Urban garbage classified collection and scheduling method based on Internet of Things

Abstract

The invention discloses an urban garbage classified collection and scheduling method based on the Internet of Things, and relates to the technical field of urban garbage classified collection, and the method comprises the steps: deploying an intelligent garbage can with the filling rate and environmental parameter perception capability in each region of a city, and uploading the collected state data to an edge gateway of the region through a low-power wide area network; the edge gateway performs localized priority evaluation based on the received real-time state data of each intelligent garbage can in combination with the capacity information, the last clearance time and the attribute of the area where the intelligent garbage can is located, and generates a dynamic priority score of each garbage can; when the dynamic priority score of any dustbin exceeds a preset response threshold value, the edge gateway triggers an event reporting mechanism and sends a clearance alarm containing position and priority information to a regional dispatching center; and on the basis of receiving the clearance alarm, the regional dispatching center combines weather forecast, festival calendar and regional people flow density data to carry out trend prediction on the garbage cans which are not alarmed yet and have high filling risks.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of urban garbage classification collection, and particularly to a method for urban garbage classification collection and scheduling based on the Internet of Things. BACKGROUND

[0002] Urban garbage classification collection technology refers to a process of effectively separating, collecting and transporting urban household garbage according to different classification standards (such as recyclables, hazardous waste, kitchen waste and other waste) through a series of technical means and management measures. This technology not only involves the construction of hardware facilities, such as setting up different types of garbage cans or bins, but also includes the development of software systems, such as the application of Internet of Things (IoT) technology, for monitoring the filling status of garbage bins. Therefore, how to use advanced technical means to improve the intelligent level and safety of urban garbage classification collection has become one of the problems to be solved at present.

[0003] In the field of urban garbage classification collection, traditional garbage bins lack real-time sensing capability of filling rate and environmental parameters, resulting in that the cleaning and transportation operation often relies on fixed cycles or citizen feedback, making it difficult to achieve precise scheduling, and in the process of executing the cleaning and transportation task, the differences in vehicle driving energy consumption are not fully considered, affecting the overall energy consumption optimization. SUMMARY

[0004] In view of the above existing problems, the present application is proposed.

[0005] Therefore, the present application provides a method for urban garbage classification collection and scheduling based on the Internet of Things to solve the problem that traditional garbage bins lack real-time sensing capability of filling rate and environmental parameters, resulting in that the cleaning and transportation operation often relies on fixed cycles or citizen feedback, making it difficult to achieve precise scheduling, and in the process of executing the cleaning and transportation task, the differences in vehicle driving energy consumption are not fully considered, affecting the overall energy consumption optimization.

[0006] To solve the above technical problems, the present application provides the following technical solutions: In a first aspect, the present application provides a method for urban garbage classification collection and scheduling based on the Internet of Things, which comprises: Intelligent garbage bins with filling rate and environmental parameter sensing capability are deployed in each region of the city, and the collected state data is uploaded to the edge gateway of the region through a low-power wide-area network; The edge gateway performs localized priority evaluation based on the real-time state data of each intelligent garbage bin, combines its capacity information, the last cleaning and transportation time and the attributes of the region, and generates a dynamic priority score for each garbage bin; When the dynamic priority score of any garbage bin exceeds a preset response threshold, the edge gateway triggers an event reporting mechanism and sends a cleaning and transportation alarm containing location and priority information to the regional scheduling center; Based on receiving collection and disposal alarms, the regional dispatch center integrates weather forecasts, holiday calendars, and regional population density data to predict trends for trash cans that have not yet received alarms but pose a high risk of overflowing, thus identifying potential high-risk points. The emergency tasks corresponding to the cleaning alarms and the predicted potential high-risk tasks are combined to form a set of tasks to be scheduled, and this set of tasks is sent to each vehicle intelligent agent participating in the cleaning operation. Each vehicle intelligent agent assesses its response capability and makes bidding decisions for each task in the task set based on its current location, loading status, and driving energy consumption model. The regional dispatch center allocates tasks based on the bidding results of each intelligent agent and generates dynamic dispatch instructions. Each vehicle intelligent agent provides real-time feedback on its operating status during the execution of scheduling instructions. When a task is delayed or a new high-priority alarm is added, the system performs a local task reallocation based on the current task allocation status and the vehicle's location. After completing the cleaning task, the system sends the execution data back.

[0007] As a preferred embodiment of the IoT-based urban waste sorting, collection, and scheduling method of the present invention, the following steps are taken: Deploying smart trash cans with filling rate and environmental parameter sensing capabilities in various urban areas, and uploading the collected status data to the edge gateway of the respective area via a low-power wide-area network: Multiple smart trash cans are deployed within the designated administrative or functional areas of the city. Each smart trash can is equipped with a non-contact distance sensor to measure the real-time distance between the trash accumulation surface inside the can and the top of the can. The smart trash can has built-in temperature and humidity sensors to collect microenvironmental parameters inside the can. The smart trash can is equipped with a communication module, which encapsulates real-time measured distance data, environmental parameters, and unique device identification information into data packets according to a preset communication protocol; The data packets are transmitted via a low-power wide-area network to an edge gateway covering the geographical area. The edge gateway receives and parses the data packets to extract the original measurement values ​​of each smart trash can.

