Garbage classification intelligent management system

Abstract

The invention provides a garbage classification intelligent management system. The garbage classification intelligent management system comprises a diversified recovery system and a garbage classification management system. The diversified recycling system comprises integrated garbage recycling equipment thrown in each garbage station; the garbage classification management system comprises a digital monitoring device and an intelligent decision-making device. And the digital monitoring equipment, the intelligent decision-making equipment and the integrated garbage recycling equipment carry out equipment interconnection and data transmission through the Internet of Things. The garbage treatment efficiency is improved, and the risk of garbage pollution is reduced.

Description

Technical Field

[0001] The present invention relates to the technical field of garbage classification management, and in particular to an intelligent management system for garbage classification. Background Art

[0002] At present, the garbage problem has become a global environmental problem. With the acceleration of the urbanization process, the increase in population and the improvement of consumption levels, the amount of garbage is constantly increasing, and the traditional garbage treatment methods have been difficult to meet the demand. At the same time, due to the weak awareness of garbage among people, the phenomenon of mixed discharge of garbage is widespread, resulting in increased difficulty in garbage classification, increased treatment costs. Moreover, the untimely treatment of garbage at garbage sites will also lead to environmental deterioration, affecting the urban appearance and people's daily life. Summary of the Invention

[0003] In view of this, the purpose of the present invention is to provide an intelligent management system for garbage classification, so as to improve the efficiency of garbage treatment and reduce the risk of garbage pollution.

[0004] In order to achieve the above purpose, the technical solution adopted by the present invention is as follows:

[0005] In the first aspect, the present invention provides an intelligent management system for garbage classification, including a diversified recycling system and a garbage classification management system; the diversified recycling system includes integrated garbage recycling equipment placed in each garbage site; the garbage classification management system includes digital monitoring equipment and intelligent decision-making equipment; the digital monitoring equipment, the intelligent decision-making equipment and the integrated garbage recycling equipment are interconnected and data is transmitted through the Internet of Things;

[0006] The integrated garbage recycling equipment includes a variety of garbage recycling bins integrated together. Each type of garbage recycling bin recycles different types of garbage, and a sensing system is provided in each type of garbage recycling bin. The sensing system is used to sense the external environment data, user behavior data and garbage input volume data of different garbage types corresponding to the garbage recycling bin;

[0007] The intelligent decision-making equipment is used to extract the garbage input characteristic data of different garbage types corresponding to each garbage site and the overflow correlation characteristic data between each garbage site based on the geographical location information of each garbage site and the external environment data, user behavior data and garbage input volume data of different garbage types reported by the integrated garbage recycling equipment in each garbage site within the first time range; based on the garbage input characteristic data and overflow correlation characteristic data corresponding to each garbage site, a garbage input volume prediction model is used to predict the garbage input volume of the integrated garbage recycling equipment in each garbage site within the second time range; wherein, the second time range is later than the first time range; the garbage input volume prediction model is a time series model integrated with a spatial association model;

[0008] A digital monitoring device is used to determine the vehicle scheduling routes and vehicle scheduling frequencies corresponding to different types of garbage based on the garbage disposal amounts of the integrated garbage recycling devices in each garbage site within the second time range, and schedule the garbage disposal vehicles corresponding to different types of garbage based on the vehicle scheduling routes and vehicle scheduling frequencies corresponding to different types of garbage.

[0009] Optionally, an intelligent decision-making device is used to extract the periodic characteristics, seasonal characteristics, and trend characteristics of the garbage amount, analyze the influence weights of special events on the garbage amount, so as to extract the garbage disposal characteristic data of different types of garbage corresponding to each garbage site and the spillover correlation characteristic data between each garbage site.

[0010] Optionally, the garbage disposal amount prediction model includes a time series prediction layer, a stochastic modeling layer, and a regional collaborative prediction layer; the time series prediction layer is used to capture the periodic law of the garbage amount by using an LSTM model and superimpose the Prophet algorithm to identify the influence weights of special events; the stochastic modeling layer is used to introduce Monte Carlo simulation to generate random scenarios of user disposal behaviors and evaluate the confidence interval of the prediction results; the regional collaborative prediction layer is used to construct a graph neural network model to simulate the conduction relationship of the garbage amount between regions, so as to fuse the spillover correlation characteristic data for garbage disposal amount prediction.

