A substation engineering waste emission reduction scheduling method and system executed by a server

Abstract

The present application belongs to the technical field of resource scheduling, and relates to a substation engineering waste emission reduction scheduling method and system executed by a server, comprising the following steps: a stripping component node generates a virtual waste pulse data packet containing an expected quality and a spatial coordinate matrix; the connection edge transportation cost weight of a bottom layer routing topology graph is initialized to an infinite large closed value; the virtual traffic lock is pressed into a to-be-processed scheduling queue and configured on the connection edge transportation cost weight; a physical impact increment signature is fused; the infinite large closed value is rewritten as a real carbon emission increment value in response to a decryption confirmation instruction; a unique vehicle matching scheduling sequence and a corresponding clean-up route graph are generated, and a closed-loop scheduling instruction is issued; and a spatial coordinate matrix instruction monitoring probe is extracted to turn and investigate. The present application solves the problems of untimely clean-up scheduling, resource mismatching and inaccurate carbon emission accounting caused by the lack of reliable synchronous verification mechanism between virtual prediction data of construction waste and physical disposal events.

Description

Technical Field

[0001] This invention belongs to the technical field of resource scheduling, and relates to a server-executed method and system for scheduling waste reduction in substation projects. Background Technology

[0002] In the management of large-scale construction projects such as substations, the removal and disposal of construction waste is a continuous and complex process. The core challenge currently lies in achieving efficient and precise coordination between waste generation and waste removal scheduling to reduce resource waste and unnecessary carbon emissions. Waste generation at construction sites is random and sudden in both time and space. Traditional management methods struggle to obtain accurate real-time information on waste generation, including its quality, material type, and specific time and location of generation, leading to delayed and unpredictable scheduling decisions.

[0003] To address this issue, the industry has explored various solutions. A common approach is periodic, scheduled collection based on fixed times or routes, where collection vehicles process all waste at designated collection points according to a pre-set schedule. Another approach is a passive response model based on manual inspection and communication reporting. On-site managers visually assess the loading status of waste bins and, when deeming collection necessary, notify the dispatch center via telephone or instant messaging. Some improvements incorporate IoT technology, such as installing ultrasonic or infrared sensors inside waste bins to monitor material levels. When a threshold is reached, an alarm is automatically triggered to initiate dispatch. Building Information Modeling (BIM) technology is also used during the project planning phase to predict the total amount and type of waste generated throughout its lifecycle, providing data support for macro-level waste management strategies.

[0004] The aforementioned traditional methods have certain limitations in dealing with complex and dynamic construction environments. Fixed-cycle collection methods may lead to ineffective transport of empty or partially full containers, or untimely removal of overflowing containers, when waste generation is uneven. Both economic and environmental aspects need improvement. The manual reporting model suffers from significant response delays and subjective judgment criteria, easily influenced by personnel responsibility and on-site communication conditions, making it unreliable for refined scheduling. While using a single material level sensor improves automation, it only reflects volume occupancy and cannot obtain actual waste mass and material information. For waste of different densities, the same material level may correspond to vastly different weights and transportation carbon footprints, and it is susceptible to misjudgments due to irregular waste shapes. While macro-prediction methods based on Building Information Modeling (BIM) provide theoretical data at the planning level, they lack closed-loop verification with real-time physical operations at the execution level. They cannot capture the instantaneous correspondence between construction activities and actual waste generation, leading to a disconnect between planning and reality.

[0005] Based on the above problems, the present invention aims to solve the problems of untimely waste disposal scheduling, resource misallocation, and inaccurate carbon emission accounting caused by the lack of a reliable synchronous verification mechanism between virtual prediction data of construction waste and physical disposal events. Summary of the Invention

[0006] In a first aspect, the present invention provides a server-executed method for scheduling waste reduction in substation engineering, comprising the following steps: S1. Extract the physical and electrical characteristic edge of the transient current signal, and generate a trigger command based on the physical and electrical characteristic edge. Strip the component node to generate a virtual waste pulse data packet containing the expected mass and spatial coordinate matrix. S2. Construct the underlying routing topology graph and initialize the transport cost weights of the connecting edges in the underlying routing topology graph to infinite locking values. S3. Calculate the expected quality and preset material attribute data to generate a virtual access lock. Push the virtual access lock into the pending scheduling queue and configure it in the transportation cost weight of the connection edge. S4. Extract the mechanical momentum pulse and height change rate data of the physical waste collection point and fuse them to form a physical impact incremental signature; S5. Drive the physical impact incremental signature to initiate a matching verification request to the virtual access lock, extract the expected quality anchored by the virtual access lock and compare it with the real physical characteristics with tolerance, and respond to the verification by rewriting the infinite locking value to the real carbon emission increment value through the instruction. S6. Perform optimization calculations on the underlying routing topology map that rewrites the actual carbon emission increment values, generate a unique vehicle matching scheduling sequence and corresponding waste disposal route map, and issue closed-loop scheduling instructions. S7. In response to an abnormal scenario where the pending scheduling queue crosses the preset safety setting line and no physical impact incremental signature is received, extract the spatial coordinate matrix and instruct the monitoring probe to turn around for reconnaissance.

[0007] A further aspect of the present invention involves stripping component nodes to generate a virtual waste pulse data packet containing the expected mass and spatial coordinate matrix, comprising the following steps: Calculate the first-order time derivative of a transient current signal to capture the rate of change; In response to a step drop in the amplitude of a transient current signal and its stabilization at a level below a preset no-load threshold, the specific signal pattern is identified as a physical electrical characteristic. Based on physical and electrical characteristics, a trigger command carrying the current timestamp and device identifier is generated; Based on the equipment identifier, the pre-set construction task database is retrieved to locate component nodes, the three-dimensional geometric model is extracted to calculate the volume, and the material density is retrieved. The expected mass is calculated by multiplying the material density by the volume, and the position and orientation of the component nodes in the global coordinate system of the project are extracted and encoded into a spatial coordinate matrix. The expected quality, spatial coordinate matrix, timestamp, and device identifier are encapsulated into a virtual waste pulse data packet.

[0008] A further aspect of this invention involves initializing the transport cost weights of the connecting edges in the underlying routing topology graph to an infinite latching value, including the following steps: Load the preset engineering site layout data, which defines the geographical coordinates of the physical waste collection points and the coordinates of the waiting positions of the collection vehicles; Instantiate a directed graph data structure in memory to generate the underlying routing topology graph, map physical waste collection points as independent nodes, and create the standby positions of collection vehicles as source nodes; Find the directed edges from the source node to the independent nodes and generate the transport cost weights of the connecting edges; The transport cost weight of the connecting edge is assigned a predefined maximum floating-point number to generate an infinite locking value. The infinite locking value is used to block the optimization path calculation for independent nodes.

[0009] A further aspect of this invention involves pushing the virtual access lock into the pending scheduling queue and configuring it on the transport cost weight of the connection edge, including the following steps: The expected quality value is converted into a string format and concatenated with the preset material attribute data to form the source information string; Input the source information string into a preset one-way hash function to perform an encrypted hash operation, and generate a fixed-length hexadecimal hash value as a virtual access lock; Send the virtual access lock to the pending scheduling queue that follows the first-in-first-out principle; The physical waste collection point is determined by querying a preset mapping table based on the internal component identifier of the virtual waste pulse data packet, and the virtual access lock is logically bound to the transportation cost weight of the connection edge pointing to the physical waste collection point.

[0010] A further aspect of the present invention involves fusing and constructing a physical impact incremental signature, comprising the following steps: In response to the virtual access lock being pushed into the pending scheduling queue, a low-level control command is sent to the embedded control unit inside the physical waste collection point; Activate the piezoelectric sensor device that is mechanically coupled to the bottom plate of the collection box and the ultrasonic array device installed on the top of the inner wall of the collection box; Mechanical momentum pulses are generated by collecting transient voltage signals from the impact of falling waste materials using piezoelectric sensors. The height value is converted from the distance to the surface of the waste accumulation in the box by continuously measuring the distance using an ultrasonic array device. The height change rate data is obtained by calculating the ratio of the difference between two adjacent height measurements to the sampling time interval. Extract the peak voltage and voltage integral of the mechanical momentum pulse, and align and package the peak voltage, voltage integral, height change rate data with the timestamp to generate a physical impact incremental signature.