[0008] As a preferred embodiment of the IoT-based urban waste sorting, collection, and scheduling method of the present invention, the edge gateway, based on the received real-time status data of each smart waste bin, combined with its capacity information, last collection time, and location attributes, performs localized priority evaluation to generate a dynamic priority score for each waste bin. The specific steps are as follows: The edge gateway calculates the ratio of the current filling height to the total height based on the received real-time distance value and the pre-stored total height of the enclosure, and records this ratio as the filling rate. ; Retrieve the timestamp of the last time the smart garbage can was emptied, calculate the time interval from the time of completion to the current time, and record it as the retention time ; According to the geographical coordinates of the smart garbage can, its area type is matched, including medical area, commercial area, transportation hub area and residential area, and different area types correspond to different area weight values , wherein the area weight of the medical area and the commercial area is higher than that of the residential area; Based on the filling rate , the retention time and the area weight , the dynamic priority score is generated by weighted combination , and the calculation process is as follows: ; Wherein, , , , is a pre-set adjustment parameter, is the generated dynamic priority score, which reflects the urgency of the garbage can.

[0009] As a preferred scheme of the city garbage classification collection and scheduling method based on Internet of Things, when the dynamic priority score of any garbage can exceeds the preset response threshold, the edge gateway triggers the event reporting mechanism, sends the emptying alarm containing the position and priority information to the regional scheduling center, and the specific steps are as follows: The edge gateway compares the currently calculated dynamic priority score with the first preset threshold ; If > , it is determined that the garbage can enters a high priority state, and an emptying alarm event is generated; The emptying alarm event contains the unique identification, geographical coordinates, current filling rate , dynamic priority score and event generation time of the garbage can; The edge gateway uploads the emptying alarm event to the regional scheduling center through a secure communication link, and records the reporting time locally to prevent the same event from being reported repeatedly within a set cooling period.

[0010] As a preferred scheme of the city garbage classification collection and scheduling method based on the Internet of Things, wherein: on the basis of receiving the cleaning and transporting alarm, the regional scheduling center fuses meteorological forecast, festival calendar and regional human flow density data, carries out trend prediction on the garbage can which has not yet alarmed but has high filling risk, and identifies potential high-risk points, and the specific steps are: The regional scheduling center receives meteorological data in a future preset time window from a meteorological service platform, and the meteorological data includes rainfall probability, average temperature and air humidity; Obtain festival information output by a city public calendar system, and the festival information is used to identify whether the current date is a legal holiday, a local activity day or a public holiday; Receive the regional human flow density index provided by the city sensing network, which is calculated based on mobile communication signals, Wi-Fi probes or video analysis technology, and reflects the personnel gathering degree of the target area in a unit time; For the intelligent garbage can which does not trigger the cleaning and transporting alarm, the filling rate change sequence in the past preset period is called, and the filling rate change sequence is composed of historical filling rate data periodically reported by the edge gateway; The filling rate change sequence is spatiotemporally aligned with the received meteorological data, festival information and regional human flow density index, so as to ensure that each data is synchronized in time dimension and matched to the same geographical grid unit in space dimension; A multi-dimensional input feature vector is constructed, and the feature vector includes historical filling rate sequence, current rainfall probability, temperature value, festival identification state and regional human flow density value; The multi-dimensional input feature vector is input into a long short-term memory neural network prediction model which is trained in advance, and the model has a memory cell structure, can capture long-term dependence in time series, and can model the nonlinear influence of external environmental variables; The long short-term memory neural network prediction model outputs the filling rate prediction value in the future preset time window, and the prediction value represents the predicted garbage accumulation degree of the garbage can at the future time; Based on the filling rate prediction value, the capacity parameter of the garbage can, the regional attribute weight and the retention time are combined to calculate the predicted dynamic priority score at the future time; If the predicted dynamic priority score exceeds the preset potential risk threshold, it is determined that the garbage can has a high filling risk in the future period, and is marked as a potential high-risk point, and a corresponding predictive cleaning and transporting task is generated, and the predictive cleaning and transporting task includes task position, predicted priority level and recommended response time window.

[0011] As a preferred scheme of the city garbage classification collection and scheduling method based on the Internet of Things, wherein: the emergency task corresponding to the clean-up alarm and the potential high-risk task generated by prediction are jointly constructed as a to-be-scheduled task set, and the task set is issued to each vehicle agent participating in the clean-up operation, and the specific steps are: The regional scheduling center defines all tasks reported by the edge gateway and having a dynamic priority score exceeding a first response threshold as first-level tasks, and the first-level tasks have the highest scheduling priority; The task generated by the potential high-risk point identified by the prediction model is defined as a second-level task, and the second-level task is used to intervene in the possible overflow event in advance; A unique task identifier is assigned to each task, and the geographic location coordinates, task type level, recommended clean-up time window, and associated garbage can number are recorded; All first-level tasks and second-level tasks are summarized to form a to-be-scheduled task set, and the to-be-scheduled task set logically constitutes a task pool that can be accessed by multiple vehicle agents; The regional scheduling center periodically issues the to-be-scheduled task set to the vehicle agents that are currently online and have job capabilities in a structured data format through a message broadcast mechanism; The vehicle terminal of each clean-up vehicle receives and parses the task set data, and updates the locally stored task list; The task set is issued at a frequency that is dynamically adjusted according to the task density of the city area, and a higher broadcast frequency is used in high-density areas.

