An intelligent scheduling and supervising system and method for construction waste transfer based on internet of things

Abstract

The application provides a building waste transfer intelligent scheduling supervision system and method based on Internet of Things, relates to the technical field of Internet of Things, and solves the problem that the existing technology is mutually separated in supervision data and scheduling decision and cannot dynamically adjust the scheduling scheme based on the real-time compliance state in the transportation process. The system comprises an Internet of Things sensing module, an edge processing module and a cloud scheduling platform; the Internet of Things sensing module comprises a stacking state sensing unit and a transportation state sensing unit, the stacking state sensing unit is used for acquiring spatial form features and quality features, and the transportation state sensing unit is used for acquiring running state features and transportation environment features; the edge processing module is used for preprocessing various sensing features and uploading to the cloud scheduling platform; and the cloud scheduling platform is used for generating an initial scheduling scheme and delivering to a vehicle-mounted execution terminal, and is also used for generating a reconstructed scheduling scheme and delivering to the vehicle-mounted execution terminal. The application is used for building waste transfer scheduling.

Description

Technical Field

[0001] This application relates to the field of Internet of Things (IoT) technology, and in particular to an IoT-based intelligent scheduling and monitoring system and method for the transfer of construction waste. Background Technology

[0002] Currently, in the field of construction waste transfer management, IoT technology is typically used to locate and track transport vehicles and monitor construction sites via video to achieve basic digital management. However, in actual operations, existing dispatching and monitoring systems are often disconnected. Dispatch decisions rely primarily on fixed time thresholds or manually reported inventory information, while compliance monitoring data during transportation (such as airtightness, spillage, and route deviation) is only used for post-event penalties or alarm alerts and cannot be fed back into the dispatching model in real time as a constraint. This disconnect between monitoring and dispatching means that when vehicles violate regulations or environmental anomalies occur, the system still executes the original optimal route or capacity allocation plan, lacking the ability to dynamically adjust based on real-time compliance status. Therefore, existing technologies suffer from the technical problem of disconnected monitoring data and dispatching decisions, and the inability to dynamically adjust dispatching plans based on real-time compliance status during transportation. Summary of the Invention

[0003] This application provides an intelligent scheduling and supervision system and method for construction waste transfer based on the Internet of Things, which solves the technical problems of existing technologies where supervision data and scheduling decisions are separated and the scheduling plan cannot be dynamically adjusted based on the real-time compliance status during transportation.

[0004] To achieve the above objectives, this application adopts the following technical solution: Firstly, an intelligent scheduling and monitoring system for construction waste transfer based on the Internet of Things (IoT) is provided, comprising: an IoT sensing module, an edge processing module, and a cloud-based scheduling platform. The IoT sensing module includes a stacking state sensing unit and a transportation state sensing unit. The stacking state sensing unit is used to acquire the spatial morphological and qualitative characteristics of the stacked materials in the target area, while the transportation state sensing unit is used to acquire the operational status characteristics of the transfer vehicles and the transportation environment characteristics. The edge processing module is communicatively connected to the IoT sensing module and is used to preprocess various sensing features and upload them to the cloud-based scheduling platform. The cloud-based scheduling platform is used to perform cross-validation based on the preprocessed spatial morphological and qualitative characteristics to generate stacking state data of the target area, and to trigger a scheduling task when the stacking state data meets the preset transfer conditions. Based on the scheduling task and a multi-objective optimization model, an initial scheduling scheme is generated and sent to the vehicle-mounted execution terminal. The cloud-based scheduling platform is also used to identify the compliance status of the transfer vehicles based on the preprocessed transportation environment characteristics and operational status characteristics, convert the compliance status into dynamic constraint parameters input to the multi-objective optimization model, dynamically adjust the initial scheduling scheme, generate a reconstructed scheduling scheme, and send it to the vehicle-mounted execution terminal.

[0005] The aforementioned system acquires both spatial morphology and mass characteristics through a stacking state sensing unit and performs cross-validation, solving the error problem of single-dimensional sensing in estimating the inventory of non-standardized construction waste and enabling precise triggering of transfer demand. At the same time, by transforming the compliance and regulatory status during transportation into dynamic constraint parameters and inputting them into a multi-objective optimization model, the system breaks down the data barriers between supervision and scheduling, enabling the scheduling scheme to be adaptively reconstructed based on real-time compliance status. This prevents non-compliant vehicles or high-risk routes from continuing to be preferred by the scheduling system, achieving real-time closed-loop intervention of supervision over scheduling.

[0006] In conjunction with the first aspect mentioned above, in one possible implementation, the system also includes a monitoring terminal; the monitoring terminal is used to synchronously display backlog status data, compliance monitoring status, and scheduling scheme execution data.

[0007] In conjunction with the first aspect mentioned above, in one possible implementation, the stacking state sensing unit includes a first detection subunit and a second detection subunit; the first detection subunit is used to scan the surface contour of the stack to generate point cloud data, and calculate the volume data and height data of the stack as spatial morphological features based on the point cloud data; the second detection subunit is used to collect the force data of the bearing surface of the stack, and calculate the weight data of the stack as mass features based on the force data and the preset gravity mapping relationship.

[0008] In conjunction with the first aspect mentioned above, in one possible implementation, the edge processing module has a built-in data verification engine and a local cache queue. The data verification engine is used to perform time-series alignment and outlier filtering on the received feature data to generate valid sensing data. The local cache queue is used to store valid sensing data when the communication link is interrupted, and to asynchronously synchronize it to the cloud scheduling platform in timestamp order after the communication link is restored.

[0009] In conjunction with the first aspect mentioned above, one possible implementation involves transforming the compliance and regulatory status into dynamic constraint parameters input into a multi-objective optimization model to dynamically adjust the initial scheduling scheme. This includes: the compliance and regulatory status includes environmental compliance status and trajectory compliance status; when a violation event is identified in either the environmental compliance status or the trajectory compliance status, the violation type and severity are determined; based on the violation type and severity, a fitness penalty factor for the transfer vehicle or the current driving path is calculated, and this fitness penalty factor is used as a dynamic constraint parameter; the fitness penalty factor is introduced into the fitness function of the multi-objective optimization model to reduce the overall fitness score of the transfer vehicle or the current driving path with the violation event; and based on the updated fitness function, candidate vehicles and candidate paths are re-evaluated, and the candidate vehicle and candidate path with the highest overall fitness score are selected as the reconstructed scheduling scheme.

[0010] In conjunction with the first aspect mentioned above, in one possible implementation, the fitness function of the multi-objective optimization model is used to calculate the comprehensive fitness score of candidate vehicles and candidate routes based on the transfer distance evaluation function, the capacity utilization evaluation function, the transportation time evaluation function, and the fitness penalty factor corresponding to each violation event; wherein, the fitness penalty factor is calculated based on the violation type and the severity of the violation through nonlinear mapping.