[0011] A further aspect of this invention involves rewriting the infinite latch value to the actual carbon emission increment value in response to a decryption confirmation command, comprising the following steps: Extract the virtual access lock and associated virtual waste pulse data packet from the head of the pending scheduling queue; The mechanical momentum pulse feature value and height change rate data in the physical impact incremental signature are input into the preset physical feature deduction model and mapped to generate a mass estimate as the real physical feature. Calculate the absolute difference between the actual quality estimate and the expected quality, and determine whether the absolute difference is within the preset quality error tolerance threshold. Calculate the difference between the generation timestamp of the physical impact incremental signature and the generation timestamp of the virtual access lock, and determine whether the time difference falls within the dynamic timing matching window extracted based on the construction flow rhythm. If both the absolute difference and the time difference meet the conditions, a verification pass instruction is generated. Based on the verification pass instruction, the infinite lock value on the transport cost weight of the connection edge is broken, and the real physical characteristics are input into the preset carbon emission model to generate the real carbon emission increment value and cover the infinite lock value.

[0012] A further aspect of this invention involves generating a unique vehicle matching and scheduling sequence and a corresponding waste collection route map, and issuing closed-loop scheduling instructions, including the following steps: The preset dynamic path planning module is invoked to receive the underlying routing topology map that rewrites the actual carbon emission increment value; Using the minimization of total carbon emissions as the objective function, a heuristic optimization algorithm is performed on the underlying routing topology graph to achieve global convergence and solution analysis. The lowest total cost travel path is selected through convergent solution analysis, and a unique vehicle matching scheduling sequence containing the target waste collection vehicle identifier and access order is output. The system calls a preset geographic information system service to generate a corresponding waste disposal route map based on the geographic coordinates in the unique vehicle matching and scheduling sequence. The unique vehicle matching and scheduling sequence and the corresponding waste collection route map are formatted and encapsulated to generate a closed-loop scheduling instruction, which is then sent to the on-board information terminal of the target waste collection vehicle through the communication network relay link.

[0013] A further aspect of this invention involves extracting the spatial coordinate matrix to instruct the monitoring probe to turn and reconnoiter, including the following steps: Set an independent timeout timer for each virtual access lock in the pending scheduling queue; Continuously count the cumulative stacked number of virtual access locks in the pending scheduling queue and compare the cumulative stacked number with the preset security setting threshold; In response to the pending scheduling queue crossing the preset security setting threshold, determine whether a decryption confirmation command has been received within the preset setting timing matching window; When the accumulated stacking quantity exceeds the preset safety threshold and the independent timeout timer expires, an abnormal scenario is determined to be triggered. Mark the status of independent nodes in the underlying routing topology graph corresponding to abnormal scenarios as unreachable to cut off channel connections; Extract the spatial coordinate matrix from the virtual waste pulse data packet, and issue control commands containing the spatial coordinate matrix to drive the high-altitude monitoring pan-tilt probe to adjust the horizontal and vertical angles to align with the target coordinates and perform optical zoom reconnaissance.

[0014] A further aspect of this invention involves inputting real physical characteristics into a preset carbon emission model to generate real carbon emission increment values ​​and covering an infinite latching value, including the following steps: Extract the actual carbon emission increment value output by the preset carbon emission model, and query the preset carbon quota database to obtain the remaining carbon emission quota for the current substation project; Calculate the difference between the remaining carbon emission allowance and the actual increase in carbon emissions, and determine whether the difference is lower than the preset carbon emission warning threshold. In response to the difference being lower than the preset carbon emission warning threshold, a dynamic carbon emission limiting strategy is generated. The dynamic carbon emission throttling strategy is fed back to the underlying routing topology, and the infinite latching value corresponding to the transportation cost weight of the unconfirmed connection edge is dynamically increased according to a preset ratio.

[0015] Secondly, the present invention provides a server-executed substation engineering waste reduction scheduling system, comprising the following modules: The virtual waste generation module is used to extract the physical and electrical characteristic edge of the transient current signal, and generate a trigger command based on the physical and electrical characteristic edge to strip the component node and generate a virtual waste pulse data packet containing the expected mass and spatial coordinate matrix. The routing environment initialization module is used to construct the underlying routing topology graph and initialize the transport cost weights of the connecting edges in the underlying routing topology graph to infinite latching values. The dynamic constraint generation module is used to calculate and generate virtual access locks based on the expected quality and preset material attribute data, push the virtual access locks into the pending scheduling queue and configure them on the transportation cost weight of the connection edge; The physical event perception module is used to extract mechanical momentum pulses and height change rate data from physical waste collection points and fuse them to form a physical impact incremental signature. The spatiotemporal coupling verification module is used to drive the physical impact incremental signature to initiate a matching verification request to the virtual access lock, extract the expected quality anchored by the virtual access lock and compare it with the real physical characteristics with tolerance, and respond to the verification by rewriting the infinite locking value to the real carbon emission increment value through the instruction. The carbon-optimal scheduling module is used to perform optimization calculations on the underlying routing topology map that rewrites the actual carbon emission increment values, generate a unique vehicle matching scheduling sequence and corresponding waste disposal route map, and issue closed-loop scheduling instructions. The anomaly monitoring and response module is used to respond to anomalies where the pending scheduling queue crosses the preset safety threshold and no physical impact incremental signature is received. It extracts the spatial coordinate matrix to instruct the monitoring probe to turn and investigate.

[0016] In summary, the present invention has the following beneficial technical effects: 1. By linking changes in the physical characteristics of construction equipment with a building information model (BIM), the system can capture the source events of waste generation in real time. When the transient current characteristics of construction equipment completing effective operations are detected, the system automatically extracts the corresponding component from the digital model and generates a virtual waste pulse data packet containing the expected mass and spatial location. This mechanism binds the abstract construction plan with specific physical actions at the timestamp level, achieving instant digitization of waste generation events. The system can obtain information about the moment of waste generation, providing a real-time and forward-looking data foundation for subsequent scheduling decisions that is not available through traditional manual reporting or periodic inspections, thereby eliminating management lag caused by information delays.

[0017] 2. By constructing a dual-locking scheduling access mechanism, blind vehicle dispatching and ineffective transportation can be effectively avoided. The system initializes the transportation cost of all collection routes to an infinite lockout value, forming a logical transportation barrier. Subsequently, an encrypted virtual access lock is generated for each virtual waste pulse data packet, and it is matched and verified against the physical waste collision signal. Only when the virtual prediction and physical reality perfectly match will the decryption confirmation command be triggered, and the system will rewrite the infinite lockout value to the actual routing cost. This mechanism ensures that the dispatching instructions for collection vehicles must be based on the confirmed collection demand in the physical world. Therefore, it reliably blocks ineffective dispatching of empty containers or unconfirmed waste, improves vehicle utilization efficiency, and reduces fuel consumption and carbon emissions caused by ineffective driving.

[0018] 3. By introducing dynamic carbon emission costs based on real physical characteristics as the core indicator for path optimization, low-carbon-oriented waste disposal scheduling can be achieved. After successful matching between virtual and real environments, the system does not use fixed distance or time as the transportation cost. Instead, it dynamically calculates and updates the real carbon emission increment generated by transporting the waste based on the real physical characteristics of the waste inferred from sensors, using this as the sole routing cost. The optimization engine then performs a global solution to generate the scheduling sequence with the lowest total carbon emissions. This mechanism ensures that every scheduling decision is directly linked to carbon reduction targets. The scheduling result no longer simply pursues the shortest path or the fastest time, but rather selects the combination that minimizes the overall system's carbon emission increment from multiple potential waste disposal tasks, realizing a shift from operational efficiency optimization to vectorized environmental benefit optimization. Attached Figure Description

[0019] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the accompanying drawings used in the description of the embodiments or the prior art will be briefly introduced below. The drawings are used to provide a further understanding of the present invention.

[0020] Figure 1 A flowchart illustrating an embodiment of this application is disclosed.

[0021] Figure 2 Structural schematic diagrams of embodiments of this application are disclosed. Detailed Implementation

[0022] The following is in conjunction with the appendix Figure 1 - Figure 2 A preferred description of the present invention is provided below.

[0023] See attached document Figure 1 This invention proposes a server-executed method for scheduling waste reduction in substation engineering, comprising the following steps: S1. Extract the physical and electrical characteristic edge of the transient current signal, and generate a trigger command based on the physical and electrical characteristic edge. Strip the component node to generate a virtual waste pulse data packet containing the expected mass and spatial coordinate matrix. S2. Construct the underlying routing topology graph and initialize the transport cost weights of the connecting edges in the underlying routing topology graph to infinite locking values. S3. Calculate the expected quality and preset material attribute data to generate a virtual access lock. Push the virtual access lock into the pending scheduling queue and configure it in the transportation cost weight of the connection edge. S4. Extract the mechanical momentum pulse and height change rate data of the physical waste collection point and fuse them to form a physical impact incremental signature; S5. Drive the physical impact incremental signature to initiate a matching verification request to the virtual access lock, extract the expected quality anchored by the virtual access lock and compare it with the real physical characteristics with tolerance, and respond to the verification by rewriting the infinite locking value to the real carbon emission increment value through the instruction. S6. Perform optimization calculations on the underlying routing topology map that rewrites the actual carbon emission increment values, generate a unique vehicle matching scheduling sequence and corresponding waste disposal route map, and issue closed-loop scheduling instructions. S7. In response to an abnormal scenario where the pending scheduling queue crosses the preset safety setting line and no physical impact incremental signature is received, extract the spatial coordinate matrix and instruct the monitoring probe to turn around for reconnaissance.

