terms of solid waste collection and scheduling, traditional on-site management often relies on fixed collection cycles or passive triggering mechanisms based on on-site stockpile capacity sensors. For example, a collection request is only issued when the weighing or volume sensor readings in the stockpile reach a preset threshold. This response mode has a significant lag, often leading to excessive accumulation of solid waste on-site, causing secondary pollution or hindering construction, or resulting in ineffective waiting of scheduling resources, increasing management costs. Summary of the Invention

[0005] To address the shortcomings of existing technologies, this invention provides a BIM-based system for predicting and scheduling solid waste generation in highway engineering projects. This system solves the problems in existing highway engineering solid waste management, such as inaccurate prediction of solid waste generation, passive and delayed waste collection and scheduling, and inability to accurately match scheduling resources according to the form of solid waste, due to the neglect of the dynamic uncertainties of the construction process.

[0006] To achieve the above objectives, the present invention provides the following technical solution: a BIM-based system for predicting and scheduling the generation of solid waste in highway engineering projects, comprising:

[0007] The multi-source data access and BIM fusion module is used to parse BIM models and 4D construction plans, access multi-source heterogeneous data from construction machinery, site environment and solid waste dumps in real time, and perform spatiotemporal alignment and fusion to form standardized activity instance data packages.

[0008] The construction process entropy dynamic calculation module is connected to the multi-source data access and BIM fusion module. It is used to receive the activity instance data packet and, based on the preset entropy factor model, calculate the construction process entropy value, which represents the degree of uncertainty of the construction process, in real time.

[0009] The material state migration network construction and evolution module is connected to the construction process entropy dynamic calculation module, and is used to dynamically adjust the migration probability from the design state to different solid waste forms in the material state migration network according to the construction process entropy value.

[0010] The solid waste form and resource profile prediction module is connected to the material state migration network construction and evolution module. It is used to perform probability simulation and deduction on the evolved material state migration network, generate the probability distribution of solid waste form within the future time window, and generate a resource demand profile that represents the scheduling resource demand.

[0011] The event-driven and adaptive scheduling decision module is connected to the solid waste morphology and resource profile prediction module and the multi-source data access and BIM fusion module. It is used to listen for scheduling events triggered by the resource demand profile or real-time on-site data, and start the multi-objective optimization solver to generate adaptive scheduling instructions that include immediate removal and resource pre-deployment.

[0012] Preferably, the multi-source heterogeneous data in the multi-source data access and BIM fusion module includes at least:

[0013] Construction machinery status data, which are real-time working efficiency, energy consumption or vibration frequency collected from embedded sensors or vehicle terminals of construction machinery;

[0014] On-site environmental data, which are real-time temperature, humidity, rainfall or wind speed obtained through weather stations deployed at the construction site;

[0015] On-site sensing data refers to the real-time weight or volume of solid waste obtained through weighing sensors, lidar, or 3D scanning equipment deployed at temporary solid waste storage sites.

[0016] Preferably, the construction process entropy dynamic calculation module calculates the construction process entropy value through a weighted fusion model, wherein the construction process entropy value is a weighted sum of mechanical efficiency factor, environmental impact factor, personnel impact factor and process complexity factor.

[0017] Preferably, the material state migration network in the material state migration network construction and evolution module is a directed graph, and its nodes are used to represent material states;

[0018] The material status includes material type, life cycle stage, and physical form;

[0019] Among them, nodes in the life cycle stage of solid waste are further subdivided into different solid waste form nodes according to their physical form.

[0020] Preferably, the material state migration network construction and evolution module is specifically used for:

[0021] Based on the entropy value of the construction process, the migration tendency score from one material state to all possible next material states is calculated. When the next material state is an undesirable solid waste form, a higher construction process entropy value will increase its migration tendency score.

[0022] The migration propensity score is normalized using the Softmax function to obtain the migration probability.