[0011] In conjunction with the first aspect mentioned above, in one possible implementation, the regulatory terminals include construction company terminals, transportation company terminals, and urban management regulatory terminals; the construction company terminals are used to query the accumulation status data and scheduling plan execution data of the target area; the transportation company terminals are used to query the operating status characteristics and compliance regulatory status of transfer vehicles; and the urban management regulatory terminals are used to query data for the entire region and to conduct traceability supervision.

[0012] In conjunction with the first aspect mentioned above, in one possible implementation, the preset transfer conditions include: the accumulated volume of the deposits in the target area reaches a preset volume threshold, or the accumulation time exceeds a preset time threshold.

[0013] In conjunction with the first aspect mentioned above, in one possible implementation, determining the violation type and severity of the violation event includes: identifying the violation type based on sensor data of the violation event; the violation events include: vehicle speeding, deviation from a preset route, unsealed cargo box, garbage spillage, illegal parking, and overloaded transportation; querying a preset violation event mapping table, which records the correspondence between various violation types and severity quantification values; and matching the identified violation type with the violation event mapping table to obtain the corresponding severity quantification value.

[0014] Secondly, an intelligent scheduling and supervision method for construction waste transfer based on the Internet of Things is provided, including: acquiring the spatial morphological and qualitative characteristics of the accumulated materials in the target area, as well as the operational status characteristics and transportation environment characteristics of the transfer vehicles; performing cross-validation based on the spatial morphological and qualitative characteristics to generate accumulation status data of the target area; triggering a scheduling task when the accumulation status data meets preset transfer conditions; generating and issuing an initial scheduling plan based on the scheduling task and a multi-objective optimization model; identifying the compliance supervision status of the transfer vehicles based on the transportation environment characteristics and operational status characteristics; converting the compliance supervision status into dynamic constraint parameters and inputting them into the multi-objective optimization model to dynamically adjust the initial scheduling plan, generate a reconstructed scheduling plan, and issue it.

[0015] This application acquires spatial morphological and qualitative features through a stacking state sensing unit and performs cross-validation. Utilizing the complementarity of multi-dimensional physical features, it eliminates the single-dimensional bias in sensing the stock of non-standardized construction waste, providing a high-precision data foundation for triggering scheduling tasks and avoiding capacity waste or overflow due to stock misjudgment. Furthermore, by transforming the compliance regulatory status identified from transportation environment and operational status features into dynamic constraint parameters and introducing them into the fitness function of a multi-objective optimization model, the scheduling algorithm can perceive changes in compliance risks during transportation in real time, automatically reducing the comprehensive fitness score of vehicles or routes with violations. This ensures transportation efficiency while proactively avoiding and intervening in violations in real time, solving the technical problems of separation between supervision and scheduling and the inability to dynamically respond to compliance status in traditional technologies. This significantly improves the intelligent supervision efficiency and operational safety of the construction waste transfer system.

[0016] It should be understood that the descriptions of technical features, technical solutions, beneficial effects, or similar language in this application do not imply that all features and advantages can be achieved in any single embodiment. Rather, it is understood that the description of a feature or beneficial effect means that a specific technical feature, technical solution, or beneficial effect is included in at least one embodiment. Therefore, the descriptions of technical features, technical solutions, or beneficial effects in this specification do not necessarily refer to the same embodiment. Furthermore, the technical features, technical solutions, and beneficial effects described in this embodiment can be combined in any suitable manner. Those skilled in the art will understand that embodiments can be implemented without one or more specific technical features, technical solutions, or beneficial effects of a particular embodiment. In other embodiments, additional technical features and beneficial effects may be identified in specific embodiments that do not embody all embodiments. Attached Figure Description

[0017] Figure 1 The system architecture of the Internet of Things-based intelligent scheduling and monitoring system for construction waste transfer provided in Embodiment 1 of this application Figure 1 ; Figure 2 The system architecture of an IoT-based intelligent scheduling and monitoring system for construction waste transfer provided in Embodiment 2 of this application Figure 2 ; Figure 3 The system architecture of an IoT-based intelligent scheduling and monitoring system for construction waste transfer provided in Embodiment 2 of this application Figure 3 ; Figure 4 The system architecture of the Internet of Things-based intelligent scheduling and monitoring system for construction waste transfer provided in Embodiment 3 of this application Figure 4 ; Figure 5 This is a flowchart illustrating an IoT-based intelligent scheduling and monitoring method for the transfer of construction waste, as provided in Embodiment 7 of this application. Detailed Implementation

[0018] In the description of this application, unless otherwise stated, " / " means "or," for example, A / B can mean A or B. "And / or" in this document is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A alone, A and B simultaneously, and B alone. Furthermore, "at least one" means one or more, and "multiple" means two or more. The terms "first," "second," etc., do not limit the quantity or order of execution, and "first," "second," etc., do not necessarily imply differences.

[0019] It should be noted that, in this application, the terms "exemplary" or "for example" are used to indicate that something is being described as an example, illustration, or illustration. Any embodiment or design described as "exemplary" or "for example" in this application should not be construed as being more preferred or advantageous than other embodiments or design solutions. Specifically, the use of terms such as "exemplary" or "for example" is intended to present the relevant concepts in a concrete manner.

[0020] Example 1: like Figure 1 As shown, this embodiment provides an intelligent scheduling and monitoring system for construction waste transfer based on the Internet of Things (IoT). This system constructs a complete data closed-loop architecture from front-end sensing to cloud-based decision-making and terminal execution. The system includes an IoT sensing module 101, an edge processing module 102, and a cloud-based scheduling platform 103.

[0021] The Internet of Things (IoT) sensing module 101 includes a stacking status sensing unit and a transportation status sensing unit. The stacking status sensing unit is used to acquire the spatial morphological and quality characteristics of the stacked objects in the target area, and the transportation status sensing unit is used to acquire the operating status characteristics of the transfer vehicles and the transportation environment characteristics.

[0022] As one implementation method, the stacking status sensing unit is deployed at various construction waste dumping sites. The spatial morphological features it collects can include geometric information such as the three-dimensional volume, stacking height, and surface contour distribution of the stacked material, while the mass features include physical attributes such as the real-time weight or density distribution of the stacked material. This multi-dimensional feature acquisition method differs from the traditional model that relies solely on single video monitoring or manual estimation, providing a heterogeneous data foundation for subsequent accurate quantification of waste inventory. The transportation status sensing unit is installed on the transfer vehicles. Its operational status features cover the vehicle's real-time location, speed, load status, and engine operating condition, while transportation environment features include the cargo box's sealing status, spillage monitoring data along the route, and images of the vehicle's interior. Through the collaborative work of these two types of sensing units, the system can simultaneously grasp the static stacking status of the "materials" and the dynamic transportation status of the "vehicles," providing comprehensive data support for intelligent scheduling.