[0024] In one embodiment of the present invention, step S1, acquiring physical characteristic change source data of construction equipment, stripping model component information, and generating virtual waste pulse data packets with spatiotemporal attributes, includes the following steps: Acquire the transient current signal continuously transmitted from the power supply terminal of the on-site substation construction equipment; monitor the physical electrical characteristic edge when the transient current signal drops sharply and falls below the preset no-load threshold, and extract the hardware-level interrupt trigger instruction based on the physical electrical characteristic edge; input the hardware-level interrupt trigger instruction into the building information parsing engine to strip the component nodes of the corresponding construction equipment, and then generate a virtual waste pulse data packet containing the expected quality and spatial coordinate matrix.

[0025] Specifically, step S1 in this embodiment is configured as a data tracing and virtualization encapsulation process originating from the physical behavior of construction, executed by an edge computing gateway deployed at the substation construction site. The edge computing gateway establishes a hard-wired connection or near-field wireless communication link with the power control cabinet of a specific construction device, such as a hydraulic shear or plasma cutter, to access and read the current transformer installed on its main power supply circuit. The current transformer continuously measures the dynamic current flowing through the construction device at a preset sampling frequency, for example, one thousand times per second, and converts the collected analog signal into a digitized time-series data stream through an analog-to-digital converter, forming a transient current signal. The signal processing unit built into the edge computing gateway analyzes the transient current signal in real time. This unit applies, for example, a Kalman filter algorithm to denoise and smooth the original signal, and calculates the first-order time derivative of the signal to capture its rate of change.

[0026] When the unit detects that the amplitude of the transient current signal has shifted from a stable high-power operating range, such as from the current value representing a full-load cutting state... Within a very short time window, a sharp, step-like drop occurred, eventually stabilizing below the preset no-load threshold. When the signal pattern reaches a certain level, the edge computing gateway identifies it as a physical electrical characteristic edge marking the end of a valid work cycle. Based on the successful capture of this physical electrical characteristic edge, the edge computing gateway generates a hardware-level interrupt trigger command carrying the current timestamp and the unique device identifier of the construction equipment.

[0027] A hardware-level interrupt trigger command is submitted as a request to the Building Information Modeling (BIM) parsing engine running on the central server. Based on the received device identifier, the engine retrieves a pre-configured construction task database, which is strictly associated with the specific BIM components that each device is responsible for dismantling or processing within a specific time period. The BIM parsing engine then locates the component node corresponding to the command, such as a specific steel beam or concrete block. The engine extracts the component node from the BIM database and updates its logical status from "existing" to "demolished." The engine then extracts its 3D geometric model from the component node's attribute data to calculate its volume. And retrieve its material properties to obtain the material density. The expected quality is calculated using the following formula. :

[0028] The engine extracts the position and orientation of the component node in the global coordinate system just before it is stripped from the geometric information of the node, and encodes it into a 4x4 homogeneous transformation matrix, which is the spatial coordinate matrix. The building information parsing engine encapsulates the calculated expected quality, the extracted spatial coordinate matrix, and metadata such as the timestamp that triggered the process and the device identifier into a structured data object, thereby generating a virtual waste pulse data packet with spatiotemporal attributes, and pushing it to the subsequent processing queue.

[0029] Transient current signals refer to time series of current values ​​that reflect the transition of construction equipment load status from full-load operation to idle standby. Preset no-load threshold. This is a key parameter for distinguishing whether equipment is in an effective working state. The value is set based on multiple measured statistics of current consumption of a specific model of equipment in standby mode. Assuming the application scenario is a hydraulic shear with a rated power of 150kW, its average current in the non-working but energized state is 3.5A. To ensure robustness, three times the standard deviation can be added to the statistical mean as the final threshold, for example, set to 5A. The physical electrical characteristics represent the dynamic process of the current rapidly changing from a high load value to an no-load value. Its identification conditions not only include the current value being lower than the preset no-load threshold, but also include the absolute value of the current change rate being greater than a certain preset slope threshold, in order to exclude interference from human operation such as slow shutdown.

[0030] The Building Information Modeling (BIM) parsing engine is a software service module deeply integrated with the 4D BIM database. It has the ability to query, modify, and extract the geometric and physical properties of model components through an application programming interface (API). The expected quality is the theoretical waste quality calculated based on the model data. The spatial coordinate matrix is ​​a standard 4x4 homogeneous transformation matrix used to uniquely determine the position and orientation of an object in three-dimensional space.

[0031] For example, suppose that at a certain moment, a hydraulic shear with device ID LC-02 deployed in substation area 2 is performing a cutting and dismantling task on an H-beam of model HW200x200. This task has been pre-associated with device LC-02 in the building information model. A current transformer installed on the LC-02 power line continuously reports transient current signals. At timestamp... The amplitude of the transient current signal drops from 157A to a stable value close to 0A, which is lower than the preset no-load threshold of 5A set for the device. The edge computing gateway captures this physical electrical characteristic edge and generates a hardware-level interrupt trigger instruction with the content of {Device ID: "LC-02", timestamp: "2025-10-26T10:30:15.120Z"}.

[0032] A hardware-level interrupt was triggered and sent to the building information parsing engine. The engine queried the database and learned that the current work object of LC-02 was an H-beam with component number BEAM-HW200-087. The engine then stripped the component's nodes and retrieved the material from its properties as Q345B steel with a density of... The volume is calculated to be 7850 kg / m³ using a geometric model. The mass is 0.04 m³. The engine performs calculations to obtain the expected mass. The engine extracts the spatial coordinate matrix of the steel beam in the engineering coordinate system, for example, a matrix containing rotation and translation information. The engine generates a matrix containing: The virtual waste pulse data packet contains fields such as {timestamp: "2025-10-26T10:30:15.120Z", device ID: "LC-02", component ID: "BEAM-HW200-087", expected mass: 314kg, spatial coordinate matrix: [[...], [...], [...], [...]]}.

[0033] In one embodiment of the present invention, step S2, constructing an initial environment structure for waste network assignment and configuring the transport algebra attribute to an infinite latch value, includes the following steps: Construct a low-level routing topology graph containing each physical waste collection point as an independent node to represent the scheduling space architecture of the engineering site; initialize all the transportation cost weights of the connecting edges of each independent node in the low-level routing topology graph pointing to the waiting position of the waste collection vehicle and assign them to an infinite locking value; use the infinite locking value to block the optimization path calculation action for each independent node, thereby freezing the dispatch time window of the corresponding physical waste collection point at the system operation level to prevent blind dispatch.

[0034] Specifically, the initial closed logical environment is designed to build for subsequent dynamic scheduling decisions. This process is executed entirely in the server's memory and initialized by the scheduling control module. The scheduling control module loads pre-configured site layout data, which defines the geographic coordinates of all physical waste collection points and the coordinates of the waiting positions of collection vehicles in a vectorized format. Based on this data, the scheduling control module calls graph theory libraries, such as Python's NetworkX library or C++'s BoostGraphLibrary, to instantiate a directed graph data structure in memory. This is the underlying routing topology graph containing each physical waste collection point as an independent node. In this graph, each physical waste collection point is mapped to an independent vertex or node object, and the waiting position of the collection vehicle is also created as a dedicated source node. This constitutes an abstract representation of the scheduling space architecture of the site. The scheduling control module then enters the weight initialization phase, traversing all nodes in the graph and identifying those nodes that represent the individual nodes.

[0035] For each node of this type, the module programmatically searches or creates directed edges in the graph from the source node at the vehicle's waiting position to that individual node. These edges represent potential transport paths from the waiting area to the collection point. After creating or accessing these edges, the scheduling control module initializes its core attribute—the transport cost weight of the edge—and assigns it a predefined infinite latching value. This is not mathematical infinity, but rather a very large floating-point number or integer within the range of computer numerical representations, such as the largest double-precision floating-point number the system can represent. This assignment ensures that the initial path cost from the vehicle's waiting position to any waste collection point is set to a value unacceptable in normal optimization calculations. The direct consequence of this initialization is that any subsequent path planning algorithm, such as Dijkstra's algorithm, will be affected. When the algorithm is invoked to find the optimal path from the waiting point, it fails to converge to any valid solution because the initial cost of all potential paths is an infinite locking value.