[0023] Preferably, the solid waste morphology and resource profile prediction module is specifically used for:

[0024] On the evolved material state migration network, a Monte Carlo simulation is performed on the planned input materials to count the number of materials entering each solid waste form node and output the probability distribution of the solid waste form.

[0025] Preferably, the solid waste form and resource profile prediction module further includes a form-resource mapping knowledge base; the solid waste form and resource profile prediction module uses the form-resource mapping knowledge base to convert the probability distribution of the solid waste form into a probabilistic demand for specific scheduling resources, so as to generate the resource demand profile.

[0026] Preferably, the scheduling events in the event-driven and adaptive scheduling decision module include:

[0027] The predicted event is triggered when the expected demand value or probability of a certain scheduled resource in the resource demand profile exceeds a preset pre-scheduling threshold.

[0028] The on-site event is triggered when the on-site perception data provided by the multi-source data access and BIM fusion module exceeds a preset immediate removal threshold.

[0029] Preferably, the multi-objective optimization solver in the event-driven and adaptive scheduling decision module constructs a cost function with the weighted sum of total scheduling cost, solid waste on-site waiting time, and potential environmental risks as the optimization objective, and solves the scheduling scheme with the objective of minimizing this cost function.

[0030] Preferably, the multi-objective optimization solver satisfies the following constraints during the solution process:

[0031] Deterministic demand constraints require that the scheduling scheme must satisfy all deterministic cleanup tasks triggered by on-site events;

[0032] Probabilistic demand constraints require that the resources pre-deployed by the scheduling scheme must meet the probabilistic demands triggered by predicted events and included in the resource demand profile in a manner no less than the preset confidence level.

[0033] This invention provides a BIM-based system for predicting and scheduling solid waste generation in highway engineering projects. It offers the following advantages:

[0034] 1. This invention sets up a dynamic calculation module for construction process entropy to calculate the construction process entropy value, which represents the degree of uncertainty in the construction process, in real time. This entropy value is then used to dynamically drive the material state migration network construction and evolution module to adjust the migration probability. This makes the prediction model for solid waste generation no longer a rigid extrapolation based on a static plan, but a dynamic adaptive model that can reflect uncertainties such as on-site mechanical efficiency and environmental changes in real time, thereby significantly improving the accuracy and timeliness of the prediction results.

[0035] 2. This invention, through a solid waste morphology and resource profiling prediction module, transforms the probabilistic prediction of future solid waste morphology into a resource demand profile for specific scheduling resources. The event-driven and adaptive scheduling decision module can monitor predicted events triggered by this profile and, in conjunction with a multi-objective optimization solver, generate adaptive scheduling instructions that include resource pre-deployment. This advances resource reserves and deployment, significantly reducing on-site waiting time for solid waste and improving scheduling efficiency.

[0036] 3. This invention refines solid waste to specific physical form nodes within a material state migration network, and uses a form-resource mapping knowledge base to accurately map the prediction results of different forms of solid waste to the demand for different types of scheduling resources. This refined management granularity ensures that the scheduling instructions generated by the scheduling decision module are highly targeted, matching the correct treatment equipment and transportation tools for solid waste of different physical forms, avoiding resource mismatch, and significantly improving the professionalism and efficiency of solid waste treatment. Attached Figure Description

[0037] Figure 1 This is a system framework diagram of the present invention;

[0038] Figure 2 This is a schematic diagram of the data flow of the multi-source data access and BIM fusion module of the present invention;

[0039] Figure 3 This is a flowchart of the entropy calculation process for the construction process of the present invention;

[0040] Figure 4 This is a schematic diagram of the material state migration network structure of the present invention. Detailed Implementation

[0041] The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0042] Please see the appendix Figure 1 - Appendix Figure 4 This invention provides a BIM-based system for predicting and scheduling the generation of solid waste in highway engineering projects, including...