[0012] As a preferred scheme of the city garbage classification collection and scheduling method based on the Internet of Things, wherein: each vehicle agent performs response capability evaluation and bidding decision on each task in the task set according to its current location, loading state, and travel energy consumption model, and the regional scheduling center performs task allocation according to the bidding results of each agent to generate a dynamic scheduling instruction, and the specific steps are: Each vehicle agent obtains its current geographic location, which is provided by the vehicle-mounted GPS or Beidou positioning module; The current load value fed back by the vehicle-mounted weighing system is read and compared with the maximum rated load of the vehicle to determine whether it has the loading capacity to undertake new tasks; For each task in the to-be-scheduled task set, the shortest path distance from the current location of the vehicle to the target location of the task is calculated, and the path calculation is based on the city road network topology and real-time traffic state data; The estimated arrival time of the vehicle at the task point is estimated according to the path distance and the road passing speed model; A vehicle driving energy consumption model is established, which considers the vehicle self-weight, current load, road slope and driving speed, and calculates the energy consumption increment per unit distance increased by performing the task; A task response score function is constructed for quantifying the response willingness and execution efficiency of the vehicle to the task, and the expression is: ; Among them, The response score of the vehicle to a certain task is represented by, The expected arrival time is represented by, The dynamic priority score of the task is represented by, The energy consumption increment per unit distance brought by performing the task is represented by, 、 、 The preset score weight coefficient adjusts the influence degree of time efficiency, task priority and energy consumption cost in the score respectively; The vehicle agent calculates the response score of all executable tasks based on the above score function, and selects the task with the highest score to bid; Each vehicle uploads the bidding information to the regional dispatch center after encryption, and the bidding information includes task identification, response score and expected arrival time; The regional dispatch center receives the bidding data of all vehicles, and combines the credit score generated by the historical performance record of the vehicle to perform weighted processing on the response score; A task allocation optimization algorithm is used to allocate each task to the vehicle with the highest comprehensive score; Generate dynamic scheduling instructions containing task sequence, driving path and time node, and issue to each vehicle agent through a secure communication link.

[0013] As a preferred scheme of the urban garbage classification collection and scheduling method based on the Internet of Things, wherein: each vehicle agent feeds back the running state in real time during the execution of the scheduling instruction, and when there is a task delay or a new high-priority alarm, the local task re-allocation is performed based on the current task allocation state and the relationship between the vehicle positions, and the execution data is returned after completing the cleaning and transportation task, and the specific steps are: The vehicle agent uploads the current position coordinates, task execution progress, remaining loading capacity and vehicle running state to the regional dispatch center at a fixed time interval during the execution of the scheduling instruction; The regional dispatch center continuously monitors the expected arrival time of each vehicle, and if the expected arrival time of a vehicle to the allocated task exceeds the preset allowable deviation range, it is determined that the task execution is delayed; When the edge gateway reports a new cleaning and transportation alarm, the regional dispatch center immediately queries the task state of the garbage can corresponding to the alarm to determine whether it has been allocated and not completed; If the task has not been assigned or the originally assigned vehicle is in a task delay state, the task is re-released to the set of tasks waiting for scheduling, forming a temporary schedulable task; The regional dispatch center notifies the vehicle agents currently located in the geographical proximity area of the task, triggering a local re-evaluation process; The neighboring vehicle agents re-acquire the task information and, based on their current location and loading state, re-calculate the response score for the task according to the response score function; Each neighboring vehicle uploads the updated response score to the regional dispatch center; The regional dispatch center re-executes the task assignment logic based on the newly reported score results, assigning the task to the currently optimal vehicle; Local adjustment instructions are generated and distributed to the newly assigned vehicle and the original executing vehicle, enabling seamless handover of the task; After receiving the adjustment instructions, the newly assigned vehicle updates the local task plan and plans the driving path to the new task point; After receiving the task release instructions, the original executing vehicle clears the local task record and continues to execute subsequent tasks or returns to the dispatch center; After all the tasks are completed, the vehicle agent packages the task execution log and uploads it to the cloud data center, including the actual cleaning and collection time, the amount of collected garbage, the driving path trajectory, the task completion status, and the environmental perception data; After receiving the execution log, the cloud data center uses the deviation between the actual cleaning and collection data and the predicted value to periodically optimize the weight parameters in the dynamic priority score model, the network parameters of the prediction model, and the adjustment coefficients in the response score function.

[0014] In a second aspect, the present application provides a computer device comprising a memory and a processor, wherein the memory stores a computer program, and wherein the computer program, when executed by the processor, implements any step of the method for urban garbage classification collection and scheduling based on Internet of Things according to the first aspect of the present application.

[0015] In a third aspect, the present application provides a computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements any step of the method for urban garbage classification collection and scheduling based on Internet of Things according to the first aspect of the present application.

[0016] The application has the beneficial effects that: by constructing an integrated technical architecture of edge dynamic evaluation-environment fusion prediction-multi-agent collaborative scheduling, the paradigm transformation of urban garbage classification and collection from passive response to active prediction, from centralized control to distributed autonomy is realized, the edge gateway is used for local priority calculation of the garbage can state, the system response delay is reduced, the timeliness of high-priority event reporting is improved, a long short-term memory neural network prediction model is constructed combined with multiple external factors such as weather, festivals and crowd density, the forward-looking identification ability of potential high-risk points is effectively improved, sudden overflow events are avoided, the garbage collection vehicle is modeled as an intelligent agent with bidding and negotiation ability, a task allocation mechanism based on response score is introduced, dynamic matching of resources and elastic optimization of path are realized, the accuracy, real-time performance and resource utilization rate of the collection operation are improved, and the operation cost and environmental risk are reduced. BRIEF DESCRIPTION OF DRAWINGS

[0017] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings needed in the embodiment description will be briefly introduced. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creating laborious work.

[0018] Figure 1 The flowchart of the method for urban garbage classification collection and scheduling based on Internet of Things in embodiment 1. DETAILED DESCRIPTION

[0019] In order to make the above-mentioned purposes, features and advantages of the present application more apparent and easy to understand, the specific embodiments of the present application will be described in detail below with reference to the drawings of the specification.

[0020] In the following description, many specific details are set forth in order to provide a thorough understanding of the present application, but the present application can also be implemented in other ways different from those described herein, and those skilled in the art can make similar generalizations without departing from the connotation of the present application, therefore the present application is not limited to the specific embodiments disclosed below.