[0036] In this way, the system uses an infinite latch value to build a logical barrier at the algorithm level, effectively blocking the optimization path calculation for each independent node. This freezes the dispatch time window of the corresponding physical waste collection point at the system operation level, and thus effectively avoids blind dispatching behavior in the absence of effective physical trigger signals at the system mechanism level.

[0037] The underlying routing topology graph is a mathematical model used to describe the connections and travel costs between different locations in physical space. It is typically stored in computer memory as an adjacency matrix or adjacency list. Each independent node is a discrete unit representing a specific physical location in this topology graph; in this case, it specifically refers to the abstract representation of each waste collection bin. The standby location for collection vehicles is a special node in the topology graph, representing the parking area and dispatch starting point for the collection fleet when it is idle. The transport cost weight of the connecting edge is a numerical value assigned to each edge in the graph, used to quantify the cost or degree of obstruction when moving from one node to another. In this step, this weight is used as a logical switch rather than a measure of physical cost.

[0038] The infinite latch value is a numerically large constant. The criterion for setting it is that it must be much larger than the sum of any actual routing cost metrics that appear in subsequent steps. For example, it could be set to... Its function is to mark certain paths as "unfeasible" or "too costly to consider" in the optimization algorithm, thereby achieving logical path blocking.

[0039] For example, suppose a substation project site has three physical waste collection points, named P1, P2, and P3, and a waiting position for transport vehicles, named D0. When the scheduling and control module in the server starts, it constructs a low-level routing topology graph G containing four nodes, with a node set V={D0, P1, P2, P3}. During the initialization phase, the module identifies the directed edges connecting the vehicle waiting point to each collection point, namely E1=(D0, P1), E2=(D0, P2), and E3=(D0, P3). When setting the transport cost weights for these connecting edges, the scheduling and control module assigns an infinite latch value, for example, 999999999, to the weight attributes of these edges sequentially. After initialization, the path cost is set to Cost(D0, P1) = 999999999, Cost(D0, P2) = 999999999, Cost(D0, P3) = 999999999.

[0040] Upon receiving a fuzzy scheduling request, the system attempts to find the most economical waste disposal route starting from point D0. Any standard optimization engine analyzing this topology will be unable to generate any effective scheduling sequence because the direct costs to P1, P2, and P3 are all locked at extremely high levels by infinite latching values. This logically freezes the dispatching of vehicles to all collection points, ensuring that vehicles are not incorrectly assigned until the actual waste disposal demand is confirmed in subsequent steps. In this state, there has been no interaction with the virtual waste pulse data packet generated in the previous step S1; the data packet is currently in the processing queue.

[0041] In one embodiment of the present invention, step S3, extracting model attribute data and using a hash algorithm to generate a virtual access lock for a dynamically constrained task sequence, includes the following steps: Extract the expected quality and material attribute data of the stripped component nodes contained in the virtual waste pulse data packet; perform a message digest combination operation on the expected quality and material attribute data using a one-way hash function to generate a virtual access lock in a dynamically encrypted state; push the virtual access lock to the server's pending scheduling queue, and float the virtual access lock on the independent node connection line corresponding to the underlying routing topology to form a physical cleaning access barrier.

[0042] Specifically, step S3 is the process of encrypting and deploying the virtual waste pulse data packet generated in step S1. This process is executed in the scheduling and control module of the central server. The scheduling and control module receives the virtual waste pulse data packet from step S1 from the input data stream. The module then accesses the internal data structure of the data packet and extracts two key fields: the expected quality and the material attribute data of the stripped component node in the building information model. The material attribute data is usually a text string representing the material type. The module initiates a message digest combination operation. This operation follows a strict data assembly and encryption protocol, converting the expected quality value into a string format without loss and performing a deterministic concatenation operation with the material attribute data string to form a unique source information string. This source information string is used as input and fed into a pre-configured one-way hash function, such as the secure hash algorithm SHA-256. This function performs an encrypted hash operation on the source information string, generating a fixed-length and irreversible hexadecimal hash value. This hash value is defined as a virtual access lock in a dynamically encrypted state.

[0043] After generation, the scheduling control module performs a dual deployment action. On one hand, it sends the newly generated virtual access lock to a dedicated FIFO (First-In, First-Out) queue within the server, allowing it to queue in the order of generation for subsequent matching and verification. On the other hand, the module queries a preset mapping table based on the spatial coordinate matrix or internal component ID contained in the virtual waste pulse data packet to determine the physical waste collection point to which the component belongs. After locating the independent node corresponding to the collection point in the underlying routing topology map constructed in step S2, the module logically binds the virtual access lock as key metadata to the connection edge pointing to that independent node. After this binding operation is completed, the virtual access lock is floated on the connection line of the corresponding independent node in the underlying routing topology map, thus formally forming a physical clearance access barrier. The existence of this barrier prevents any scheduling of the node before obtaining physical confirmation in the next step from being blocked by the underlying algorithm logic.

[0044] The information digest combination operation here is formally represented as follows, where the product is... Represents a virtual access lock: , It is a virtual access lock, which is a unique digital identifier generated by combining information digests and calculations. It is used to "lock" the physical clearance access barrier in the underlying routing architecture. It is a one-way hash function, representing the process of performing cryptographic hash operations. It uses industry-standard algorithms such as SHA-256 or SHA-3 to convert the input source information string into a fixed-length hash value. This refers to string conversion functions, which convert numerical values ​​into string formats without loss of quality. This refers to the expected mass, which is the predicted mass value of the waste extracted from the data packet. For example, the expected mass value of 314 mentioned in the example scenario is in kg. It is a string concatenation operator; It is material attribute data, which is a standardized text information describing the material composition and specifications of components, derived from the text strings representing material types provided by the Building Information Model.

[0045] Material attribute data refers to standardized textual information extracted from Building Information Modeling (BIM) that describes the material composition and specifications of building components, such as "Q345BSteel" or "C30Concrete". One-way hash functions are cryptographic functions that can map data of arbitrary length to a fixed-length hash value. They possess one-wayness, meaning the original data cannot be deduced from the hash value, and the avalanche effect, where a small change in input can lead to a large difference in output. In this invention, industry-standard algorithms such as SHA-256 or SHA-3 can be selected.

[0046] A virtual access lock is a unique digital credential generated by a one-way hash function, condensing the core physical attributes of a specific waste generation event in cryptographic form. The pending scheduling queue is a data structure in server memory that manages virtual access locks, ensuring the sequential nature of waste generation events is preserved. The physical waste removal access barrier is not a physical obstacle, but rather a software-based conditional lock, associated with the virtual access lock on a connecting edge in the underlying routing topology. Only when specific conditions are met, such as the access lock being correctly decrypted, can the infinite latch value on that connecting edge be modified, making the path available.

[0047] For example, the scheduling control module receives a virtual scrap pulse data packet generated in step S1, containing a expected mass of 314 kg and a source component of an H-beam. The module extracts the expected mass of 314 kg and the material attribute data "Q345BSteel" stored in the BIM. Next, the module converts the mass value into the string "314", concatenates it with the material string to form the source information string "314Q345BSteel", and calls the SHA-256 hash function to perform a calculation on this string, generating a 64-bit hexadecimal virtual access lock, for example: "e1b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855".

[0048] The virtual access lock "e1b0c442..." is pushed to the head of the pending scheduling queue. The system, by querying the mapping relationship between the construction area and the collection point, confirms that the H-shaped steel beam scrap belongs to node P1 in the underlying routing topology. The virtual access lock "e1b0c442..." is bound as a metadata attribute to the connection edge between nodes D0 and P1, initialized in step S2. This operation establishes a physical access barrier for waste collection on the path to the P1 scrap collection point, constructed from the virtual access lock. Unless the lock is unlocked by the correct key generated in subsequent steps, the transportation cost weight of the connection edge leading to P1 will remain an infinitely large locking value of 999999999, and the system will reject any dispatch requests to P1.

[0049] In one embodiment of the present invention, step S4, activating the leading edge sensing hardware and picking up the impact momentum signal of the base plate to generate a physical impact increment signature for cross-domain matching, includes the following steps: Send low-level control commands to activate piezoelectric sensors and ultrasonic array devices deployed inside physical waste collection points.

[0050] By combining sensing elements that continuously sample at high frequency, the mechanical momentum pulse and height change rate data generated instantaneously when waste falls and impacts the bottom plate of the collection box are extracted. Following the data assembly logic specifications, the mechanical momentum pulse and height change rate data are timestamped and dimensionally fused to form a physical impact incremental signature used to characterize the variation characteristics of real physical quantities.