[0043] The multi-source data access and BIM fusion module is used to parse BIM models and 4D construction plans, access multi-source heterogeneous data such as construction machinery, site environment and solid waste dump in real time, and perform spatiotemporal alignment and fusion to form standardized activity instance data packages.

[0044] Specifically, in this embodiment, the multi-source data access and BIM fusion module integrates static engineering design and planning data with dynamic construction process and site status data in a structured manner, providing a unified and standardized input for the subsequent prediction and scheduling module.

[0045] The functions of the source data access and BIM integration module are implemented through the following processing units:

[0046] The data structured processing unit for BIM and 4D construction planning parses static engineering design and planning data. First, it parses highway engineering BIM models conforming to Industry Basic Class (IFC) standards or other formats, extracting the complete set of components required for the project. Each component Each element contains globally unique identifiers (GUIDs), material properties, geometric information, and design parameters such as quantities. Simultaneously, this unit parses construction schedule files output by project management software (such as Primavera P6 or MS Project) to obtain the construction activity set. Every construction activity All include the event number and the planned start time. and the planned end time .

[0047] The data structuring processing unit builds a component set by reading preset component-activity relationships (e.g., linked in BIM software or defined through an external mapping table). With construction activities Spatiotemporal correlation mapping between This mapping clarifies any construction activity. The specific subset of BIM components associated with and consumed within its planned execution timeframe. At this point, a 4D construction foundation model incorporating the time dimension has been completed.

[0048] The real-time access and spatiotemporal alignment unit for multi-source heterogeneous data during the construction process is responsible for the dynamic data acquisition and binding. Multi-source heterogeneous data explicitly includes, but is not limited to, the following types:

[0049] Construction machinery status data, such as real-time work efficiency, energy consumption, vibration frequency, and GPS location collected from embedded sensors or on-board terminals of excavators, mixing plants, and transport vehicles;

[0050] On-site environmental data, such as real-time temperature, humidity, rainfall, and wind speed obtained through weather stations deployed at the construction site;

[0051] On-site sensing data, such as the real-time weight and volume of solid waste obtained through weighing sensors, lidar or 3D scanning equipment deployed in temporary solid waste storage sites, as well as image data collected by cameras.

[0052] Data access is achieved through specific communication protocols. For example, IoT data for construction machinery can use MQTT or OPCUA protocols, while environmental and meteorological data can be accessed by calling the RESTful API of third-party services. All accessed data carries precise timestamps and source device identifiers.

[0053] The core function of the spatiotemporal alignment unit is to precisely bind these timestamped dynamic data streams with the static construction activities in the aforementioned 4D construction foundation model. When a piece of real-time data (e.g., from excavator Exc-01 at time...) is generated... When the vibration frequency data enters the system, the spatiotemporal alignment unit performs a query operation: in the 4D construction foundation model, it retrieves the data at time... Construction activities that are currently active and whose resource lists include Exc-01 Once the query is successful, the real-time data entry is marked and categorized under construction activities. In the current instance record.

[0054] Through the above processing, the multi-source data access and BIM fusion module ultimately outputs a standardized activity instance data package. This data packet is a structured collection that provides direct input for subsequent processing. An active instance data packet. The data structure is defined as follows:

[0055] :

[0056] In the formula, Indicates at time Construction activities Instantiated data packets, Indicates construction activities A unique identifier, Indicates the timestamp of the current data collection. Indication and construction activities Associated BIM component information, Indicates at time The set of construction machinery status data associated with this activity, Indicates at time The set of environmental data associated with this activity, Indicates at time A collection of on-site perception data related to the event.

[0057] The construction process entropy dynamic calculation module is connected to the multi-source data access and BIM fusion module. It is used to receive the activity instance data packet and, based on the preset entropy factor model, calculate the construction process entropy value, which represents the degree of uncertainty of the construction process, in real time.