[0023] The edge processing module 102 is communicatively connected to the IoT sensing module 101 and is used to preprocess various sensing features and upload them to the cloud scheduling platform 103. In practical application scenarios, the network environment at construction sites is often complex and unstable. As an intermediate hub connecting the sensing layer and the cloud, the edge processing module 102 undertakes the key functions of data cleaning and buffering. For example, the edge processing module 102 can perform noise reduction, outlier removal, and format standardization on the raw sensor data collected at high frequencies, uploading only the effective high-value data to the cloud, thereby significantly reducing the cloud's computing load and communication bandwidth pressure. It should be understood that although the edge processing module 102 is described as an independent module in this embodiment, in other embodiments, its function can also be integrated into a smart gateway or a sensing device with computing power, as long as local data preprocessing and reliable uploading can be achieved.

[0024] The cloud-based scheduling platform 103 is used to perform cross-validation based on preprocessed spatial morphological and quality characteristics to generate stacking status data for the target area. When the stacking status data meets preset transfer conditions, a scheduling task is triggered. An initial scheduling plan is generated based on the scheduling task and a multi-objective optimization model and sent to the vehicle-mounted execution terminal. This is the first core closed-loop mechanism for achieving precise scheduling triggering in this application. Due to the non-standardized characteristics of construction waste, such as complex composition, irregular stacking, and large fluctuations in moisture content, single-dimensional sensing data is prone to errors. For example, lightweight coverings may cause inflated volume, or sensor obstruction may lead to abnormal weight readings. Therefore, this embodiment introduces a cross-validation mechanism, utilizing the physical correlation between spatial morphological and quality characteristics for mutual verification. For example, when a significant increase in volume data is detected but the weight data does not change synchronously, the system can determine it as interference from lightweight floating objects or sensor malfunction, thereby automatically correcting or eliminating abnormal data and generating high-confidence stacking status data. Only when the validated stacking status data meets preset transfer conditions (such as volume thresholds or time thresholds) is a scheduling task triggered, effectively avoiding invalid dispatches or wasted transport capacity due to data misjudgment. Subsequently, the cloud-based scheduling platform 103 invokes a multi-objective optimization model to generate an initial scheduling plan by comprehensively considering factors such as distance, capacity, and time consumption, and then sends it to the vehicle-mounted execution terminal to guide vehicle operations.

[0025] The cloud-based scheduling platform 103 is also used to identify the compliance and regulatory status of transfer vehicles based on preprocessed transportation environment characteristics and operational status characteristics. This compliance and regulatory status is then transformed into dynamic constraint parameters input into a multi-objective optimization model, dynamically adjusting the initial scheduling plan, generating a reconstructed scheduling plan, and distributing it to the on-board execution terminal. This is the second core closed-loop mechanism for achieving the integration of supervision and scheduling in this application, and it is also the key difference from the separation of supervision and scheduling in existing technologies. In traditional technologies, violations (such as unsealed cargo, spillage, and route deviation) typically only trigger post-event alarms or penalties, failing to affect ongoing scheduling decisions in real time. In this embodiment, however, the compliance and regulatory status is transformed into dynamic constraint parameters at the algorithm level in real time. For example, when an environmental compliance issue such as an unsealed cargo box or minor spillage is identified in a vehicle, or a trajectory compliance issue such as frequent violation records is identified on a route, the system quantifies these issues as penalty factors or weight coefficients, directly inputting them into the fitness function of the multi-objective optimization model. This means that vehicles or routes with violation risks will have their comprehensive scores automatically reduced in the algorithm evaluation, thus enabling the model to actively avoid these high-risk objects when iteratively searching for the optimal solution, generating a reconstructed scheduling plan. This mechanism internalizes qualitative regulatory rules into quantitative scheduling constraints, realizing a shift from "passive regulation" to "proactive intervention," and ensuring that the scheduling scheme is not only optimal in terms of efficiency, but also under real-time control in terms of compliance.

[0026] Through the aforementioned dual closed-loop mechanism, this embodiment establishes an intelligent system driven by perception data for scheduling triggering and driven by regulatory status feedback for scheduling reconstruction. On the one hand, multi-dimensional feature cross-validation addresses the industry pain point of inaccurate perception of non-standardized construction waste inventory, improving the accuracy of scheduling triggering. On the other hand, real-time parameterized feedback of compliance regulatory status breaks down business silos, enabling the scheduling system to dynamically adapt to compliance risks, thereby significantly improving the overall regulatory efficiency and operational safety while ensuring transfer efficiency.

[0027] Example 2: Based on Example 1, such as Figure 2 As shown, the system of this application also includes a monitoring terminal 201, which is used to synchronously display the backlog status data, compliance monitoring status and scheduling scheme execution data.

[0028] As one implementation method, the monitoring terminal 201, serving as a carrier for human-computer interaction and data visualization, is not simply a data display screen, but an intelligent client integrating multi-source heterogeneous data fusion and rendering capabilities. Synchronous display means that the monitoring terminal 201 maintains millisecond-level or second-level data heartbeats with the cloud-based dispatch platform 103 through long-term connections or high-frequency polling mechanisms. This ensures that the accumulated volume, vehicle violation records, and work order progress presented on the front-end interface remain consistent with the real-time status of the cloud database, eliminating the information lag caused by manual reporting or periodic report export in traditional models. For example, when the cloud-based dispatch platform 103 updates the accumulated status data at a certain location based on cross-validation, or generates a new compliance monitoring status alarm due to the identification of spillage, the monitoring terminal 201 can automatically refresh the corresponding view within a preset time window, allowing managers to obtain the latest situation without manual operation. This real-time synchronization mechanism provides a unified and immediate data benchmark for multi-entity collaborative management, avoiding scheduling command conflicts or regulatory blind spots caused by information asymmetry.

[0029] As an optional implementation method, such as Figure 3As shown, the monitoring terminal 201 includes a construction company terminal, a transportation company terminal, and an urban management monitoring terminal. The construction company terminal is used to query the accumulation status data and scheduling plan execution data of the target area; the transportation company terminal is used to query the operational status characteristics and compliance monitoring status of transfer vehicles; and the urban management monitoring terminal is used to query data for the entire area and perform traceability supervision. This three-tiered permission architecture is designed not only to meet the division of responsibilities in administrative management but also for technical decoupling from the perspectives of system security and data processing efficiency. Specifically, the construction company terminal is configured to only access point data within its bound specific geofence. Its data request interface has its query scope parameters restricted at the gateway layer, thus preventing the leakage of sensitive data across construction sites. Furthermore, since only local data needs to be loaded, this terminal can adopt a high-frequency (e.g., once every 5 seconds) proactive push mode to meet the strong dependence of construction sites on the timeliness of waste removal. The transportation company terminal focuses on the full lifecycle management of its transportation assets. Its data view covers the GPS tracks, load curves, sealed sensor status, and historical violation records of all its vehicles, but does not include operational data of other companies or internal project progress information of construction parties. This terminal typically adopts a hybrid mode combining event-triggered updates and scheduled data retrieval to balance real-time monitoring needs with mobile data consumption. The urban management supervision terminal has the highest level of data access permissions, capable of accessing historical ledgers for the entire region, cross-enterprise correlation analysis, and macro-statistical reports. It also has traceability and supervision functions, that is, it can trace back to the original data of specific sensing devices, edge processing logs, and scheduling decision snapshots through any abnormal record. Given the huge bandwidth and computing power overhead of full data query, this terminal uses an on-demand loading and asynchronous query mechanism by default, retrieving detailed data from cold storage or data warehouses only when the user explicitly initiates a search command.