[0051] Specifically, step S4 is a physical phenomenon perception and data encapsulation process remotely triggered by the central server and executed by the field edge devices. After the virtual access lock generated in step S3 is successfully pushed into the pending scheduling queue, the scheduling logic of the central server will identify the target independent node bound to the lock, and based on this information, send low-level control commands to the embedded control unit deployed inside the physical waste collection point corresponding to the node. The commands are sent through the IoT MQTT or CoAP protocol, and their core content is to wake up the sensor group of the collection point from low-power sleep mode to high-frequency sampling activation.

[0052] In response to this command, the embedded control unit synchronously activates the piezoelectric sensor mechanically coupled to the bottom plate of the collection box and the ultrasonic array device mounted on the top of the inner wall of the collection box. Once activated, the signal conditioning circuit of the piezoelectric sensor and the transmitter / receiver controller of the ultrasonic array device enter a continuous high-frequency operating state. In this state, the ultrasonic array device emits ultrasonic pulses at a frequency of, for example, 20Hz and detects their echo time, thereby continuously measuring the distance to the surface of the waste accumulation inside the box and calculating the height value. The sampling circuit of the piezoelectric sensor continuously performs analog-to-digital conversion on the weak voltage signal output by the sensor at a frequency of, for example, 20kHz. When waste falls from a height and impacts the bottom plate of the collection box, the resulting mechanical stress wave causes the piezoelectric sensor to instantaneously generate a voltage pulse signal, which is a mechanical momentum pulse.

[0053] The accumulation of waste inevitably leads to changes in the height measured by ultrasound. The embedded control unit calculates the ratio of the difference between two adjacent height measurements to the sampling time interval in real time, thereby obtaining the height change rate data. The processor in the embedded control unit then follows the preset data assembly logic specifications to fuse these two data streams from different sources with different sampling rates. First, it extracts features from the acquired mechanical momentum pulse waveform, such as calculating its peak voltage, voltage integral, and other key indicators.

[0054] The processor correlates the height change rate data sequences calculated within a very short time window before and after the pulse occurrence time, aligning them with timestamps synchronized by a network time protocol. The processor then packages the integrated timestamps, mechanical momentum pulse characteristic values, and height change rate data sequences into a compact binary or JSON format data packet. This data packet serves as the physical impact increment signature characterizing the actual physical quantity variation. The height change rate here can be approximated by the following formula:

[0055] Physical shock incremental signature It can be formally represented as a tuple containing multidimensional features:

[0056] In the formula, It is a moment The measured height, This refers to the time interval between two height measurements. It is the timestamp of the impact event. It is the peak voltage of the mechanical momentum pulse. It is the voltage integral of the pulse. It is the rate of change of altitude at the moment of impact.

[0057] Sending low-level control commands refers to the commands issued by the server to a designated IoT terminal to change its operating state, and their format follows a specific communication protocol. Piezoelectric sensors are typically elements that convert mechanical force or pressure into charge or voltage, such as PZT ceramic sheets, mounted on a base plate to sensitively capture vibrations generated by impacts. Ultrasonic array devices consist of multiple ultrasonic transducers; compared to single probes, array designs provide a wider detection coverage, and beamforming technology can suppress noise and more accurately measure the average height of irregularly piled surfaces. Mechanical momentum pulses are transient voltage signals output by piezoelectric sensors when impacted; their amplitude and waveform contain information related to the mass and velocity of the impacting object. The rate of change of height is the first derivative of the waste pile height with respect to time, directly reflecting the velocity of material accumulation. Physical impact incremental signatures are structured data packets that encapsulate the multi-dimensional physical characteristics generated by a real waste throwing event, serving as a "physical key" for subsequent verification.

[0058] For example, after the virtual access lock “e1b0c442…” is pushed into the queue and bound to node P1, the server sends an activation command to the embedded control unit deployed in the P1 collection box. Around timestamp 2025-10-26T10:30:18.450Z, the lifting equipment throws the aforementioned 314kg H-shaped steel beam into the P1 collection box. The piezoelectric sensor fixed to the bottom plate of the P1 box generates a strong voltage pulse at the moment of impact. After high-speed ADC sampling, the control unit analyzes the waveform and obtains a peak voltage of 3.8V and a voltage integral of 12.5V·ms. An ultrasonic array device operating at 50ms intervals monitors a sharp increase in the height of the waste surface from 1.21m to 1.45m within 100ms, calculating the average height change rate during this period as (1.45-1.21) / 0.1 = 2.4m / s.

[0059] The embedded control unit then executes the data assembly logic specification, encapsulating data such as the impact timestamp, peak voltage, voltage integral, and height change rate to generate a physical impact incremental signature data packet in the form of {timestamp: "2025-10-26T10:30:18.450Z", pulse_peak_voltage: 3.8, pulse_energy: 12.5, height_gradient: 2.4}, which is then prepared to be transmitted back to the solver module of the central server via the uplink.

[0060] In one embodiment of the present invention, step S5, which involves executing a spatiotemporal coupling comparison process and rewriting the penalty operator into a true routing cost indicator based on the decryption confirmation action, includes the following steps: The physical impact incremental signature is sent back to the solver module at the bottom layer of the scheduling center, and drives the physical impact incremental signature to attempt to decrypt the virtual access lock located at the head of the pending scheduling queue; multi-dimensional comparison verifies whether the real physical characteristics represented by the physical impact incremental signature and the expected quality embedded in the virtual access lock are mutually matched within the content difference range of the set timing matching window; in response to the decryption confirmation instruction that the real physical characteristics and the expected quality are completely matched, the infinite locking value on the transport cost weight of the connection edge is broken down and the penalty operator is rewritten as the real carbon emission increment value derived from the real physical characteristics.

[0061] Specifically, step S5 is a closed-loop confirmation and weight rewriting process executed on the central server to verify the consistency between physical events and virtual predictions. The physical impact increment signature generated in step S4 and returned from the field is transmitted back to the solver module in the scheduling center via a wireless communication link such as 4G or 5G. After receiving this data packet in its message queue, the module is driven to execute the spatiotemporal coupling comparison process. The first step of the process is to extract the currently pending virtual access lock from the head of the pending scheduling queue, and at the same time extract the original virtual waste pulse data packet generated in step S1 associated with the lock, so as to access the internally encapsulated expected quality and generation timestamp. The solver module deduces the real physical characteristics represented by the physical impact increment signature. The module inputs the feature values ​​of the mechanical momentum pulse and the height change rate data into a pre-trained physical characteristic deduction model based on multiple regression or shallow neural networks. This physical characteristic deduction model is established based on a large amount of calibration experimental data. Its function is to map the diverse sensor signals collected back to a single quality estimate, which is the real physical characteristic.

[0062] The module performs multi-dimensional comparison and verification, which includes two levels. The first is content matching: the module calculates the absolute difference between the predicted actual physical characteristics and the expected quality of the virtual access lock's embedded root cause, and determines whether this difference is within a preset content difference range. The second is temporal matching: the module calculates the difference between the generation timestamp of the physical impact incremental signature and the generation timestamp of the virtual access lock, and determines whether this time difference falls within a set temporal matching window, ensuring that the physical event occurs within a reasonable time after the virtual prediction. Only when both content and temporal dimension comparisons are successful—that is, when the physical facts and the virtual prediction match within the allowable error range—will the system generate a decryption confirmation command with a Boolean value of true.

[0063] In response to this instruction, the solver module will trigger subsequent breakdown and rewriting actions. Based on the node information bound to the decrypted virtual access lock, it will locate the specific connection edge in the underlying routing topology graph. The module will break down the infinite locking value set in step S2 on this connection edge and completely overwrite it. In its place, a newly calculated real carbon emission increment value will be used as the penalty operator. This value is derived from a dedicated carbon emission model based on the mass corresponding to the real physical characteristics, reflecting the marginal carbon emission cost required to transport this batch of waste, making this path a cost-defined option in subsequent optimization.

[0064] The solver module is the core software component on the server side responsible for performing complex data comparisons, logical judgments, and numerical calculations. Multi-dimensional comparison verification is a complex logical judgment process designed to ensure a high degree of correlation between physical events and virtual predictions in both numerical and temporal aspects. The time-series matching window is set as a time interval, for example, from 2 seconds to 5 minutes after the virtual access lock is generated. Its lower limit ensures a necessary delay in signal transmission and physical descent, while the upper limit prevents outdated predictions from mismatching with new physical events.

[0065] The content difference tolerance range is a tolerance threshold set for the allowable difference between the actual physical characteristics and the expected quality, for example, ±7%. This tolerance range takes into account sensor measurement errors, minor losses during the waste's fall, and model extrapolation errors. The decryption confirmation command is a logical signal generated internally by the system after successful multi-dimensional comparison verification, serving as the sole credential for triggering subsequent actions. The actual carbon emission increment is a quantified cost indicator, calculated based on the carbon emission factor of the transportation vehicle published by the country or industry, the expected mileage of the transport, and the transport quality extrapolated from the actual physical characteristics.