[0058] Specifically, in this embodiment, the construction process entropy dynamic calculation module receives standardized activity instance data packets output by the multi-source data access and BIM fusion module. It also calculates a key quantitative indicator, the entropy of the construction process, in real time.

[0059] The construction process entropy is decomposed into four independent influencing factors: mechanical efficiency factor, environmental impact factor, personnel impact factor, and process complexity factor. The construction process entropy (CPE) dynamic calculation module calculates the construction process entropy through the following processing units;

[0060] Mechanical efficiency factor The computational unit. This computational unit is based on the active instance data package. Real-time construction machinery status data Perform calculations. For example, calculate the real-time monitored mechanical efficiency. Compared with the standard efficiency preset in the system knowledge base In comparison, any deviation is considered an increase in uncertainty. Simultaneously, the degree of dispersion of the mechanical vibration frequency is considered. As a negative indicator of stability, the mechanical efficiency factor... Through a normalization function Perform the calculation:

[0061] :

[0062] In the formula, Indicates at time Mechanical efficiency factor Represents the normalization function. Indicates at time The working efficiency of construction machinery is obtained through real-time monitoring. This represents the standard mechanical working efficiency preset in the system's knowledge base. Indicates the relative deviation of mechanical efficiency. It indicates the degree of dispersion or standard deviation of mechanical vibration frequency.

[0063] Environmental impact factors The computational unit. This computational unit is based on the active instance data package. Real-time environmental data Calculations are performed. For specific construction techniques (such as concrete pouring), the system's knowledge base pre-defines ideal environmental parameter ranges, such as ideal temperature ranges. When the real-time ambient temperature Outside this range, or real-time rainfall Exceeding the allowable threshold of the process will increase the environmental impact factor. Environmental Impact Factor Through a normalization function Perform calculations;

[0064] ;

[0065] In the formula, Indicates environmental impact factors, Represents a function. The first part represents the temperature deviation. This represents the ideal minimum temperature. Indicates ambient temperature, as of time Changing temperature value, The second part representing the temperature deviation This represents the ideal maximum temperature. Indicates rainfall amount, indicates time. The amount of rainfall in the environment.

[0066] Personnel Influence Factor The calculation unit. The input to this calculation unit comes from the project management database and current construction activities. Information about the associated construction work teams. This information includes the average skill level of the teams. and average operational error rate based on historical data A lower skill level or a higher historical error rate corresponds to a higher personnel influence factor. Personnel Influence Factor Through a normalization function Perform the calculation:

[0067] :

[0068] In the formula, Indicates employee influence factor. Represents a function. The average skill level of employees Employees' historical operational error rate.

[0069] Process complexity factor The calculation unit. Unlike the aforementioned dynamic factors, the process complexity factor is a static value, which is calibrated during the project initialization phase based on the construction activity attributes in the 4D construction basic model. The system calculates the process complexity factor based on the construction activities. The inherent technological attributes, such as precision requirements and material specialties, are used to assign a complexity classification level to it. Process complexity factor Through a normalization function Perform the calculation:

[0070] ;

[0071] In the formula, Represents the process complexity factor. Represents a function. This indicates the classification level of process complexity.

[0072] Finally, the Construction Process Entropy (CPE) dynamic calculation module uses a weighted fusion model to integrate the above factors into a final value for construction activities. At any moment Entropy value of the construction process This ensures a comprehensive assessment of process uncertainties.

[0073] ;

[0074] In the formula, Represents the entropy of the construction process. The weighting coefficients representing the machine efficiency factor. This represents the machine efficiency factor. The weighting coefficients of environmental impact factors. Environmental impact factors This represents the weighting coefficient of the employee influence factor. Indicates employee influence factor. This represents the weighting coefficient of the process complexity factor. This represents the process complexity factor.

[0075] The material state migration network construction and evolution module is connected to the construction process entropy dynamic calculation module, and is used to dynamically adjust the migration probability from the design state to different solid waste forms in the material state migration network according to the construction process entropy value.