[0030] It should be noted that the aforementioned hierarchical access control relies on a combination of a role-based access control model and a dynamic data masking engine. When accounts with different identities log in to the monitoring terminal 201, the cloud scheduling platform 103 dynamically generates API response payloads based on a pre-set permission matrix. Fields that the user does not have permission to access are directly masked or removed, rather than hidden on the front end, thus avoiding the risk of obtaining unauthorized data through packet sniffing or tampering with front-end code. Furthermore, the differentiated data refresh strategies for different terminals effectively reduce cloud concurrency pressure and network bandwidth consumption, enabling the system to support stable access from a large number of terminals while ensuring the real-time performance of core business operations.

[0031] Example 3: In some embodiments of this application, such as Figure 4As shown, the stacking status sensing unit includes a first detection subunit and a second detection subunit. This hardware architecture is designed to provide a dual verification basis for estimating the stock of non-standardized construction waste through the physical complementarity of heterogeneous sensors, thereby solving the technical challenge of single-dimensional sensing being susceptible to interference in complex construction site environments.

[0032] In one implementation, the first detection subunit scans the surface contour of the accumulation to generate point cloud data. Based on the point cloud data, it calculates the volume and height of the accumulation as spatial morphological features. In actual deployment, the first detection subunit typically uses active 3D imaging equipment such as LiDAR or depth cameras, mounted on a fixed bracket or gantry above the stacking area. Taking LiDAR as an example, it acquires a dense set of spatial coordinate points on the surface of the accumulation, i.e., point cloud data, by emitting a high-frequency laser beam and receiving the reflected echo. Then, after denoising, ground segmentation, and surface fitting processing of the original point cloud, a 3D geometric model of the accumulation is constructed. The accurate volume data is then calculated by integration, and the Z-axis coordinate of the highest point is extracted as the height data. This optical or electromagnetic wave-based geometric measurement method has the advantages of being non-contact and having a fast response, enabling it to keenly capture dynamic changes in the morphology of the accumulation.

[0033] The second detection subunit is used to collect stress data on the bearing surface of the stockpile and calculates the weight data of the stockpile as a mass characteristic based on the stress data and a preset gravity mapping relationship. The second detection subunit typically uses a high-precision gravity sensor or pressure transmitter array embedded in the foundation of the stockpile site or a weighbridge system. These sensors directly sense the vertical load applied by the stockpile to the bearing surface and output analog or digital electrical signals. The system converts the raw stress data into standardized weight data according to a pre-calibrated gravity mapping relationship (which takes into account sensor sensitivity, the leverage ratio of the installation position, and the foundation settlement compensation coefficient). Unlike optical measurements, weight data directly reflects the inertial mass of the material and is unaffected by color, texture, light transmittance, or surface coverings, providing extremely high measurement reliability for typical construction waste such as high-density bricks and concrete blocks.

[0034] Based on the physical differences between the two sensing modes, this embodiment achieves a qualitative leap in data quality through a cross-validation mechanism. The specific logic of cross-validation is not a simple numerical comparison, but a dynamic verification process built on the principles of matter conservation and physical correlation. For example, if the volume data reported by the first detection subunit increases significantly in a short period (e.g., an increase exceeding 20%), but the weight data reported by the second detection subunit does not show a synchronous growth trend (e.g., an increase less than 5%), the system can determine that the current volume increase is mainly caused by lightweight floating objects (e.g., plastic sheeting or foam board blown in by the wind), or that the first detection subunit has generated false point clouds due to specular reflection or multipath effects. In this case, the system will automatically reduce the confidence weight of the volume data at that moment, or even mark it as an outlier and remove it to avoid triggering unnecessary scheduling tasks. Conversely, if the weight data suddenly increases while the volume data does not change significantly, it may indicate the presence of high-density metal waste or interference from external impacts on the gravity sensor, and the system will also initiate a review process.

[0035] As an example, in actual operation at a construction site, when a strong gust of wind blew down a large area of ​​dust-proof netting and covered the top of the waste pile, the volume scanned by the lidar instantly increased by 3.5 cubic meters, but the weighbridge reading only fluctuated by 15 kilograms. The system performed cross-validation based on a preset reasonable density range (typical density range for construction waste is 800-1600 kg / m³), calculating that the current apparent density was only 4.3 kg / m³, far below the lower threshold. Therefore, it was determined to be an environmental disturbance event, maintaining the original pile status data unchanged, and sending a sensor status alarm to the monitoring terminal 201. After the windbreak netting was removed, the volume and weight data returned to the linear correlation range, and the system automatically resumed high-confidence output. This mechanism effectively avoids frequent invalid dispatches caused by false alarms from a single sensor, improving the accuracy of dispatch triggering.

[0036] It should be noted that although this embodiment uses the combination of lidar and gravity sensor as the preferred implementation, in other application scenarios, the first detection subunit can also be replaced by a binocular stereo vision camera, structured light scanner or millimeter-wave radar, and the second detection subunit can also be replaced by a distributed fiber optic strain sensor or hydraulic weighing module. As long as the two provide independent spatial morphology information and mass information respectively, and can form a physical cross-verification relationship, they are all within the protection scope of this application.

[0037] Based on the above technical solution, this embodiment overcomes the industry pain points of "inaccurate measurement and unreliable data" in the perception of non-standard construction waste inventory by hardware-level fusion and physical correlation verification of heterogeneous sensing units. It provides a highly reliable and interference-resistant data foundation for the intelligent scheduling system, ensuring that subsequent scheduling triggering and capacity allocation decisions are always based on real and reliable physical states.

[0038] Example 4: In some embodiments of this application, the edge processing module 102 incorporates a data verification engine and a local cache queue. This architecture is specifically optimized for the complex network environment and multi-source heterogeneous sensor characteristics of construction sites. In actual operation sites, wireless network signals often experience high latency, packet loss, or even long-term interruptions due to tower crane obstruction, metal structure reflection, or base station load fluctuations, and the sampling frequencies of different sensors vary significantly. If the raw data is directly transmitted to the cloud, it will not only lead to bandwidth waste but may also cause the cross-validation of the cloud scheduling model to fail due to data disorder or missing data. Therefore, this embodiment constructs a "firewall" and "reservoir" to ensure data quality by deploying a data verification engine with computing power and a local cache queue with persistent capabilities on the edge side, ensuring that the data uploaded to the cloud scheduling platform 103 always has temporal consistency and content integrity.