[0021] Secondly, the "one embodiment" or "embodiment" referred to herein means that the specific features, structures or characteristics can be included in at least one implementation of the present application. "In one embodiment" appearing in different places in the specification does not mean the same embodiment, nor is it an independent or alternative embodiment that excludes other embodiments.

[0022] Embodiment, refer to Figure 1 For the embodiments of the present application, the embodiment provides a method for urban garbage classification collection and scheduling based on Internet of Things, comprising the following steps: S1, Deploy intelligent garbage cans with filling rate and environmental parameter sensing capability in each area of the city, upload the collected state data to the edge gateway of the corresponding area through low-power wide-area network; Further, multiple intelligent garbage cans are arranged in the administrative or functional areas designated by the city, and each intelligent garbage can is equipped with a non-contact distance sensor for measuring the real-time distance between the garbage accumulation surface in the can and the top of the can; The intelligent garbage can is provided with a temperature and humidity sensor for collecting the micro-environmental parameters inside the can; The intelligent garbage can is provided with a communication module, which encapsulates the real-time distance data, environmental parameters and device unique identification information into a data packet according to a preset communication protocol; The data packet is transmitted to the edge gateway covering the geographic range through a low-power wide-area network, and the edge gateway receives and analyzes the data packet to extract the original measurement values of each intelligent garbage can; It should be noted that the non-contact distance sensor uses ultrasonic or infrared ranging technology to avoid pollution and wear caused by direct contact with garbage, improving the stability of long-term operation of the device. The communication module supports LoRa or NB-IoT protocol to ensure reliable data transmission with low power consumption, long distance and high penetration in complex urban environments. The edge gateway is deployed at the regional aggregation node and has local storage and computing capability, which can temporarily store data when the network is interrupted and supplement the transmission after the network is restored, ensuring data integrity.

[0023] S2, Based on the real-time state data of each intelligent garbage can received by the edge gateway, the capacity information, the last cleaning time and the regional attribute are combined to perform localized priority evaluation, and the dynamic priority score of each garbage can is generated; Further, the edge gateway calculates the ratio of the current filling height to the total height based on the received real-time distance value and the pre-stored total height of the can, which is denoted as filling rate ; The timestamp of the last time the intelligent garbage can was cleaned is retrieved, and the time interval from the cleaning completion time to the current time is calculated, which is denoted as retention time ; According to the geographic coordinates of the intelligent garbage can, the type of the area to which it belongs is matched, and the area type includes medical area, business area, transportation hub area and residential area. Different area types correspond to different area weight values , wherein the area weight of the medical area and the business area is higher than that of the residential area; Based on the filling rate , the retention time and the area weight , the dynamic priority score is generated by weighted combination , and the calculation process is as follows: ; wherein, , , , is a preset adjustment parameter, is a generated dynamic priority score, which reflects the urgency of the garbage bin to be emptied; It should be noted that the calculation process of the dynamic priority score is completed on the edge side, avoiding uploading all raw data to the central server, reducing network bandwidth pressure and central processing delay, and the adjustment parameter , , , According to the historical emptying data and actual operation feedback, the scoring model is calibrated to ensure good adaptability and discrimination in different seasons and different regional scenarios. The introduction of the retention time effectively identifies garbage bins that are not full but have been long empty, preventing secondary pollution caused by microbial fermentation or leachate leakage.

[0024] S3, when the dynamic priority score of any garbage bin exceeds the preset response threshold, the edge gateway triggers the event reporting mechanism, and sends the emptying alarm containing the location and priority information to the regional dispatch center; Further, the edge gateway compares the currently calculated dynamic priority score with the first preset threshold ; If > , it is determined that the garbage bin enters a high priority state, and an emptying alarm event is generated; The emptying alarm event contains the unique identifier, geographic coordinates, current filling rate , dynamic priority score and event generation time of the garbage bin; The edge gateway uploads the emptying alarm event to the regional dispatch center through a secure communication link, and records the reporting time locally to prevent the same event from being reported repeatedly within a set cooling period; It should be noted that the event reporting mechanism adopts a threshold comparison and cooling period double control strategy, which ensures timely reporting of high priority events and avoids frequent false alarms caused by sensor fluctuations or temporary overflow. The secure communication link is established based on the TLS encryption protocol to ensure that the emptying alarm information is not tampered with or stolen during transmission. The binding of the unique identifier and the geographic coordinates provides accurate basis for subsequent task tracking and responsibility tracing.

[0025] S4, on the basis of receiving the emptying alarm, the regional dispatch center integrates weather forecast, festival calendar and regional crowd density data to predict the trend of garbage bins that have not been alarmed but have high filling risk, and identify potential high-risk points; Further, the regional dispatch center receives meteorological data within a future preset time window from the meteorological service platform, including rainfall probability, average temperature, and air humidity; Obtain the festival information output by the city public calendar system, which is used to identify whether the current date is a statutory holiday, a local event day, or a public holiday; Receive the regional passenger flow density index provided by the city perception network, which is calculated based on mobile communication signals, Wi-Fi probes, or video analysis technology, and reflects the degree of personnel gathering in the target area per unit time; For the intelligent garbage can that has not triggered the clean-up alarm, retrieve its filling rate change sequence in the past preset period, which is composed of historical filling rate data periodically reported by the edge gateway; Align the filling rate change sequence with the received meteorological data, festival information, and regional passenger flow density index in time and space, ensuring that all data are synchronized in time and matched to the same geographic grid unit in space; Construct a multi-dimensional input feature vector, which includes the historical filling rate sequence, current rainfall probability, temperature value, festival identification status, and regional passenger flow density value; Input the multi-dimensional input feature vector into the pre-trained long short-term memory neural network prediction model, which has a memory cell structure and can capture long-term dependencies in time series and model the nonlinear effects of external environmental variables; The long short-term memory neural network prediction model outputs the filling rate prediction value within the future preset time window, which represents the predicted garbage accumulation level of the garbage can at the future time; Based on the filling rate prediction value, combined with the capacity parameters of the garbage can, regional attribute weights, and retention time, calculate its predicted dynamic priority score at the future time; If the predicted dynamic priority score exceeds the preset potential risk threshold, it is determined that the garbage can has a high filling risk in the future period, and it is marked as a potential high-risk point, and a corresponding predictive clean-up task is generated, which includes task location, predicted priority level, and recommended response time window; It should be noted that the long short-term memory neural network prediction model is trained offline with a large amount of historical filling data and external environmental factors before deployment, and is periodically fine-tuned online with newly collected actual clean-up data to adapt to changes in urban population flow patterns, dynamic scenarios such as new areas being put into use, and spatial and temporal alignment to ensure that meteorological and passenger flow data accurately match the spatial location of the target garbage can, avoiding prediction bias due to data misalignment. The festival identification status is input into the model as a binary variable to improve the prediction accuracy of garbage generation peaks during holidays.