[0076] Specifically, in this embodiment, the material state migration network construction and evolution module receives the construction process entropy value calculated in real time. The function of this module is to build and maintain a probabilistic graphical model that can describe all physical transformation paths of materials throughout the entire construction cycle.

[0077] The graph structure definition of a material state transition network. In one specific embodiment, the material state transition network is a directed graph. .in, It is a collection of material status nodes. It is the set of state transition edges.

[0078] Nodes of a directed graph Used to represent a specific state of a material at a given moment. A material state node. It is innovatively defined as a triple. ,in: This indicates the type of material, such as concrete, steel bars, earthwork, etc. This represents the life cycle stage of a material, such as design stage, transportation stage, finished product stage, solid waste stage, etc. This represents the physical form of the material. A core technical feature of this invention lies in the refined definition of the solid waste state node, that is, based on the physical form of solid waste, the traditional single "solid waste" node is expanded into multiple specific solid waste form nodes. For example, the solid waste form nodes for concrete materials may include: (concrete, solid waste state, slurry state), (concrete, solid waste state, slurry state), (concrete, solid waste state, large-scale hardening), (concrete, solid waste state, large-scale hardening), etc. The solid waste form nodes for steel reinforcement materials may include: (steel reinforcement, solid waste state, short material), (steel reinforcement, solid waste state, short material), (steel reinforcement, solid waste state, corrosion), (steel reinforcement, solid waste state, corrosion), etc.

[0079] Edges of a directed graph Indicates the material from state to state Each migration is driven by a specific construction activity. Driven by this, the material state migration network generates an initial graph structure and baseline migration paths during system initialization, based on the standard construction process flow preset in the system knowledge base.

[0080] The migration probability is driven by a dynamic evolution mechanism of construction process entropy. The migration probability in the material state migration network is not static, but rather determined by the real-time entropy value output by the construction process entropy dynamic calculation module. Driven by this mechanism, the process model can reflect the uncertainties of the field conditions in real time.

[0081] For any activity Driven, from state Migrate to the next state The path, its migration probability A score based on migration tendency To calculate. Migration propensity score. It is the entropy of the construction process. Functions:

[0082] :

[0083] In the formula, Indicates from state After the event, Transition to state The path transition probability, This represents the transfer value within the pre-defined system knowledge base. Indicates from state to state Sensitivity coefficient of the construction process entropy corresponding to the path. This represents the entropy of the construction process.

[0084] When the migration path points to a desired finished product state, the operator in the formula is a minus sign, indicating that a higher entropy value will reduce the tendency of the path; when the path points to an undesirable solid waste state, the operator is a plus sign, indicating that a higher entropy value will increase the tendency of the path.

[0085] After obtaining the migration tendency scores for all possible paths starting from state sisi, the Softmax function is used to normalize these scores to obtain the final migration probability. :

[0086] :

[0087] In the formula, Indicates from state Through activities Transition to state The transition probability, Representing state From state Through activities The score or preference value upon arrival Represents an exponential function. This indicates that for all slave states Through activities The next state that can be transitioned to Summation, Indicates from state Through activities The set of all possible next states.

[0088] The solid waste form and resource profile prediction module is connected to the material state migration network construction and evolution module. It is used to perform probability simulation and deduction on the evolved material state migration network, generate the probability distribution of solid waste form within the future time window, and generate a resource demand profile representing the scheduling resource demand.

[0089] Specifically, in this embodiment, the solid waste morphology and resource profiling prediction module receives a material state migration network containing real-time migration probabilities. The core function of this module is to perform forward-looking probabilistic predictions, ultimately transforming the uncertainties of the construction process into quantifiable and probabilistic demands for future resource allocation.