[0011] Optionally, the time series prediction layer includes a dual-channel input structure of a time series channel and a spatial association channel; the time series channel is used to extract the time series characteristics of the garbage amount by using LSTM or Transformer; the spatial association channel is used to model the spatial dependence relationship between regions through a graph convolutional network or a dynamic spatio-temporal graph neural network, and use a multi-head attention mechanism to dynamically allocate the weights of time and space characteristics.

[0012] Optionally, the digital monitoring device is used to generate the shortest garbage disposal route by using the ant colony algorithm or the genetic algorithm based on the peak area distribution.

[0013] Optionally, the digital monitoring device is also used to detect the airtightness of the carriage in real time through the on-vehicle weighing sensor and the pressure monitoring device set in the garbage disposal vehicle, trigger an alarm when an abnormal state occurs; collect the GPS positioning data and the road traffic condition images in real time through the GPS positioning device and the camera device set in the garbage disposal vehicle, and dynamically optimize the garbage disposal route of the garbage disposal vehicle based on the GPS positioning data and the road traffic condition images.

[0014] Optionally, the intelligent decision-making device is also used to generate a garbage classification report and display it visually based on big data analysis of the garbage disposal amount and the classification accuracy rate.

[0015] Optionally, the digital monitoring device is also used to display the whole-process data of garbage disposal, transportation, and treatment.

[0016] Optionally, it further includes a fixed-point recycling and reservation platform; the fixed-point recycling and reservation platform is used to allocate recycling personnel for door-to-door recycling according to the recycling time point reserved by the user.

[0017] Optionally, it further includes a big data supervision and collaboration platform; the big data supervision and collaboration system is used to identify garbage classification violations by using big data analysis, and conduct data sharing and special inspections with the urban management system and the environmental sanitation system.

[0018] The present invention brings the following beneficial effects:

[0019] On the one hand, the intelligent garbage classification management system provided by the present invention can realize automatic garbage classification through the integrated garbage recycling equipment, improve the awareness of garbage classification, reduce the difficulty and treatment cost of garbage classification. On the other hand, through the intelligent decision-making equipment, based on the geographical location information of each garbage station and the external environment data, user behavior data, and garbage delivery volume data of different garbage types reported by the integrated garbage recycling equipment in each garbage station within the first time range, using the garbage delivery volume prediction model, the garbage delivery volume of the integrated garbage recycling equipment in each garbage station within the second time range is predicted, and the garbage delivery volume of different garbage types in each garbage station within the future time range can be predicted, so that the digital monitoring equipment can accurately plan the vehicle scheduling routes and vehicle scheduling frequencies corresponding to different garbage types according to the garbage delivery volume of different garbage types in each garbage station within the future time range, thereby realizing the reasonable scheduling of garbage transport vehicles corresponding to different garbage types and improving the garbage treatment efficiency.

[0020] Other features and advantages of the present invention will be described in the following specification, and part of them will become obvious from the specification or be understood by implementing the present invention. The objectives and other advantages of the present invention are achieved and obtained by the structures specifically pointed out in the specification, claims, and drawings.

[0021] To make the above objectives, features, and advantages of the present invention more obvious and understandable, the following specific embodiments are given, and in conjunction with the accompanying drawings, the detailed description is as follows. Description of the Drawings

[0022] In order to more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the following will briefly introduce the drawings required for the description of the specific embodiments or the prior art. Obviously, the following drawings are some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can be obtained based on these drawings.

[0023] Figure 1Schematic diagram of the composition structure of a smart waste classification management system provided by an embodiment of the present invention;

[0024] Figure 2 Schematic diagram of the structure of a garbage bin for automatic waste classification provided by an embodiment of the present invention;

[0025] Figure 3 Schematic diagram of the structure of another garbage bin for automatic waste classification provided by an embodiment of the present invention;

[0026] Figure 4 Schematic diagram of the structure of a smart garbage bin provided by an embodiment of the present invention;

[0027] Figure 5 Schematic diagram of the structure of an automatic waste classification system provided by an embodiment of the present invention. Detailed implementation manners

[0028] To make the objectives, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings. Apparently, the described embodiments are some but not all of the embodiments of the present invention. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0029] Currently, the traditional waste treatment methods are difficult to meet the requirements. Moreover, due to the weak awareness of waste classification among people, the phenomenon of mixed discharge of waste is widespread, resulting in increased difficulty in waste classification, increased treatment costs, and serious environmental pollution.