[0066] For example, with a timestamp of 2025-10-26T10:30:18.500Z, the solver module of the scheduling center receives the physical impact increment signature {timestamp: "...18.450Z", pulse_peak_voltage: 3.8, ...} from the P1 collection box. The module then retrieves the virtual access lock "e1b0c442..." and its associated metadata from the head of the pending scheduling queue, which includes the expected mass of 314kg and the generation timestamp "...15.120Z". The module inputs the multi-dimensional features of the physical impact increment signature into the physical feature deduction model to calculate the actual physical feature, i.e., the equivalent mass, as 310kg. Performing multi-dimensional comparison verification, at the content level, the calculated mass difference is |310-314|=4kg, with a difference rate of 4 / 314≈1.27%, which falls within the preset content difference range of ±7%. At the timing level, the difference between the event timestamps is 18.450s - 15.120s = 3.33s, which falls within the preset timing matching window of 2s to 5min.

[0067] After successful dual verification, the solver module generates a decryption confirmation command. In response, the module locates the edge connecting D0 and P1 on the underlying routing topology. Based on the vehicle type and estimated mileage required to transport 310kg of waste to the designated disposal point, the module calculates the actual carbon emission increment to be 16.8 kg of CO2 equivalent. Finally, the module accesses the weight attribute of this edge and rewrites its value from 999999999 to 16.8. This series of operations allows the previously logically frozen waste disposal path P1 to be officially activated due to confirmation of physical facts, and it is assigned a routing cost that can be compared by subsequent optimization algorithms.

[0068] In one embodiment of the present invention, step S6, which involves using an optimization engine to generate a spatiotemporally optimal matching sequence with the lowest carbon emission loss and issuing scheduling operation instructions, includes the following steps: The dynamic path planning module is invoked to perform a low-carbon oriented global convergence solution operation and analysis on the underlying routing topology network that rewrites the actual carbon emission increment value. The convergence solution operation results are used to filter and output a unique vehicle matching scheduling sequence and corresponding waste removal route map with the core orientation of minimizing the total fleet transportation mileage and eliminating the overall cumulative carbon emission cost of the system. A closed-loop scheduling instruction carrying the unique vehicle matching scheduling sequence and corresponding waste removal route map is generated and sent to the on-board information terminal of the target waste removal vehicle through the relay link of the communication network to start the physical waste removal operation.

[0069] Specifically, step S6 is executed in the optimization engine of the central server. This process is automatically triggered immediately after the successful rewriting of cost weights in step S5. The optimization engine calls the built-in dynamic path planning module and submits the underlying routing topology graph, which has been modified in step S5 and is currently in its current state, as input data to this module. In this topology graph, at least one connecting edge has had its transportation cost weight rewritten to a real and finite value, while the remaining unconfirmed connecting edges still retain infinite locking values. After receiving the graph, the dynamic path planning module performs a global convergence solution analysis oriented towards low carbon emissions. This analysis process uses heuristic optimization algorithms such as tabu search or simulated annealing, with minimizing the total carbon emissions of the scheduling scheme as the single objective function. The algorithm's operation logic is to plan one or more travel paths for the target node with finite cost weights among all available waste collection vehicles and calculate the total cost of these paths, which is the sum of the weights of all connecting edges on the path.

[0070] After multiple rounds of iterative optimization, the dynamic route planning module filters and outputs the current optimal solution through convergent solution calculations. This solution is defined as a unique vehicle matching and scheduling sequence and corresponding waste collection route map, with the core objective of minimizing the total fleet transportation mileage and eliminating the overall cumulative carbon emission cost of the system. This sequence explicitly specifies the ID of the unique vehicle that should be dispatched, the list of collection points that the vehicle needs to serve, and the order in which they are accessed.

[0071] The corresponding waste collection route map is a specific navigation path generated by the geographic information system service integrated within the dispatch system based on the geographic coordinates in the sequence. The instruction generation unit of the central server formats and encapsulates the output unique vehicle matching dispatch sequence and the corresponding waste collection route map, and integrates them to generate a structured closed-loop dispatch instruction. This instruction adopts, for example, JSON or XML format and is ultimately sent to the on-board information terminal installed on the selected target waste collection vehicle through a communication network relay link, such as through a secure MQTT message broker, so as to instruct the driver or autonomous driving system to start the physical waste collection operation.

[0072] The dynamic route planning module is a set of software functions that solves vehicle routing problems, capable of handling complex constraints such as time windows, multiple vehicle types, and dynamic nodes. The global convergence solution analysis is an iterative optimization process aimed at finding solutions that minimize or nearly minimize the overall objective function from a vast number of potential path combinations. The unique vehicle matching scheduling sequence is the optimal waste disposal plan calculated based on all available information at a specific time point, specifying which vehicle goes where and when to perform which task. The corresponding waste disposal route map transforms the abstract nodes in the scheduling sequence into detailed route guidance for navigation devices, representing actual geographical roads.

[0073] A closed-loop scheduling instruction is a data packet encapsulating all necessary information for the scheduling scheme. Its "closed-loop" nature is reflected in the fact that after the instruction is executed, the vehicle status is transmitted back to the system to update the global status and influence the next round of scheduling. The target waste collection vehicle is the most suitable vehicle selected from the fleet based on its current location, load status, and capacity matching degree to perform the task. The onboard information terminal is a hardware device installed in the driver's cab, possessing data display, human-machine interaction, and communication functions.

[0074] For example, after discovering that the edge weight connecting D0 and P1 in the underlying routing topology graph has been rewritten to 16.8, the optimization engine initiates a global convergence solution analysis. At this point, the topology graph state is: Cost(D0, P1) = 16.8, Cost(D0, P2) = 999999999, Cost(D0, P3) = 999999999. The objective function of the dynamic path planning module is to minimize the total carbon emission cost. The algorithm evaluates all possible paths starting from D0. The path cost to P1 is 16.8, while the path costs to P2 and P3 are infinite. The algorithm converges quickly, and the only optimal solution is to dispatch a vehicle to P1. The system then checks the status of the waste collection fleet and finds that the electric heavy-duty truck with vehicle ID CV-07 is idle and closest to D0. The module outputs a unique vehicle matching and scheduling sequence: assign vehicle CV-07, with the task order being D0->P1->disposal point DS-02. The system queries the GIS service to generate a corresponding transport route map from the vehicle's current location to collection point P1 and then to steel recycling and disposal point DS-02. The system integrates this information into a closed-loop scheduling instruction in JSON format. {"taskId":"TSK-20251026-001","vehicleId":"CV-07","route":[{"point":"P1","action":"load","cargo":"Q345 BSteel", "est_mass_kg": 310}, {"point": "DS-02", "action": "unload"}], "nav_path": "[…detailedcoordinates…]"}.

[0075] The closed-loop dispatch instruction in JSON format is sent to the on-board information terminal on the CV-07 truck via the 5G network. The terminal receives and parses the instruction, and displays the task details and navigation route to the driver on the screen. After the driver confirms, the engine is started and the physical collection operation to the P1 collection point begins.

[0076] In one embodiment of the present invention, step S7, tracking the key matching environment survival status and implementing remote anomaly countermeasures based on the video network channel for the unlocked blocking queue, includes the following steps: Throughout the entire operating cycle of the virtual access lock, the pending scheduling queue is continuously accumulated and pushed into the virtual access lock. The feedback authentication message and ciphertext pairing status of the physical impact incremental signature are polled and monitored frequently. In response to the abnormal scenario of the virtual access lock being lost due to the cumulative stacking number exceeding the security setting threshold but not yet receiving the physical impact incremental signature message within the legal timing window, the closed-loop scheduling channel connection pointing to the abnormal node is cut off.

[0077] Extract the spatial coordinate matrix position of the virtual waste pulse data packet that was registered when it was generated, and instruct the high-altitude monitoring pan-tilt probe at the engineering site to turn and aim at the target coordinate position in order to conduct close reconnaissance and collect evidence of the environmental hazards of illegal construction piles or foreign materials stuck at the corresponding location.

[0078] Specifically, step S7 is a long-term monitoring and countermeasure process running on the central server to ensure the stability of the scheduling system and handle abnormal situations. This process is enabled along with the main service after the system starts. The server's built-in health monitoring module is configured with a high-frequency polling timer, for example, every 500 milliseconds, the module checks the status of the core components of the entire system. During the entire runtime of the virtual access lock being pushed into the pending scheduling queue, the health monitoring module not only polls the queue, but more importantly, monitors and records the runtime logs and status messages output by the solver module in step S5, polling and monitoring the feedback authentication messages and ciphertext pairing status of the physical impact incremental signature at high frequency. The module tracks the lifecycle of each lock from generation to successful decryption or timeout by setting an independent timeout timer for each virtual access lock entering the queue. The system will then enter the abnormal scenario judgment logic.