[0090] The solid waste morphology and resource profiling prediction module implements its functions through the following processing units:

[0091] This unit is based on a probabilistic simulation and extrapolation unit of the evolved material state migration network. It first obtains a preset time window from the multi-source data access and BIM fusion module. The unit takes into account the construction plan, including the construction activities to be carried out and the types and quantities of materials to be used. Then, using these planned material inputs as initial inputs, the unit performs Monte Carlo simulation (MCS) on the Material State Transition Network (MSTN) construction and evolution module, which has been dynamically adjusted by the real-time construction process entropy value.

[0092] The specific process of Monte Carlo simulation is as follows: for a planned input of materials, the system executes... Sub-independent simulation path derivation (e.g., In each simulation, the material starts from its initial design state node, and at each non-terminal state node, the migration probability is determined based on all outgoing edges from the current node. A random sample is taken to determine its next state. This process continues until the material reaches a final state (such as a finished product or any form of solid waste).

[0093] After completion After each simulation, the probabilistic simulation unit statistically analyzes the quantity of all materials that ultimately fall into each solid waste morphology node, thereby outputting the future time window. Inside, each form of solid waste Production The probability distribution is used to construct a probability prediction set for solid waste morphology. It includes the expected value and variance of the amount of solid waste generated in various forms, which is used to characterize the central trend and degree of uncertainty of the forecast.

[0094] The Resource Demand Profile (RDP) generation unit. This unit's function is to transform forecasts of materials into forecasts of resource demand. The input to this unit is the solid waste form probability prediction set output by the probability simulation and deduction unit. Internally, it contains a pre-defined, configurable form-resource mapping knowledge base. .

[0095] Morphology-Resource Mapping Knowledge Base It must contain at least two parts of information:

[0096] Morphology-Resource Type Mapping It defines the processing of specific physical forms The collection of resource types required for solid waste For example, solid waste in the form of large, hardened blocks corresponds to resource types such as {rock breaker, heavy-duty dump truck}; waste in the form of slurry corresponds to resource types such as {slurry pump, sealed tanker truck}.

[0097] Resource quantification function It defines a certain quantity Solid waste, converted into corresponding resources Demand The specific calculation rules. For example, this function can... Tons of hardened solid waste converted to Hourly operation time of the hydraulic breaker, or The demand for heavy-duty dump truck transportation.

[0098] Resource Requirement Profile (RDP) generation unit utilization form - Resource Mapping Knowledge Base The probability prediction set of solid waste form Each item in the data, representing the probability distribution of the generation amount of each form of solid waste, is converted into a probability distribution of the corresponding processing resource requirements. Finally, this unit integrates all the probabilistic resource requirements to generate a structured Resource Demand Profile (RDP). The image is a vector, and its mathematical expression is:

[0099] :

[0100] In the formula, Image representing resource requirements. Indicates the first Types of adjustment resources Representing resources In time period Expected demand within the country Representing resources In time period The variance of domestic demand This indicates the total number of resource types.

[0101] The event-driven and adaptive scheduling decision module is connected to the solid waste form and resource profile prediction module and the multi-source data access and BIM fusion module. It is used to listen to scheduling events triggered by the resource demand profile or real-time on-site data, and start the multi-objective optimization solver to generate adaptive scheduling instructions that include immediate removal and resource pre-deployment.

[0102] Specifically, in this embodiment, the event-driven and adaptive scheduling decision module replaces the traditional rigid scheduling plan based on a fixed timetable, and realizes a forward-looking adaptive scheduling triggered by real-time events.

[0103] The event-driven and adaptive scheduling decision module implements its functions through the following processing units:

[0104] The triggering unit of an event-driven architecture.

[0105] The event-driven architecture's triggering unit continuously listens for three types of pre-defined events that can trigger scheduling decisions. This multi-event-source triggering mechanism is key to achieving a combination of proactive and reactive scheduling in this invention. The three types of events are clearly defined as follows:

[0106] Predicting events Resource demand profile generated by the Solid Waste Form and Resource Profile (RDP) prediction module Triggered directly. When A certain type of scheduling resource Demand Expectations or its demand probability Exceeding a preset pre-scheduling threshold When a predicted event is triggered, the event serves to drive the system to proactively deploy or reserve resources.