[0026] S5, the emergency task corresponding to the clean-up alarm is constructed into a to-be-scheduled task set together with the potential high-risk task generated by prediction, and the task set is issued to each vehicle agent participating in the clean-up operation; Further, the regional scheduling center defines all tasks reported by the edge gateway and having a dynamic priority score exceeding a first response threshold as first-level tasks, and the first-level tasks have the highest scheduling priority; The task generated by the potential high-risk point identified by the prediction model is defined as a second-level task, and the second-level task is used to intervene in a possible overflow event in advance; A unique task identifier is assigned to each task, and the geographic location coordinates, task type level, recommended clean-up time window, and associated garbage can number are recorded; All first-level tasks and second-level tasks are aggregated to form a to-be-scheduled task set, and the to-be-scheduled task set logically constitutes a task pool that can be accessed by multiple vehicle agents; The regional scheduling center periodically issues the to-be-scheduled task set to the vehicle agents currently online and having operation capability in a structured data format through a message broadcast mechanism; The vehicle terminal of each clean-up vehicle receives and parses the task set data, and updates the locally stored task list; The frequency of issuing the task set is dynamically adjusted according to the task density of the city area, and a higher broadcast frequency is used in high-density areas; It should be noted that the structured data format of the task set is packaged using a lightweight serialization protocol, such as Protocol Buffers or JSON-LD, which balances data expression ability and parsing efficiency, the message broadcast mechanism supports multicast and groupcast modes, reduces the overhead of repeated sending, and the task type level is not only used for scheduling priority sorting, but also guides vehicle resource allocation, for example, first-level tasks are preferentially assigned to large clean-up vehicles, and second-level tasks can be responded by small mobile vehicles, improving resource allocation flexibility.

[0027] S6, each vehicle agent performs response capability evaluation and bidding decision on each task in the task set according to its current position, loading state and driving energy consumption model, and the regional scheduling center performs task allocation according to the bidding results of each agent to generate a dynamic scheduling instruction; Further, each vehicle agent obtains its current geographic position, which is provided by a vehicle-mounted GPS or Beidou positioning module; The current load value fed back by the vehicle-mounted weighing system is read and compared with the maximum rated load of the vehicle to determine whether it has the loading capacity to undertake a new task; For each task in the set of tasks to be scheduled, calculate the shortest path distance from the vehicle's current location to the task's target location. The path calculation is based on the city's road network topology and real-time traffic status data. Based on the path distance and road traffic speed model, estimate the estimated arrival time of the vehicle to the task point; Establish a vehicle driving energy consumption model that takes into account the vehicle's weight, current load, road gradient, and driving speed, and calculate the increase in energy consumption per unit distance for performing this task. Construct a task response scoring function to quantify the vehicle's willingness to respond to the task and its execution efficiency. The expression is as follows: ; in, This indicates the vehicle's response score for a specific task. Indicates the estimated arrival time. This represents the dynamic priority score of the task. This indicates the increase in energy consumption per unit distance resulting from performing this task. , , The pre-set scoring weight coefficients are used to adjust the influence of time efficiency, task priority, and energy cost on the scoring. Based on the above scoring function, the vehicle intelligent agent calculates the response score for each executable task and selects the task with the highest score for bidding. Each vehicle will encrypt and upload its bidding information to the regional dispatch center. The bidding information includes the task identifier, response score, and estimated arrival time. The regional dispatch center receives bidding data from all vehicles and, in conjunction with the credit score generated from the vehicle's historical performance record, performs weighted processing on the response score. A task allocation optimization algorithm is used to assign each task to the vehicle with the highest overall score; Generate dynamic scheduling instructions containing task sequences, driving routes, and time nodes, and send them to each vehicle intelligent agent through a secure communication link; It should be noted that the driving energy consumption model comprehensively considers the characteristics of the vehicle's power system and road conditions. The energy consumption increment per unit distance reflects the impact of performing a specific task on overall energy efficiency, avoiding the sacrifice of energy efficiency in the pursuit of rapid response. The design of the response scoring function enables the vehicle's intelligent agent to automatically weigh the three factors of time, priority, and cost during the bidding process, achieving optimal matching without human intervention. The credit scoring mechanism effectively suppresses false bidding or malicious task grabbing, ensuring the fairness and reliability of the scheduling system.