[0039] As one implementation method, the data verification engine performs temporal alignment and outlier filtering on the received feature data to generate valid sensing data. Temporal alignment aims to solve the data misalignment problem caused by inconsistent sampling rates of multiple sensors. For example, the point cloud acquisition frequency of the first detection subunit (LiDAR) may be 10Hz, while the weight reporting frequency of the second detection subunit (gravity sensor) is only 1Hz. The data verification engine uses a sliding time window algorithm, taking the sampling time of the low-frequency sensor as the base frame, and performs linear interpolation or weighted average aggregation on the high-frequency sensor data, mapping both to a unified timestamp, thereby providing a strictly synchronized input pair for cross-validation. Outlier filtering goes beyond conventional noise smoothing processing, introducing dual discrimination rules based on physical constraints and business logic. On the one hand, the system pre-sets the physical extreme ranges of various sensors (e.g., weight 0-50 tons, volume 0-100 cubic meters). Readings exceeding these ranges are directly identified as sensor malfunctions or electromagnetic interference and discarded. On the other hand, a dynamic mutation threshold is set. When the rate of change of values ​​between two adjacent sampling periods exceeds a preset upper limit (e.g., weight changes by more than 2 tons within 1 second) and there is no corresponding loading / unloading operation signal, it is identified as transient impact interference and filtering is triggered. Only data that passes the above dual verification is marked as valid sensing data and enters the subsequent transmission or caching process. This edge-side cleaning mechanism significantly reduces the computing power overhead of cloud processing dirty data and prevents misscheduling caused by single-point sensor anomalies.

[0040] Furthermore, a local cache queue is used to store valid sensing data when the communication link is interrupted, and asynchronously synchronizes it to the cloud scheduling platform 103 in timestamp order after the communication link is restored. The core value of this mechanism lies in ensuring the time-series fidelity of the data, rather than simply preventing data loss. When the heartbeat monitoring mechanism of the edge processing module 102 detects an interruption in the communication link with the cloud, the valid sensing data generated by the data verification engine is encapsulated into data packets with precise UTC timestamps and appended to a circular buffer queue in the local non-volatile storage area (such as an industrial-grade SD card or eMMC) in chronological order. It should be noted that this storage is strictly ordered; even if there is a slight delay between the data generation time and the writing time, the business timestamp within the data packet still retains the original collection time, ensuring the traceability of historical data. Once the communication link is restored, the system does not immediately push all backlogged data at once to avoid network congestion or triggering cloud-based rate limiting. Instead, it initiates an asynchronous batch synchronization strategy: background threads read data packets from the cache queue in batches, in ascending order of timestamps, and upload them. After each batch is uploaded, it waits for cloud confirmation before sending the next batch, until the queue is cleared or the real-time data stream is reached. This "old first, new later, orderly retransmission" strategy ensures that the data stream received by the cloud scheduling platform 103 is continuous and monotonically increasing on the timeline, completely avoiding out-of-order arrival of old and new data due to network jitter, and thus preventing the scheduling model from making incorrect capacity allocation decisions due to receiving delayed backlogged data.

[0041] Example 5: In some embodiments of this application, the process of transforming the compliance regulatory status into dynamic constraint parameters input into a multi-objective optimization model and dynamically adjusting the initial scheduling scheme is a key step in establishing a closed loop between regulatory data and scheduling algorithms.

[0042] As one implementation method, compliance monitoring status includes environmental compliance status and trajectory compliance status. Environmental compliance status primarily reflects the impact of transport vehicles on the urban environment during transportation. Its data sources may include the status of the vehicle's airtightness sensors, images of spillage monitoring along the route, and dust concentration readings at loading and unloading areas. Trajectory compliance status reflects the vehicle's adherence to traffic regulations and predetermined routes. Its data sources may include the comparison results of GPS / BeiDou positioning trajectories and electronic fences, vehicle speed sensor readings, parking duration, and location information. This classification method decouples the abstract concept of "compliance" into two orthogonal dimensions, enabling the system to adopt differentiated response strategies for different types of risks. For example, for environmental violations, the focus is more on preventing pollution spread, while for trajectory violations, the focus is more on safety and order control.

[0043] When a violation event is detected in the environmental compliance status or trajectory compliance status, the system determines the type and severity of the violation event.

[0044] In this embodiment, the system first identifies the violation type based on sensor data of the violation event. Violations include speeding, deviation from a preset path, unsealed cargo box, garbage spillage, illegal parking, and overloading. For example, when the cargo box sealing sensor reports an "open" signal and the vehicle speed exceeds 20 km / h, it is identified as "unsealed cargo box"; when the vehicle's location point deviates from the preset path buffer zone by more than 50 meters for more than 2 minutes, it is identified as "deviation from the preset path". Subsequently, the system queries a preset violation event mapping table, which records the correspondence between various violation types and severity quantification values. By matching the identified violation type with the violation event mapping table, the corresponding severity quantification value can be obtained.

[0045] For example, the baseline quantification value for "unsealed cargo container" in the mapping table is 0.3, for "garbage spillage" it is 0.6, and for "overloaded transportation" it is 0.8. In practical applications, this baseline quantification value can be dynamically adjusted based on factors such as the duration and frequency of the violation, for example, by adding an additional 0.05 points for each minute of violation, thus more accurately reflecting the actual severity of the violation. This standardized quantification mechanism based on the mapping table provides a unified benchmark for subsequent accurate calculation of penalty factors, ensuring the consistency and interpretability of the scheduling reconfiguration logic.

[0046] Based on the type and severity of the violation, the system calculates the fitness penalty factor for the transfer vehicle or the current travel route, and uses the fitness penalty factor as a dynamic constraint parameter.

[0047] The fitness penalty factor is a dimensionless numerical value, whose magnitude directly represents the "resistance" to which the vehicle or route is selected by the dispatching system at the current moment. To reflect the guidance of regulatory policies and ensure the stability of the dispatching system, the fitness penalty factor is calculated through a nonlinear mapping based on the type and severity of the violation. Specifically, the nonlinear mapping relationship typically takes the form of an exponential function. For example, for minor violations (such as short-term, low-speed inadequate sealing), the penalty factor increases linearly or exponentially with severity, allowing the system to still consider it as an alternative when capacity is extremely tight, with only a moderate reduction in weight. For serious violations (such as malicious dumping, severe overloading, or entering restricted areas), the penalty factor increases exponentially, causing its overall score to instantly drop outside the feasible region, thus achieving "hard isolation" at the algorithm level. This nonlinear design solves the dilemma of traditional linear deduction mechanisms: if the slope is too small, serious violations cannot be effectively eliminated; if the slope is too large, even slight disturbances will cause drastic fluctuations in the dispatching scheme. Through nonlinear mapping, the system maintains flexible dispatching capabilities to cope with unexpected situations while ensuring compliance.

[0048] Furthermore, a fitness penalty factor is introduced into the fitness function of the multi-objective optimization model to reduce the overall fitness score of transfer vehicles or current travel routes that have violated regulations.