[0079] The health monitoring module continuously counts the cumulative stacked number of virtual access locks in the current pending scheduling queue and compares it with a preset security threshold. For the virtual access lock at the head of the queue, the module checks whether its associated timeout timer has expired, i.e., whether a physical impulse incremental signature message capable of successfully decrypting the lock has been received within the preset valid time window. When the module detects a typical unlocking anomaly scenario, i.e., the cumulative stacked number of virtual access locks exceeds the security threshold, and the lock at the head of the queue has not received a valid decryption key within the specified time window, the system triggers an anomaly countermeasure procedure. As the first step of the countermeasure, the scheduling control module cuts off the closed-loop scheduling channel connection to the abnormal node. The module identifies the independent node corresponding to the virtual access lock causing the queue blockage and temporarily marks the node's status in the underlying routing topology as "faulty" or "unreachable," thereby preventing any new scheduling instructions from attempting to interact with this waste collection point. As the second step of the countermeasure, the scheduling control module extracts the spatial coordinate matrix position that was registered at the time of generation from the original data source of the virtual access lock that caused the blockage, namely the virtual waste pulse data packet generated in step S1.

[0080] After obtaining the three-dimensional coordinates, the module sends a command through its interface to the high-definition video monitoring system deployed at the substation construction site. The command instructs the high-altitude monitoring pan-tilt-zoom (PTZ) camera at the construction site to forcibly turn and align with the target coordinates. The monitoring system interprets the command, controls the designated PTZ camera to adjust its horizontal and vertical angles, and performs optical zoom to closely observe the actual situation around the coordinate point and transmit the real-time video stream back to the dispatch center. In this way, dispatchers can remotely collect evidence and analyze the root cause of the spatiotemporal key mismatch at that location. For example, illegal construction dumping may have resulted in waste not falling into the collection bin, or foreign materials may be obstructing the normal operation of sensors. This provides direct visual evidence and environmental hazard information for subsequent manual intervention and problem-solving.

[0081] Among these, high-frequency polling and monitoring is an automated system status check process designed to promptly detect any deviations from normal operating procedures. The cumulative stack of virtual access locks is a key indicator of the balance between waste generation rate and clearance capacity. The safety setting threshold is a threshold, for example, set at 10. When the number of pending virtual access locks exceeds this, it indicates that the system's clearance capacity may be unable to keep up with the waste generation rate, posing a potential systemic risk. An unlocking anomaly scenario refers to a virtually predicted waste event occurring, but no corresponding sensor-confirmed event in the physical world matches it, leading to a disruption in the scheduling chain. Disconnecting the closed-loop scheduling channel connection to the abnormal node is an emergency isolation measure to prevent fault propagation and resource waste. The high-altitude monitoring pan-tilt-zoom (PTZ) probe at the engineering site is a monitoring camera with remote control panning, tilting, and zoom functions, providing flexible visual coverage of a large area on site.

[0082] Close-range reconnaissance and evidence collection involves confirming abnormal situations on-site via remote video, aiming to provide irrefutable evidence for subsequent troubleshooting or management accountability. Illegal construction stockpiling refers to construction workers failing to dispose of waste materials into designated collection containers as required. Foreign material jamming refers to the phenomenon where non-target waste or excessively large waste materials mistakenly enter the collection point, preventing normal waste from falling smoothly or obstructing sensors.

[0083] For example, assuming the current security threshold is 10, over a period of time, due to accelerated on-site construction progress, the number of virtual access locks in the pending scheduling queue continues to accumulate, reaching 11, which exceeds the security threshold. The timeout timer for the virtual access lock at the head of the queue, targeting the P2 collection point, has also expired. This indicates that within the legal time window of more than 5 minutes, the system has not received a physical impact incremental signature message capable of unlocking the virtual access lock at point P2. The system therefore determines that it has entered an unlocking anomaly scenario, and the countermeasure program is activated. The scheduling control module temporarily marks the state of node P2 in the underlying routing topology diagram as "SUSPENDED," and any scheduling instructions for P2 will be rejected by the system. The module traces back to the source from the record of the virtual access lock that caused the blockage, extracts its spatial coordinate matrix position generated in step S1, and parses the global coordinates as (150.32, 210.88, 5.50).

[0084] The system generates the command: {“camera_id”:“PTZ-04”,“action”:“goto_preset”,“coordinates”:[150.32, 210.88, 5.50]}. This command is sent to the video monitoring server, which then controls the high-altitude monitoring pan-tilt-zoom (PTZ) camera on tower crane No. 4 at the construction site to be forcibly turned and aligned with this coordinate. The monitoring screen in the dispatch center displays a real-time image of this location, showing a pile of scaffolding steel pipes mistakenly stacked next to the P2 collection box, blocking the main passage and preventing the waste transport vehicle from properly approaching to dispose of waste.

[0085] See appendix Figure 2 The present invention also proposes a server-executed substation engineering waste reduction scheduling system, comprising the following modules: The virtual waste generation module is used to extract the physical and electrical characteristic edge of the transient current signal, and generate a trigger command based on the physical and electrical characteristic edge to strip the component node and generate a virtual waste pulse data packet containing the expected mass and spatial coordinate matrix. The routing environment initialization module is used to construct the underlying routing topology graph and initialize the transport cost weights of the connecting edges in the underlying routing topology graph to infinite latching values. The dynamic constraint generation module is used to calculate and generate virtual access locks based on the expected quality and preset material attribute data, push the virtual access locks into the pending scheduling queue and configure them on the transportation cost weight of the connection edge; The physical event perception module is used to extract mechanical momentum pulses and height change rate data from physical waste collection points and fuse them to form a physical impact incremental signature. The spatiotemporal coupling verification module is used to drive the physical impact incremental signature to initiate a matching verification request to the virtual access lock, extract the expected quality anchored by the virtual access lock and compare it with the real physical characteristics with tolerance, and respond to the verification by rewriting the infinite locking value to the real carbon emission increment value through the instruction. The carbon-optimal scheduling module is used to perform optimization calculations on the underlying routing topology map that rewrites the actual carbon emission increment values, generate a unique vehicle matching scheduling sequence and corresponding waste disposal route map, and issue closed-loop scheduling instructions. The anomaly monitoring and response module is used to respond to anomalies where the pending scheduling queue crosses the preset safety threshold and no physical impact incremental signature is received. It extracts the spatial coordinate matrix to instruct the monitoring probe to turn and investigate.

[0086] Each of the modules can be implemented in whole or in part through software, hardware, or a combination thereof. It supports hardware embedded in or independent of the processor in the computer device, and also supports software stored in the memory of the computer device, so that the processor can call and execute the operations corresponding to each of the above modules.

[0087] The above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention, and should all be included within the protection scope of the present invention.

Claims

1. A server-executed method for scheduling waste reduction in substation engineering, characterized in that, Includes the following steps: S1. Extract the physical and electrical characteristic edge of the transient current signal, and generate a trigger command based on the physical and electrical characteristic edge. Strip the component node to generate a virtual waste pulse data packet containing the expected mass and spatial coordinate matrix. S2. Construct the underlying routing topology graph and initialize the transport cost weights of the connecting edges in the underlying routing topology graph to infinite locking values. S3. Calculate the expected quality and preset material attribute data to generate a virtual access lock. Push the virtual access lock into the pending scheduling queue and configure it in the transportation cost weight of the connection edge. S4. Extract the mechanical momentum pulse and height change rate data of the physical waste collection point and fuse them to form a physical impact incremental signature; S5. Drive the physical impact incremental signature to initiate a matching verification request to the virtual access lock, extract the expected quality anchored by the virtual access lock and compare it with the real physical characteristics with tolerance. After the verification is passed, generate a decryption confirmation instruction. Respond to the decryption confirmation instruction to rewrite the infinite locking value into the real carbon emission increment value. S6. Perform optimization calculations on the underlying routing topology map that rewrites the actual carbon emission increment values, generate a unique vehicle matching scheduling sequence and corresponding waste disposal route map, and issue closed-loop scheduling instructions. S7. In response to an abnormal scenario where the pending scheduling queue crosses the preset safety setting line and no physical impact incremental signature is received, extract the spatial coordinate matrix and instruct the monitoring probe to turn around for reconnaissance.