[0028] S7. Each vehicle intelligent agent provides real-time feedback on its operating status during the execution of scheduling instructions. When a task delay occurs or a new high-priority alarm is added, it performs local task reallocation based on the current task allocation status and the relationship between vehicle location. After completing the cleaning task, it sends the execution data back. Furthermore, during the execution of scheduling instructions, the vehicle intelligent agent uploads its current location coordinates, task execution progress, remaining loading capacity, and vehicle operating status to the regional scheduling center at fixed time intervals. The regional dispatch center continuously monitors the estimated arrival time of each vehicle. If the estimated arrival time of a vehicle for an assigned task exceeds the preset allowable deviation range, it is determined to be a task execution delay. When the edge gateway reports a new waste collection alarm, the regional dispatch center immediately queries the task status of the corresponding waste bin to determine whether it has been assigned and has not yet been completed. If the task has not yet been assigned, or the originally assigned vehicle is in a task delay state, the task will be released back into the set of tasks to be scheduled, forming a temporary schedulable task. The regional dispatch center notifies the vehicle agents currently located in the geographical vicinity of the task, triggering a local reassessment process. The nearby vehicle agent reacquires the task information and, based on its current location and loading status, recalculates the response score for the task according to the response scoring function; Each nearby vehicle will upload its updated response score to the regional dispatch center; Based on the newly reported scoring results, the regional dispatch center re-executes the task allocation logic and assigns the task to the current best vehicle. Generate local adjustment instructions and issue them to the newly assigned vehicle and the original executing vehicle to achieve seamless handover of tasks; After receiving the adjustment instructions, the newly assigned vehicle updates its local mission plan and plans its route to the new mission location. After receiving the task release command, the original vehicle clears its local task record and continues to execute subsequent tasks or returns to the dispatch center. After all the cleaning tasks are completed, the vehicle's intelligent agent will package the task execution log and upload it to the cloud data center. The execution log includes the actual cleaning time, the amount of garbage collected, the driving route trajectory, the task completion status, and environmental perception data. After receiving the execution logs, the cloud data center uses the deviation between the actual cleanup data and the predicted values ​​to periodically optimize the weight parameters in the dynamic priority scoring model, the network parameters of the prediction model, and the adjustment coefficients in the response scoring function. It should be noted that the local task reassignment mechanism only re-optimizes the affected tasks to avoid large-scale disturbances caused by global path replanning; the determination of nearby vehicles is based on geographic grid division and real-time location comparison to ensure that task handover is spatially feasible and economical; the closed-loop feedback link formed by the execution data feedback provides real operating samples for the self-learning of model parameters, promotes the evolution of the system from "experience-driven" to "data-driven", and continuously improves the overall intelligence level of scheduling.

[0029] This embodiment also provides a computer device applicable to the Internet of Things-based urban waste sorting, collection, and scheduling method, including: a memory and a processor; the memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions to realize the Internet of Things-based urban waste sorting, collection, and scheduling method proposed in the above embodiment.

[0030] The computer device can be a terminal, comprising a processor, memory, communication interface, display screen, and input devices connected via a system bus. The processor provides computing and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs stored in the non-volatile storage media. The communication interface is used for wired or wireless communication with external terminals; wireless communication can be achieved through Wi-Fi, carrier networks, NFC (Near Field Communication), or other technologies. The display screen can be an LCD screen or an e-ink screen. The input devices can be a touch layer covering the display screen, buttons, a trackball, or a touchpad on the computer device's casing, or an external keyboard, touchpad, or mouse.

[0031] This embodiment also provides a storage medium on which a computer program is stored. When executed by a processor, the program implements the IoT-based urban waste sorting, collection, and scheduling method proposed in the above embodiments. The storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as Static Random Access Memory (SRAM), Electrically Erasable Programmable Read-Only Memory (EEPROM), Erasable Programmable Read Only Memory (EPROM), Programmable Red-Only Memory (PROM), Read-Only Memory (ROM), magnetic storage, flash memory, magnetic disk, or optical disk.

[0032] In summary, this invention achieves a paradigm shift in urban waste sorting and collection from passive response to proactive prediction and from centralized control to distributed autonomy by constructing an integrated technical architecture of edge dynamic assessment, environmental fusion prediction, and multi-agent collaborative scheduling. It utilizes edge gateways to perform localized priority calculations on waste bin status, reducing system response latency and improving the timeliness of reporting high-priority events. By combining multi-source external factors such as weather, festivals, and pedestrian density, a long short-term memory neural network prediction model is constructed, effectively enhancing the forward-looking identification capability of potential high-risk points and avoiding sudden overflow events. Furthermore, by modeling collection vehicles as agents with bidding and negotiation capabilities and introducing a task allocation mechanism based on response scoring, dynamic resource matching and flexible path optimization are achieved, improving the accuracy, real-time performance, and resource utilization of collection operations while reducing operating costs and environmental risks.

[0033] It should be noted that 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 preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A method for urban waste sorting, collection, and scheduling based on the Internet of Things, characterized by: include: Smart trash cans with filling rate and environmental parameter sensing capabilities are deployed in various urban areas, and the collected status data is uploaded to the edge gateway of the area through a low-power wide area network. The edge gateway performs localized priority assessment based on the real-time status data of each smart trash can received, combined with its capacity information, last collection time and location attributes, and generates a dynamic priority score for each trash can. When the dynamic priority score of any trash can exceeds the preset response threshold, the edge gateway triggers the event reporting mechanism and sends a collection alarm containing location and priority information to the regional dispatch center. Based on receiving collection and disposal alarms, the regional dispatch center integrates weather forecasts, holiday calendars, and regional population density data to predict trends for trash cans that have not yet received alarms but pose a high risk of overflowing, thus identifying potential high-risk points. The emergency tasks corresponding to the cleaning alarms and the predicted potential high-risk tasks are combined to form a set of tasks to be scheduled, and this set of tasks is sent to each vehicle intelligent agent participating in the cleaning operation. Each vehicle intelligent agent assesses its response capability and makes bidding decisions for each task in the task set based on its current location, loading status, and driving energy consumption model. The regional dispatch center allocates tasks based on the bidding results of each intelligent agent and generates dynamic dispatch instructions. Each vehicle intelligent agent provides real-time feedback on its operating status during the execution of scheduling instructions. When a task is delayed or a new high-priority alarm is added, the system performs a local task reallocation based on the current task allocation status and the vehicle's location. After completing the cleaning task, the system sends the execution data back.