[0049] The fitness function of the multi-objective optimization model is used to calculate the comprehensive fitness score of candidate vehicles and candidate routes based on the transfer distance evaluation function, capacity utilization evaluation function, transportation time evaluation function, and fitness penalty factors corresponding to each violation event. In a specific implementation, the fitness function... It can be represented as:

[0050] in, This is the evaluation function value for transit distance; the shorter the distance, the higher the score. The calculation formula is as follows: ,in As a reference distance, This represents the actual distance the candidate vehicle travels from its current location to the target loading point. To prevent extremely small positive numbers with a denominator of zero, this function ensures that candidate vehicles that are closer to each other receive higher distance scores. This is the load factor assessment function value. The closer the load factor is to the rated load capacity, the higher the score. The calculation formula is as follows: ,in This represents the actual load capacity of the candidate vehicles. The rated load capacity of the candidate vehicles; this function encourages the scheduling system to prioritize vehicles with loading rates close to full capacity in order to reduce wasted capacity. This is a function value for assessing transportation time; the earlier the estimated arrival time, the higher the score. The calculation formula is as follows: ,in The latest time required to complete the task. The function estimates the arrival time of candidate vehicles; it allows candidate solutions that can complete the task earlier to receive higher scores. , , These are the weighting coefficients for the three efficiency indicators mentioned above, and The weighting coefficients can be dynamically adjusted according to the actual business scenario, for example, they can be appropriately increased when the project schedule is tight. The value of . For the first The fitness penalty factor corresponding to each violation event is calculated using the following formula: ,in For the first A quantitative value for the severity of each violation. For the first The frequency of violations in each violation incident The frequency sensitivity coefficient is used to control the amplification of the penalty factor by the frequency of violations; this nonlinear mapping causes the penalty factor for high-frequency violations or high-severity violations to grow exponentially. This is the adjustment coefficient for this type of violation, used to balance the weighting of different types of violations on the total score; This represents the total number of violations associated with the current candidate.

[0051] The candidate vehicles and paths are re-evaluated based on the updated fitness function, and the candidate vehicle and path with the highest comprehensive fitness score are selected as the reconfiguration scheduling scheme. This process is not a simple threshold filtering, but a natural selection process completed during the algorithm's iterative search. For example, suppose that in a certain scheduling task, candidate vehicle A initially has the highest fitness score of 0.85 due to its closest distance and shortest empty driving distance. However, during transportation, the on-board terminal reports a violation of "cargo box not sealed," which the system identifies as an environmental compliance violation with a moderate severity quantification value. A fitness penalty factor is calculated through nonlinear mapping. After substituting into the fitness function, vehicle A's comprehensive score drops to 0.55. At the same time, candidate vehicle B, although slightly farther away and with an initial efficiency score of only 0.70, has a good historical compliance record and no current violations, resulting in a penalty term of 0 and a comprehensive score remaining at 0.70. At this point, the system re-evaluates and finds that vehicle B's score is higher than vehicle A's. Therefore, it automatically switches the optimal solution from vehicle A to vehicle B, generates a reconfiguration scheduling scheme containing vehicle B's information and a new path, and issues it. This restructuring process is entirely data-driven and requires no human intervention. It not only promptly avoids the potential for escalating violations but also ensures the optimal global allocation of remaining transportation capacity resources, truly achieving real-time integration and dynamic collaboration between supervision and scheduling.

[0052] Example 6: In some embodiments of this application, the preset transfer conditions include the accumulation volume of the deposits in the target area reaching a preset volume threshold, or the accumulation time exceeding a preset time threshold.

[0053] As one implementation method, the preset volume threshold can be set to 5 cubic meters, and the preset time threshold can be set to 24 hours. This dual threshold triggering mechanism is not a simple numerical superposition, but a complementary protection logic built based on the physical characteristics of construction waste and the needs of environmental governance.

[0054] The threshold of 5 cubic meters for accumulated volume is primarily used to prevent overflow of physical space on site. When the three-dimensional volume detected by lidar or depth camera exceeds this limit, it indicates that the storage area is nearing saturation. If not handled promptly, construction waste may encroach on construction access roads or surrounding public areas. Triggering this condition ensures the operational order and safety of the construction site. The threshold of 24 hours for accumulation time focuses on the timeliness of environmental management. Even if some locations do not reach the upper limit due to small amounts of waste, if construction waste is left in the open for a long time, it is very easy for dust pollution to occur due to wind and sun exposure, leachate to occur due to rainwater leaching, or foul odors to occur due to fermentation, seriously affecting the lives of surrounding residents and the city's appearance.

[0055] Based on the above technical solution, by introducing the time dimension as a fallback trigger condition, the system can effectively identify and deal with those hidden pollution sources that are "small in quantity but long-lasting", avoiding the management blind spot of low-volume construction waste that has been left uncollected for a long time due to relying solely on volume thresholds.

[0056] Example 7: like Figure 5 As shown, this embodiment provides an intelligent scheduling and monitoring method for construction waste transfer based on the Internet of Things. This method is applied to the system described in any of the foregoing embodiments and achieves closed-loop control of the entire chain from perception triggering to scheduling reconfiguration through time-sequential steps. The method includes the following steps: S501. Obtain the spatial morphological and qualitative characteristics of the accumulated materials in the target area, as well as the operational status and transportation environment characteristics of the transfer vehicles.

[0057] Among them, spatial morphological characteristics refer to data that characterize the geometric properties of the accumulation, mass characteristics refer to data that characterize the physical weight of the accumulation, operational status characteristics refer to data that reflect the working condition of the transport vehicle itself, and transportation environment characteristics refer to data that reflect the compliance of the vehicle's external operations.

[0058] In this embodiment, the system uses a stacking status sensing unit deployed at the construction site to collect real-time spatial morphological features of construction waste within the target area, such as three-dimensional volume and stacking height, as well as mass features such as real-time weight obtained through weighing equipment. Simultaneously, through vehicle-mounted sensing devices installed on the transport vehicles, it continuously acquires operational status features such as GPS positioning, speed, and load factor, as well as transportation environment features such as the cargo box sealing sensor status and spillage monitoring images along the route. This data stream acquisition is parallel and continuous, providing comprehensive raw information input for subsequent processing.

[0059] It should be noted that the various feature data acquired in this step should undergo preprocessing by the edge processing module 102, including time-series alignment and outlier filtering, to ensure the validity and synchronization of the data uploaded to the cloud. Furthermore, the acquisition process is performed automatically by the system, rather than relying on manual data entry, thus ensuring the objectivity and timeliness of the data.

[0060] Based on the above steps, the automated, multi-dimensional, and real-time acquisition of all elements of construction waste transfer data has been achieved, eliminating information lag and subjective errors in the traditional manual reporting mode, and providing a reliable data foundation for subsequent accurate perception and intelligent scheduling.

[0061] S502. Based on spatial morphological features and quality features, cross-validation is performed to generate stacking state data of the target area.