[0107] On-site incident Triggered by real-time on-site sensing data provided by the multi-source data access and BIM fusion module. When a temporary solid waste storage site... Weighing sensor readings Or lidar volume reading Exceeding a preset immediate removal threshold When this happens, an on-site event is triggered. The function of this event is to respond to a confirmed, existing waste removal need on-site.

[0108] Planned events This is triggered by continuous monitoring of the 4D construction plan by the multi-source data access and BIM fusion module. When a major change in the construction schedule is detected (such as the advancement or postponement of key processes), a planning event is triggered to drive the scheduling system to perform a global reassessment.

[0109] Scheduling Resource Pool Management Unit. This unit maintains in real time a virtualized resource pool containing all available scheduling resources. Each resource in the resource pool Each has its current state recorded. (e.g., idle, in use, under maintenance) and available quantity This unit provides real-time and accurate resource supply information for subsequent scheduling optimization.

[0110] A multi-objective scheduling optimization unit based on resource demand profiling. This unit is activated once any event is captured by the triggering unit. It constructs and solves a multi-objective optimization problem to generate the optimal scheduling instruction. The decision variable of this optimization problem is the scheduling scheme. This refers to the allocation strategy for available resources in the resource pool over a future period. The optimization objective is to minimize a comprehensive cost function. This function is a weighted sum of total scheduling cost, on-site waiting time for solid waste, and potential environmental risks:

[0111] :

[0112] In the formula, This represents the comprehensive optimization objective function. Indicate the scheduling scheme, The weighting coefficients representing scheduling costs. Representing the scheduling scheme The total cost involved, The weighting coefficient representing the waiting time. Representing the scheduling scheme The following waiting time, The weighting coefficients representing environmental risks Representing the scheduling scheme Environmental risks.

[0113] The optimization solution process must satisfy the following core constraints:

[0114] Deterministic requirement constraints: Scheduling scheme It must contain information that fully satisfies all on-site events. Triggered, confirmed, and immediately executed waste removal tasks.

[0115] Probabilistic demand constraints: for events predicted Triggered pre-scheduled tasks, scheduling scheme The reserved or pre-deployed resources must meet the resource demand profile in a manner no less than a pre-set confidence level. The probabilistic requirements in the process.

[0116] Resource availability constraints: in scheduling schemes At any time, any type of resource is allocated. The quantity cannot exceed the available quantity provided by the scheduling resource pool management unit. .

[0117] The optimization unit calculates the objective function by employing a constraint-satisfying programming (CSP) solver or other operations research optimization algorithms, thereby satisfying all the above constraints. Minimize the optimal scheduling scheme Ultimately, the optimal scheduling scheme is parsed into specific, executable scheduling instructions, which include not only immediate responses to on-site events but also pre-scheduling instructions based on resource demand profiles.