2. The urban waste sorting, collection, and scheduling method based on the Internet of Things as described in claim 1, characterized in that: The deployment of smart trash cans with filling rate and environmental parameter sensing capabilities in various urban areas, and the uploading of collected status data to the edge gateway of the respective area via a low-power wide-area network, involves the following steps: Multiple smart trash cans are deployed within the designated administrative or functional areas of the city. Each smart trash can is equipped with a non-contact distance sensor to measure the real-time distance between the trash accumulation surface inside the can and the top of the can. The smart trash can has built-in temperature and humidity sensors to collect microenvironmental parameters inside the can. The smart trash can is equipped with a communication module, which encapsulates real-time measured distance data, environmental parameters, and unique device identification information into data packets according to a preset communication protocol; The data packets are transmitted via a low-power wide-area network to an edge gateway covering the geographical area. The edge gateway receives and parses the data packets to extract the original measurement values ​​of each smart trash can.

3. The urban waste sorting, collection, and scheduling method based on the Internet of Things as described in claim 2, characterized in that: The edge gateway, based on the received real-time status data of each smart trash can, combined with its capacity information, last collection time, and location attributes, performs localized priority assessment to generate a dynamic priority score for each trash can. The specific steps are as follows: The edge gateway calculates the ratio of the current filling height to the total height based on the received real-time distance value and the pre-stored total height of the enclosure, and records this ratio as the filling rate. ; Retrieve the timestamp of the last time the smart trash can was emptied, calculate the time interval from the time the emptying was completed to the current time, and record it as the dwell time. ; The smart trash can is matched to its corresponding area type based on its geographical coordinates. These area types include medical areas, commercial areas, transportation hub areas, and residential areas, with different area weight values ​​corresponding to different area types. Among them, the regional weight of medical and commercial areas is higher than that of residential areas; Based on fill rate Duration of stay and regional weights Dynamic priority scores are generated through weighted combination. The calculation process is as follows: ; in, , , , For the preset adjustment parameters, This generates a dynamic priority score, which reflects the urgency of clearing the trash can.

4. The urban waste sorting, collection, and scheduling method based on the Internet of Things as described in claim 3, characterized in that: When the dynamic priority score of any trash can exceeds a preset response threshold, the edge gateway triggers an event reporting mechanism to send a collection alarm containing location and priority information to the regional dispatch center. The specific steps are as follows: The edge gateway will use the currently calculated dynamic priority score. With the first preset threshold Compare; like > If the trash can is deemed to have entered a high-priority state, a collection alarm event will be generated. The waste collection alarm event includes the unique identifier of the waste bin, its geographical coordinates, and its current fill rate. Dynamic priority score and the time of event generation; The edge gateway uploads the cleaning alarm events to the regional dispatch center through a secure communication link and records the reporting time locally to prevent the same event from being reported repeatedly within a set cooling period.

5. The urban waste sorting, collection, and scheduling method based on the Internet of Things as described in claim 4, characterized in that: Based on receiving collection alarms, the regional dispatch center integrates weather forecasts, holiday calendars, and regional pedestrian density data to predict trends for trash cans that have not yet received alarms but pose a high risk of overflow, identifying potential high-risk points. The specific steps are as follows: The regional dispatch center receives meteorological data for a future preset time window from the meteorological service platform. The meteorological data includes the probability of rainfall, average temperature, and air humidity. Obtain festival information output from the city's public calendar system, wherein the festival information is used to identify whether the current date is a statutory holiday, a local event day, or a public holiday; Receive the regional population density index provided by the urban sensing network. The regional population density index is calculated based on mobile communication signals, Wi-Fi probes or video analysis technology, and reflects the degree of population gathering in the target area per unit time. For smart trash cans that have not yet triggered a collection alarm, retrieve their filling rate change sequence over a preset period of time. The filling rate change sequence consists of historical filling rate data periodically reported by the edge gateway. The fill rate change sequence is spatiotemporally aligned with the received meteorological data, festival information and regional population density index to ensure that all data are synchronized in the time dimension and matched to the same geographic grid unit in the spatial dimension. Construct a multidimensional input feature vector, which includes historical fill rate sequence, current rainfall probability, temperature value, festival status and regional population density value; The multidimensional input feature vector is input into a pre-trained long short-term memory neural network prediction model. This model has a memory unit structure, which can capture long-term dependencies in time series and model the nonlinear effects of external environmental variables. The long short-term memory neural network prediction model outputs a predicted value of the filling rate within a preset time window in the future. The predicted value represents the expected degree of garbage accumulation in the trash can at a future time. Based on the predicted filling rate, combined with the capacity parameters, area attribute weights and retention time of the trash can, its predicted dynamic priority score for future moments is calculated. If the predicted dynamic priority score exceeds the preset potential risk threshold, it is determined that the trash can has a high risk of filling in the future, and it is marked as a potential high-risk point. A corresponding predictive collection task is generated, which includes the task location, the predicted priority level, and the suggested response time window.