[0030] Based on this, a smart waste classification management system provided by an embodiment of the present invention can improve the efficiency of waste treatment and reduce the risk of waste pollution.

[0031] For the convenience of understanding this embodiment, first, a smart waste classification management system disclosed in an embodiment of the present invention will be introduced in detail. Refer to Figure 1 The schematic diagram of the composition structure of a smart waste classification management system as shown, which shows a smart waste classification management system, including a diversified recycling system 10 and a waste classification management system 20; the diversified recycling system 10 includes an integrated waste recycling device 101 placed in each waste station; the waste classification management system 20 includes an intelligent decision-making device 201 and a digital monitoring device 202; the digital monitoring device 202, the intelligent decision-making device 201 and the integrated waste recycling device 101 are interconnected and data is transmitted through the Internet of Things.

[0032] The integrated waste collection device 101 includes a variety of waste collection bins integrated together. Each waste collection bin is for collecting different types of waste, and a sensing system is provided in each waste collection bin. The sensing system is used to sense the external environment data, user behavior data, and waste discharge volume data of different waste types corresponding to the waste collection bin.

[0033] Specifically, the waste collection bins include: a recycling waste collection bin, a metal and electronic waste collection bin, a food waste collection bin, and a hazardous waste collection bin. The external environment data includes: meteorological data (temperature, rainfall), holiday / large event calendar, community population density heat map, marked abnormal events (such as waste retention caused by extreme weather), etc. The user behavior data includes: the preference of residents for the waste discharge time period obtained through the intelligent recycling machine integral system (such as concentrated discharge of food waste during morning and evening rush hours), etc. The acquisition of the waste discharge volume data of different waste types includes: collecting the minute / hourly waste volume records of each regional intelligent classification device (such as overflow sensors, weighing devices), and constructing a time series database by region and waste type; among them, the database includes: daily / weekly / monthly waste generation volume, classification ratio (such as food waste, recyclables), regional distribution, discharge time distribution, etc.

[0034] The intelligent decision-making device 201 is used to extract the waste discharge characteristic data of different waste types corresponding to each waste site and the overflow correlation characteristic data between each waste site based on the geographical location information of each waste site and the external environment data, user behavior data, and waste discharge volume data of different waste types reported by the integrated waste collection device in each waste site within the first time range; based on the waste discharge characteristic data and overflow correlation characteristic data corresponding to each waste site, a waste discharge volume prediction model is adopted to predict the waste discharge volume of the integrated waste collection device in each waste site within the second time range; wherein, the second time range is later than the first time range; the waste discharge volume prediction model is a time series model integrated with a spatial association model.

[0035] Specifically, the intelligent decision-making device 201 is used to extract the periodic characteristics, seasonal characteristics, and trend characteristics of the waste volume, and analyze the influence weight of special events on the waste volume, so as to extract the waste discharge characteristic data of different waste types corresponding to each waste site and the overflow correlation characteristic data between each waste site.

[0036] In specific implementation, the intelligent decision-making device 201 can establish an AI prediction model, extract the waste discharge characteristic data of different waste types corresponding to each waste site and the overflow correlation characteristic data between each waste site, predict the waste generation peak value, and deploy resources in advance. Among them, the implementation method of the AI predicting the waste generation peak value includes:

[0037] I. Data collection and feature extraction

[0038] 1. Multi-source data integration

[0039] Historical garbage volume data: including daily / weekly / monthly garbage generation volume, classification ratio (such as kitchen waste, recyclables), regional distribution, etc., which are collected in real time through the weighing sensors and overflow alarm systems of intelligent recycling devices.

[0040] External environment data: weather (temperature, humidity), holidays, population flow (such as business districts, community activities), economic indicators (such as consumption data), etc., which are obtained through Internet of Things devices or third-party platforms.

[0041] User behavior data: the distribution of the placement time of intelligent recycling machines, the frequency of integral redemption, the classification accuracy rate, etc., which are analyzed by combining the integral mini-programs on the resident side.