Claims

1. A BIM-based system for predicting and scheduling solid waste generation in highway engineering, characterized in that, include: The multi-source data access and BIM fusion module is used to parse BIM models and 4D construction plans, access multi-source heterogeneous data from construction machinery, site environment and solid waste dumps in real time, and perform spatiotemporal alignment and fusion to form standardized activity instance data packages. The construction process entropy dynamic calculation module is connected to the multi-source data access and BIM fusion module. It is used to receive the activity instance data packet and, based on the preset entropy factor model, calculate the construction process entropy value, which represents the degree of uncertainty of the construction process, in real time. The construction process entropy dynamic calculation module calculates the construction process entropy value through a weighted fusion model. The construction process entropy value is a weighted sum of mechanical efficiency factor, environmental impact factor, personnel impact factor and process complexity factor. The material state migration network construction and evolution module is connected to the construction process entropy dynamic calculation module, and is used to dynamically adjust the migration probability from the design state to different solid waste forms in the material state migration network according to the construction process entropy value. The material state migration network construction and evolution module is a directed graph, and its nodes are used to represent material states. The material status includes material type, life cycle stage, and physical form; Among them, nodes in the life cycle stage of solid waste are further subdivided into different solid waste form nodes according to their physical form. The solid waste form and resource profile prediction module is connected to the material state migration network construction and evolution module. It is used to perform probability simulation and deduction on the evolved material state migration network, generate the probability distribution of solid waste form within the future time window, and generate a resource demand profile that represents the scheduling resource demand. The solid waste form and resource profile prediction module also includes a form-resource mapping knowledge base; the solid waste form and resource profile prediction module uses the form-resource mapping knowledge base to convert the probability distribution of the solid waste form into a probabilistic demand for specific scheduling resources in order to generate the resource demand profile. The event-driven and adaptive scheduling decision module is connected to the solid waste morphology and resource profile prediction module and the multi-source data access and BIM fusion module. It is used to listen for scheduling events triggered by the resource demand profile or real-time on-site data, and start the multi-objective optimization solver to generate adaptive scheduling instructions that include immediate removal and resource pre-deployment.

2. The BIM-based highway engineering solid waste generation prediction and scheduling system according to claim 1, characterized in that, The multi-source heterogeneous data in the multi-source data access and BIM fusion module includes at least: Construction machinery status data, which are real-time working efficiency, energy consumption or vibration frequency collected from embedded sensors or vehicle terminals of construction machinery; On-site environmental data, which are real-time temperature, humidity, rainfall or wind speed obtained through weather stations deployed at the construction site; On-site sensing data refers to the real-time weight or volume of solid waste obtained through weighing sensors, lidar, or 3D scanning equipment deployed at temporary solid waste storage sites.

3. The BIM-based highway engineering solid waste generation prediction and scheduling system according to claim 1, characterized in that, The material state migration network construction and evolution module is specifically used for: Based on the entropy value of the construction process, the migration tendency score from one material state to all possible next material states is calculated. When the next material state is an undesirable solid waste form, a higher construction process entropy value will increase its migration tendency score. The migration propensity score is normalized using the Softmax function to obtain the migration probability.

4. The BIM-based highway engineering solid waste generation prediction and scheduling system according to claim 1, characterized in that, The solid waste form and resource profile prediction module is specifically used for: On the evolved material state migration network, a Monte Carlo simulation is performed on the planned input materials to count the number of materials entering each solid waste form node and output the probability distribution of the solid waste form.

5. The BIM-based highway engineering solid waste generation prediction and scheduling system according to claim 1, characterized in that, The event-driven and adaptive scheduling decision module includes the following scheduling events: The predicted event is triggered when the expected demand value or probability of a certain scheduled resource in the resource demand profile exceeds a preset pre-scheduling threshold. The on-site event is triggered when the on-site perception data provided by the multi-source data access and BIM fusion module exceeds a preset immediate removal threshold.

6. The BIM-based highway engineering solid waste generation prediction and scheduling system according to claim 1, characterized in that, The event-driven and adaptive scheduling decision module's multi-objective optimization solver constructs a cost function with the weighted sum of total scheduling cost, solid waste site waiting time, and potential environmental risks as the optimization objective, and solves the scheduling scheme with the objective of minimizing this cost function.

7. The BIM-based highway engineering solid waste generation prediction and scheduling system according to claim 1, characterized in that, The multi-objective optimization solver satisfies the following constraints during the solution process: Deterministic demand constraints require that the scheduling scheme must satisfy all deterministic cleanup tasks triggered by on-site events; Probabilistic demand constraints require that the resources pre-deployed by the scheduling scheme must meet the probabilistic demands triggered by predicted events and included in the resource demand profile in a manner no less than the preset confidence level.

Images