6. The urban waste sorting, collection, and scheduling method based on the Internet of Things as described in claim 5, characterized in that: The steps for constructing a task set to be scheduled by combining the emergency tasks corresponding to the cleaning alarms with the predicted potential high-risk tasks, and then distributing this task set to the intelligent agents of each vehicle participating in the cleaning operation, are as follows: The regional dispatch center defines all tasks reported by the edge gateways and whose dynamic priority scores exceed the first response threshold as Level 1 tasks, which have the highest scheduling priority. The tasks generated from potential high-risk points identified by the prediction model are defined as secondary tasks, which are used to intervene in advance in case of overflow events. Assign a unique task identifier to each task and record its geographical coordinates, task type level, suggested collection time window and associated trash can number; All primary and secondary tasks are aggregated to form a set of tasks to be scheduled. Logically, this set of tasks to be scheduled constitutes a task pool that can be accessed by multiple vehicle intelligent agents. The regional dispatch center periodically sends the set of tasks to be dispatched in a structured data format to the currently online vehicle intelligent agents that are capable of operation through a message broadcasting mechanism. Each garbage truck's onboard terminal receives and parses the task set data, and updates the locally stored task list; The frequency of task distribution is dynamically adjusted based on the task density in the urban area, with a higher broadcast frequency used in high-density areas.

7. The urban waste sorting, collection, and scheduling method based on the Internet of Things as described in claim 6, characterized in that: Each vehicle agent assesses its response capability and makes bidding decisions for each task in the task set based on its current location, loading status, and driving energy consumption model. The regional dispatch center allocates tasks based on the bidding results of each agent and generates dynamic dispatch instructions. The specific steps are as follows: Each vehicle intelligent agent obtains its current geographical location, which is provided by the vehicle's GPS or BeiDou positioning module; Read the current load value fed back by the vehicle weighing system and compare it with the vehicle's maximum rated load to determine whether it has the loading capacity to undertake new tasks; For each task in the set of tasks to be scheduled, calculate the shortest path distance from the vehicle's current location to the task's target location. The path calculation is based on the urban road network topology and real-time traffic status data. Based on the path distance and road traffic speed model, estimate the estimated arrival time of the vehicle to the task point; Establish a vehicle driving energy consumption model, which takes into account the vehicle's weight, current load, road gradient, and driving speed, and calculate the increase in energy consumption per unit distance for performing the task. Construct a task response scoring function to quantify the vehicle's willingness to respond to the task and its execution efficiency. The expression is as follows: ； in, This indicates the vehicle's response score for a specific task. Indicates the estimated arrival time. This represents the dynamic priority score of the task. This indicates the increase in energy consumption per unit distance resulting from performing this task. , , The pre-set scoring weight coefficients are used to adjust the influence of time efficiency, task priority, and energy cost on the scoring. Based on the above scoring function, the vehicle intelligent agent calculates the response score for each executable task and selects the task with the highest score for bidding. Each vehicle will encrypt and upload its bidding information to the regional dispatch center. The bidding information includes the task identifier, response score, and estimated arrival time. The regional dispatch center receives bidding data from all vehicles and, in conjunction with the credit score generated from the vehicle's historical performance record, performs weighted processing on the response score. A task allocation optimization algorithm is used to assign each task to the vehicle with the highest overall score; The system generates dynamic scheduling instructions that include task sequences, driving routes, and time nodes, and sends them to each vehicle agent via a secure communication link.

8. The method for urban waste sorting, collection, and scheduling based on the Internet of Things as described in claim 7, characterized in that: Each vehicle intelligent agent provides real-time feedback on its operating status during the execution of scheduling instructions. When a task delay occurs or a new high-priority alarm is added, a local task reallocation is performed based on the current task allocation status and the vehicle's location. After completing the cleaning task, the execution data is transmitted back. The specific steps are as follows: During the execution of scheduling instructions, the vehicle intelligent agent uploads its current location coordinates, task execution progress, remaining loading capacity, and vehicle operating status to the regional scheduling center at fixed time intervals. The regional dispatch center continuously monitors the estimated arrival time of each vehicle. If the estimated arrival time of a vehicle for an assigned task exceeds the preset allowable deviation range, it is determined to be a task execution delay. When the edge gateway reports a new waste collection alarm, the regional dispatch center immediately queries the task status of the corresponding waste bin to determine whether it has been assigned and has not yet been completed. If the task has not yet been assigned, or the originally assigned vehicle is in a task delay state, the task will be released back into the set of tasks to be scheduled, forming a temporary schedulable task. The regional dispatch center notifies the vehicle agents currently located in the geographical vicinity of the task, triggering a local reassessment process. The nearby vehicle agent reacquires the task information and, based on its current location and loading status, recalculates the response score for the task according to the response scoring function; Each nearby vehicle will upload its updated response score to the regional dispatch center; Based on the newly reported scoring results, the regional dispatch center re-executes the task allocation logic and assigns the task to the current best vehicle. Generate local adjustment instructions and issue them to the newly assigned vehicle and the original executing vehicle to achieve seamless handover of tasks; After receiving the adjustment instructions, the newly assigned vehicle updates its local mission plan and plans its route to the new mission location. After receiving the task release command, the original vehicle clears its local task record and continues to execute subsequent tasks or returns to the dispatch center. After all the waste collection tasks are completed, the vehicle's intelligent agent will package the task execution log and upload it to the cloud data center. The execution log includes the actual collection time, the amount of waste collected, the driving route trajectory, the task completion status, and environmental perception data. After receiving the execution logs, the cloud data center uses the deviation between the actual cleanup data and the predicted values ​​to periodically optimize the weight parameters in the dynamic priority scoring model, the network parameters of the prediction model, and the adjustment coefficients in the response scoring function.

9. A computer device, comprising a memory and a processor, wherein the memory stores a computer program, characterized in that: When the processor executes the computer program, it implements the steps of the Internet of Things-based urban waste sorting, collection, and scheduling method according to any one of claims 1 to 8.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by the processor, it implements the steps of the Internet of Things-based urban waste sorting, collection, and scheduling method as described in any one of claims 1 to 8.

Images