[0062] Cross-validation refers to the logical processing of verifying data by utilizing the inherent correlation between features of different physical dimensions. Stacked state data refers to standardized data that has high confidence after verification and is used to characterize the true state of waste inventory.

[0063] In this embodiment, the system performs correlation analysis on spatial morphological features and mass features within the same time window. As one implementation, based on the typical density distribution range of construction waste, it determines whether the ratio of current volume data to weight data is within a reasonable range. If the trends of both are consistent and the ratio is within a preset threshold, the data is deemed valid, and they are fused to generate stacking state data containing precise volume and weight. If abnormal situations occur, such as a sudden increase in volume without a change in weight, or a sudden increase in weight without a change in volume, then according to a preset confidence assessment strategy, the values ​​of abnormal dimensions are removed or their weights are reduced, retaining only the data of the reliable dimensions as the stacking state data output.

[0064] It should be noted that the core purpose of cross-validation is to address the problem of single-dimensional perception distortion caused by the complex composition and irregular stacking of non-standard construction waste. The generated stacking status data not only includes the numerical values ​​themselves, but also data confidence indicators, which can serve as a reference factor for subsequent scheduling decisions.

[0065] As an example, when the system detects that the volume data of a certain point has increased by 3 cubic meters in 1 minute, but the weight data has only increased by 0.1 tons, the apparent density is calculated to be far below the minimum density threshold of construction waste. The system determines that the volume increase is caused by dust netting or interference from lightweight floating objects. Therefore, when generating the accumulation status data, the system automatically removes the volume increase, maintains the original stock record, and avoids the generation of false overflow signals.

[0066] Based on the above steps, the complementary verification mechanism of multi-dimensional physical features effectively filters out environmental interference and sensor noise, generates highly reliable stacking state data, significantly improves the accuracy of scheduling triggers, and prevents capacity waste or invalid responses caused by data misjudgment.

[0067] S503. When the accumulated data meets the preset transfer conditions, a scheduling task is triggered, and an initial scheduling scheme is generated and issued based on the scheduling task and the multi-objective optimization model.

[0068] Among them, the preset transfer conditions refer to the threshold rules that are pre-configured to determine whether the cleaning operation needs to be started, and the initial scheduling scheme refers to the theoretically optimal order dispatch and route planning results calculated based on the static road network and transportation resources without considering real-time dynamic violation factors.

[0069] In this embodiment, the system continuously monitors the generated stacking status data. When the data meets preset transfer conditions (such as the stacking volume reaching a preset volume threshold or the stacking time exceeding a preset time threshold), a scheduling task request is automatically generated. Subsequently, the cloud scheduling platform 103 calls a multi-objective optimization model, using the shortest transfer distance, highest capacity utilization, and least transportation time as objective functions, to search for the optimal solution in the currently available capacity pool. An initial scheduling plan is generated, containing information such as the specified vehicle ID, recommended driving route, and estimated arrival time. This plan is then distributed to the corresponding vehicle-mounted execution terminal via a wireless network.

[0070] It should be noted that the initial scheduling scheme is generated based on static snapshot data at the current moment, representing the optimal efficiency solution for the system under ideal compliance conditions. The actions are explicitly sent to the onboard execution terminal, rather than directly to the driver's personal device, demonstrating the system's standardized control over the execution end.

[0071] As an example, when the accumulated volume at a certain location reaches the threshold of 5 cubic meters, the system triggers a scheduling task. Through model calculation, the system selects the nearest and idle vehicle B, plans an optimal route to avoid the restricted road section during the morning rush hour, and pushes the initial scheduling plan, which includes navigation instructions and work order details, to the vehicle's onboard tablet.

[0072] Based on the above steps, an automated connection was achieved from waste inventory perception to scheduling task triggering, ensuring timely response to transfer needs, and the efficiency of initial resource allocation was optimized through a multi-objective optimization model.

[0073] S504. Identify the compliance and regulatory status of transfer vehicles based on transportation environment characteristics and operational status characteristics.

[0074] Among them, compliance and regulatory status refers to the comprehensive evaluation result of whether the transfer vehicle complies with urban management regulations and safety standards during the transportation process, which usually includes two dimensions: environmental compliance status and trajectory compliance status.

[0075] In this embodiment, the system analyzes the received transportation environment and operational status characteristics in real time and performs compliance determination through a preset rule engine or recognition model. For example, it correlates cargo box sealing sensor signals with vehicle speed signals to determine whether there is any unsealed behavior during driving; it compares GPS tracks with electronic fences to determine whether there is any route deviation or illegal parking; and it compares load data with the rated load to determine whether there is any overloading. The recognition results are structured into a compliance regulatory status data stream containing the type of violation, time of occurrence, duration, and severity.

[0076] It is important to note that the identification of compliance status is performed in real time and is independent of the lifecycle of the dispatch task. Even if a vehicle is not performing a current dispatch task, its compliance status is continuously monitored and recorded so that it can be used as historical credit data in subsequent dispatch assessments. This step transforms previously discrete sensor readings into regulatory information with business semantics.

[0077] Based on the above steps, real-time digital perception of compliance throughout the entire transfer process is achieved, transforming physical violations into logical states that the system can process, thus providing the necessary prerequisites for regulatory data-driven scheduling decisions.

[0078] S505. Transform the compliance and regulatory status into dynamic constraint parameters and input them into the multi-objective optimization model to dynamically adjust the initial scheduling scheme, generate a reconstructed scheduling scheme, and issue it.

[0079] Among them, dynamic constraint parameters refer to mathematical variables that can quantify the degree of impact of compliance risks on scheduling objectives, and reconstructed scheduling schemes refer to new scheduling decisions that take into account both efficiency and security, obtained by re-optimizing the model after introducing real-time compliance constraints.

[0080] In this embodiment, the system first maps the identified compliance and regulatory status to a specific fitness penalty factor, which is injected as a dynamic constraint parameter into the fitness function of the multi-objective optimization model. Next, based on the updated fitness function, the model re-evaluates and scores the current candidate vehicles and candidate paths. Since objects with violations are penalized, resulting in a lower overall score, the model automatically avoids these high-risk objects during iterative search, thereby selecting a new optimal solution as the reconstructed scheduling scheme. Finally, the system distributes the reconstructed scheduling scheme to the onboard execution terminal, replacing or correcting the original initial scheduling scheme.

[0081] It's important to note that this "dynamic adjustment" is not simply manual intervention or rule filtering, but rather an adaptive reconstruction at the algorithmic level. The transformation process typically employs a non-linear mapping mechanism, resulting in differentiated gradients in the impact of minor and severe violations on scheduling outcomes. Furthermore, the reconstructed scheduling plan is issued in real time. Once a significant change in compliance risk is detected, the system can complete recalculation and instruction updates within milliseconds, ensuring that vehicles always follow the currently optimal and compliant path to execute tasks.