2. The server-executed substation engineering waste reduction scheduling method according to claim 1, characterized in that, The process of stripping component nodes to generate virtual scrap pulse data packets containing the expected mass and spatial coordinate matrix includes the following steps: Calculate the first-order time derivative of a transient current signal to capture the rate of change; In response to a step drop in the amplitude of a transient current signal and its stabilization at a level below a preset no-load threshold, the specific signal pattern is identified as a physical electrical characteristic. Based on physical electrical characteristics, a trigger command carrying the current timestamp and device identifier is generated; Based on the equipment identifier, the component node is located by searching the preset construction task database, the three-dimensional geometric model is extracted to calculate the volume, and the material density is retrieved. The expected mass is calculated by multiplying the material density by the volume, and the position and orientation of the component nodes in the global coordinate system of the project are extracted and encoded into a spatial coordinate matrix. The expected quality, spatial coordinate matrix, timestamp, and device identifier are encapsulated into a virtual waste pulse data packet.

3. The server-executed substation engineering waste reduction scheduling method according to claim 1, characterized in that, Initialize the transport cost weights of the connecting edges in the underlying routing topology graph to infinite latching values, including the following steps: Load the preset engineering site layout data, which defines the geographical coordinates of the physical waste collection points and the coordinates of the waiting positions of the collection vehicles; Instantiate a directed graph data structure in memory to generate the underlying routing topology graph, map physical waste collection points as independent nodes, and create the standby positions of collection vehicles as source nodes; Find the directed edges from the source node to the independent nodes and generate the transport cost weights of the connecting edges; The transport cost weight of the connecting edge is assigned a predefined maximum floating-point number to generate an infinite locking value. The infinite locking value is used to block the optimization path calculation for independent nodes.

4. The server-executed substation engineering waste reduction scheduling method according to claim 1, characterized in that, Pushing the virtual access lock into the pending scheduling queue and configuring it on the transport cost weight of the connection edge includes the following steps: The expected quality value is converted into a string format and concatenated with the preset material attribute data to form the source information string; Input the source information string into a preset one-way hash function to perform an encrypted hash operation, and generate a fixed-length hexadecimal hash value as a virtual access lock; Send the virtual access lock to the pending scheduling queue that follows the first-in-first-out principle; The physical waste collection point is determined by querying a preset mapping table based on the internal component identifier of the virtual waste pulse data packet, and the virtual access lock is logically bound to the transportation cost weight of the connection edge pointing to the physical waste collection point.

5. A server-executed substation engineering waste reduction scheduling method according to claim 1, characterized in that, The process of fusing and constructing a physical impact incremental signature includes the following steps: In response to the virtual access lock being pushed into the pending scheduling queue, a low-level control command is sent to the embedded control unit inside the physical waste collection point; Activate the piezoelectric sensor device that is mechanically coupled to the bottom plate of the collection box and the ultrasonic array device installed on the top of the inner wall of the collection box; Mechanical momentum pulses are generated by collecting transient voltage signals from the impact of falling waste materials using piezoelectric sensors. The height value is converted from the distance to the surface of the waste accumulation in the box by continuously measuring the distance using an ultrasonic array device. The height change rate data is obtained by calculating the ratio of the difference between two adjacent height measurements to the sampling time interval. Extract the peak voltage and voltage integral of the mechanical momentum pulse, and align and package the peak voltage, voltage integral, height change rate data with the timestamp to generate a physical impact incremental signature.

6. The server-executed waste reduction scheduling method for substation projects according to claim 1, characterized in that, The process of rewriting the infinite latch value to the actual carbon emission increment value in response to the decryption confirmation command includes the following steps: Extract the virtual access lock and associated virtual waste pulse data packet from the head of the pending scheduling queue; The mechanical momentum pulse feature value and height change rate data in the physical impact incremental signature are input into the preset physical feature deduction model and mapped to generate a mass estimate as the real physical feature. Calculate the absolute difference between the actual quality estimate and the expected quality, and determine whether the absolute difference is within the preset quality error tolerance threshold. Calculate the difference between the generation timestamp of the physical impact incremental signature and the generation timestamp of the virtual access lock, and determine whether the time difference falls within the dynamic timing matching window extracted based on the construction flow rhythm. In response to the condition that both the absolute difference and the time difference meet the requirements, a decryption confirmation command is generated; Based on the decryption confirmation command, the infinite lock value on the transport cost weight of the connection edge is broken, and the real physical characteristics are input into the preset carbon emission model to generate the real carbon emission increment value and cover the infinite lock value.

7. A server-executed substation engineering waste reduction scheduling method according to claim 1, characterized in that, Generate a unique vehicle matching and scheduling sequence and corresponding waste collection route map, and issue closed-loop scheduling instructions, including the following steps: The preset dynamic path planning module is invoked to receive the underlying routing topology map that rewrites the actual carbon emission increment value; Using the minimization of total carbon emissions as the objective function, a heuristic optimization algorithm is performed on the underlying routing topology graph to achieve global convergence and solution analysis. The lowest total cost travel path is selected through convergent solution analysis, and a unique vehicle matching scheduling sequence containing the target waste collection vehicle identifier and access order is output. The system calls a preset geographic information system service to generate a corresponding waste disposal route map based on the geographic coordinates in the unique vehicle matching and scheduling sequence. The unique vehicle matching and scheduling sequence and the corresponding waste collection route map are formatted and encapsulated to generate a closed-loop scheduling instruction, which is then sent to the on-board information terminal of the target waste collection vehicle through the communication network relay link.

8. A server-executed substation engineering waste reduction scheduling method according to claim 1, characterized in that, Extracting the spatial coordinate matrix to instruct the monitoring probe to turn and detect, includes the following steps: Set an independent timeout timer for each virtual access lock in the pending scheduling queue; Continuously count the cumulative stacked number of virtual access locks in the pending scheduling queue and compare the cumulative stacked number with the preset security setting threshold; In response to the pending scheduling queue crossing the preset security setting threshold, determine whether a decryption confirmation command has been received within the preset setting timing matching window; When the accumulated stacking quantity exceeds the preset safety threshold and the independent timeout timer expires, an abnormal scenario is determined to be triggered. Mark the status of independent nodes in the underlying routing topology graph corresponding to abnormal scenarios as unreachable to cut off channel connections; Extract the spatial coordinate matrix from the virtual waste pulse data packet, and issue control commands containing the spatial coordinate matrix to drive the high-altitude monitoring pan-tilt probe to adjust the horizontal and vertical angles to align with the target coordinates and perform optical zoom reconnaissance.

9. A server-executed substation engineering waste reduction scheduling method according to claim 6, characterized in that, The process involves inputting real physical characteristics into a preset carbon emission model to generate real carbon emission increment values ​​and overriding the infinite latch value, including the following steps: Extract the actual carbon emission increment value output by the preset carbon emission model, and query the preset carbon quota database to obtain the remaining carbon emission quota for the current substation project; Calculate the difference between the remaining carbon emission allowance and the actual increase in carbon emissions, and determine whether the difference is lower than the preset carbon emission warning threshold. In response to the difference being lower than the preset carbon emission warning threshold, a dynamic carbon emission limiting strategy is generated. The dynamic carbon emission throttling strategy is fed back to the underlying routing topology, and the infinite latching value corresponding to the transportation cost weight of the unconfirmed connection edge is dynamically increased according to a preset ratio.

10. A server-executed substation engineering waste reduction scheduling system, characterized in that, Includes the following modules: The virtual waste generation module is used to extract the physical and electrical characteristic edge of the transient current signal, and generate a trigger command based on the physical and electrical characteristic edge to strip the component node and generate a virtual waste pulse data packet containing the expected mass and spatial coordinate matrix. The routing environment initialization module is used to construct the underlying routing topology graph and initialize the transport cost weights of the connecting edges in the underlying routing topology graph to infinite latching values. The dynamic constraint generation module is used to calculate and generate virtual access locks based on the expected quality and preset material attribute data, push the virtual access locks into the pending scheduling queue and configure them on the transportation cost weight of the connection edge; The physical event perception module is used to extract mechanical momentum pulses and height change rate data from physical waste collection points and fuse them to form a physical impact incremental signature. The spatiotemporal coupling verification module is used to drive the physical impact incremental signature to initiate a matching verification request to the virtual access lock, extract the expected quality anchored by the virtual access lock and compare it with the real physical characteristics with tolerance. After the verification is passed, a decryption confirmation instruction is generated. In response to the decryption confirmation instruction, the infinite locking value is rewritten as the real carbon emission increment value. The carbon-optimal scheduling module is used to perform optimization calculations on the underlying routing topology map that rewrites the actual carbon emission increment values, generate a unique vehicle matching scheduling sequence and corresponding waste disposal route map, and issue closed-loop scheduling instructions. The anomaly monitoring and response module is used to respond to anomalies where the pending scheduling queue crosses the preset safety threshold and no physical impact incremental signature is received. It extracts the spatial coordinate matrix to instruct the monitoring probe to turn and investigate.

Images