[0082] Based on the above steps, the data barrier between the regulatory system and the scheduling system is broken down, and the compliance and regulatory status is internalized as the core constraint variable of the scheduling algorithm. This enables the scheduling scheme to dynamically evolve with the real-time compliance situation, ensuring transfer efficiency while achieving proactive avoidance and real-time intervention of compliance risks.

[0083] Although this application has been described herein in conjunction with various embodiments, those skilled in the art, by reviewing the accompanying drawings, disclosure, and appended claims, will understand and implement other variations of the disclosed embodiments in carrying out the claimed application. In the claims, the word "comprising" does not exclude other components or steps, and "a" or "an" does not exclude multiple instances. A single processor or other unit can implement several functions listed in the claims. While different dependent claims may recite certain measures, this does not mean that these measures cannot be combined to produce good results.

[0084] Although this application has been described in conjunction with specific features and embodiments, it is obvious that various modifications and combinations can be made thereto without departing from the spirit and scope of this application. Accordingly, this specification and drawings are merely exemplary illustrations of this application as defined by the appended claims, and are considered to cover any and all modifications, variations, combinations, or equivalents within the scope of this application. Clearly, those skilled in the art can make various alterations and modifications to this application without departing from the spirit and scope of this application. Thus, if such modifications and modifications of this application fall within the scope of the claims of this application and their equivalents, this application is also intended to include such modifications and modifications.

Claims

1. An intelligent scheduling and monitoring system for construction waste transfer based on the Internet of Things, characterized in that, include: IoT sensing modules, edge processing modules, and cloud scheduling platforms; The IoT sensing module includes a stacking status sensing unit and a transportation status sensing unit. The stacking status sensing unit is used to acquire the spatial morphological and mass characteristics of the stacked objects in the target area, and the transportation status sensing unit is used to acquire the operating status characteristics of the transfer vehicles and the transportation environment characteristics. The edge processing module is communicatively connected to the IoT sensing module and is used to preprocess various sensing features and upload them to the cloud scheduling platform. The cloud-based scheduling platform is used to perform cross-validation based on the preprocessed spatial morphological features and quality features to generate stacking status data of the target area, and to trigger a scheduling task when the stacking status data meets the preset transfer conditions. An initial scheduling scheme is generated based on the aforementioned scheduling task and multi-objective optimization model and then sent to the vehicle-mounted execution terminal. The cloud-based scheduling platform is also used to identify the compliance and regulatory status of the transfer vehicle based on the preprocessed transportation environment characteristics and the operating status characteristics, convert the compliance and regulatory status into dynamic constraint parameters and input them into the multi-objective optimization model, dynamically adjust the initial scheduling scheme, generate a reconstructed scheduling scheme and send it to the vehicle-mounted execution terminal.

2. The system according to claim 1, characterized in that, The system also includes a monitoring terminal; the monitoring terminal is used to simultaneously display the backlog status data, the compliance monitoring status, and the scheduling scheme execution data.

3. The system according to claim 1, characterized in that, The stacking state sensing unit includes a first detection subunit and a second detection subunit; The first detection subunit is used to scan the surface contour of the accumulation to generate point cloud data, and calculate the volume and height data of the accumulation as spatial morphological features based on the point cloud data; The second detection subunit is used to collect the force data of the bearing surface of the accumulation, and calculate the weight data of the accumulation as a quality feature based on the force data and the preset gravity mapping relationship.

4. The system according to claim 1, characterized in that, The edge processing module has a built-in data verification engine and a local cache queue; The data verification engine is used to perform time-series alignment and outlier filtering on the received feature data to generate valid perception data. The local cache queue is used to store valid sensing data when the communication link is interrupted, and asynchronously synchronizes it to the cloud scheduling platform in timestamp order after the communication link is restored.

5. The system according to claim 1, characterized in that, The step of converting the compliance and regulatory status into dynamic constraint parameters and inputting them into the multi-objective optimization model to dynamically adjust the initial scheduling scheme includes: The compliance and regulatory status includes environmental compliance status and trajectory compliance status; When a violation event is detected in the environmental compliance status or the trajectory compliance status, the violation type and severity corresponding to the violation event are determined. Based on the violation type and the violation severity, calculate the fitness penalty factor for the transfer vehicle or the current driving route, and use the fitness penalty factor as the dynamic constraint parameter; The fitness penalty factor is introduced into the fitness function of the multi-objective optimization model to reduce the overall fitness score of the transfer vehicle or the current driving route that has the violation event. Based on the updated fitness function, candidate vehicles and candidate paths are re-evaluated, and the candidate vehicle and candidate path with the highest comprehensive fitness score are selected as the reconfiguration scheduling scheme.

6. The system according to claim 5, characterized in that, The fitness function of the multi-objective optimization model is used to calculate the comprehensive fitness score of candidate vehicles and candidate routes based on the transfer distance evaluation function, the capacity utilization evaluation function, the transportation time evaluation function, and the fitness penalty factor corresponding to each violation event; wherein, the fitness penalty factor is calculated based on the violation type and the violation severity through nonlinear mapping.

7. The system according to claim 2, characterized in that, The monitoring terminals include construction party terminals, transportation company terminals, and urban management monitoring terminals; The construction terminal is used to query the stacking status data and scheduling plan execution data of the target area; The transportation company terminal is used to query the operational status characteristics of the transfer vehicles and the compliance supervision status. The urban management monitoring terminal is used to query data across the entire region and to conduct source tracing and supervision.

8. The system according to claim 3, characterized in that, The preset transfer conditions include: the accumulated volume of the accumulated material in the target area reaches a preset volume threshold, or the accumulation time exceeds a preset time threshold.

9. The system according to claim 5, characterized in that, Determining the type and severity of the violation corresponding to the violation event includes: Based on the sensor data of the violation, the violation type corresponding to the violation is identified; the violation includes: vehicle speeding, deviation from the preset route, unsealed cargo box, garbage spillage, illegal parking, and overloading. Query the preset violation event mapping table, which records the correspondence between various violation types and severity quantification values; The identified violation type is matched with the violation event mapping table to obtain the corresponding severity quantification value.

10. A method for intelligent scheduling and monitoring of construction waste transfer based on the Internet of Things, applied to the system described in any one of claims 1-9, characterized in that, The method includes: Acquire the spatial morphological and qualitative characteristics of the accumulated materials in the target area, as well as the operational status and transportation environment characteristics of the transfer vehicles; Based on the spatial morphological features and the quality features, cross-validation is performed to generate the stacking state data of the target area; When the accumulated data meets the preset transfer conditions, a scheduling task is triggered, and an initial scheduling scheme is generated and issued based on the scheduling task and the multi-objective optimization model. Based on the transportation environment characteristics and the operational status characteristics, the compliance and regulatory status of the transfer vehicle is identified; The compliance and regulatory status is transformed into dynamic constraint parameters and input into the multi-objective optimization model to dynamically adjust the initial scheduling scheme, generate a reconstructed scheduling scheme, and issue it.

Images