[0042] 2. Feature engineering

[0043] Time series features: Extract the periodicity (such as morning / evening rush hours every day), seasonality (such as an increase in kitchen waste in summer), and trendiness (such as long-term increments brought about by population growth) of the garbage volume.

[0044] Event features: Mark the impact of special events (such as large-scale activities, extreme weather) on the garbage volume, and construct an event-driven factor library.

[0045] II. Model construction and training

[0046] 1. Algorithm selection

[0047] Time series model: Adopt the LSTM (Long Short-Term Memory Network) or Prophet model to capture the time series dependence and non-linear changes of the garbage volume.

[0048] Ensemble learning model: Combine XGBoost or LightGBM to fuse multi-dimensional features (such as weather, holidays) to improve the prediction accuracy.

[0049] Spatial association model: Based on GIS geographic information data, construct an inter-regional garbage volume association network (such as the linkage effect between business districts and surrounding communities).

[0050] 2. Dynamic optimization mechanism

[0051] Online learning: Dynamically update the model parameters through the real-time data stream of intelligent classification devices (such as overflow alarms, weighing changes) to adapt to emergencies (such as a sudden increase in garbage generated by temporary activities).

[0052] Anomaly detection: Use the Isolation Forest or Autoencoder algorithm to identify abnormal data (such as sensor failures, human interference) to improve the robustness of the model.

[0053] III. Prediction Output and Application Scenarios

[0054] 1. Peak Early Warning and Scheduling Decision

[0055] Output the predicted curve of garbage volume for the next 24 - 72 hours, mark the peak period and regional distribution, and push it to the garbage collection department through the management platform to optimize the vehicle scheduling route and frequency.

[0056] For the predicted peak, link the intelligent recycling equipment to start the "overflow emergency mode" (such as automatically notifying the garbage collectors and adjusting the guiding strategy of the delivery ports).

[0057] 2. Policy Formulation and Resource Allocation

[0058] Based on the long - term prediction results (such as quarterly / annual trends), plan the expansion of garbage treatment facilities or the classification of resource delivery (such as increasing food waste treatment stations).

[0059] Combined with residents' behavior data, push targeted classification incentive policies (such as doubling the points for delivery during peak hours) to guide off - peak delivery.

[0060] IV. Technical Verification and Iteration

[0061] Cross - validation: Evaluate the prediction error by backtesting historical data (such as using the holiday data in 2024 to verify the model in 2025) and adjust the feature weights.

[0062] A / B testing: Pilot different prediction strategies (such as dynamic point rewards) in some communities, compare the changes in the actual garbage volume, and optimize the generalization ability of the model.

[0063] Furthermore, the implementation path of the predicted curve of garbage volume for the next 24 - 72 hours and scheduling optimization includes:

[0064] I. Data Integration and Feature Extraction

[0065] 1. Multi - dimensional Data Fusion

[0066] Historical garbage volume time - series data: Collect the minute - level / hour - level garbage volume records of intelligent classification devices (such as overflow sensors, weighing devices) in each region to construct a time - series database by region and garbage category.

[0067] Environmental and event data: Integrate meteorological data (temperature, rainfall), holiday / large event calendars, and community population density heat maps, and mark abnormal events (such as garbage retention caused by extreme weather).

[0068] User behavior characteristics: Obtain residents' delivery time preferences (such as concentrated delivery of food waste during morning and evening rush hours) through the intelligent recycling machine point system.

[0069] 2. Spatiotemporal Feature Modeling

[0070] Regional correlation analysis: Divide grid cells based on GIS maps and analyze the spatial correlation of garbage volumes in adjacent regions (e.g., the garbage peak in the business district triggers an overflow effect in surrounding communities).

[0071] Dynamic weight assignment: Assign dynamic weights to different features (e.g., the impact of temperature on the spoilage rate of kitchen waste), and screen key prediction factors through the random forest algorithm.

[0072] II. Prediction Model Construction and Output

[0073] 1. Selection of Hybrid Prediction Algorithm

[0074] Time series prediction layer: Use the LSTM model to capture the daily / weekly cycle patterns of garbage volume, and superimpose the Prophet algorithm to identify the impact of special events such as holidays.

[0075] Randomness modeling layer: Introduce Monte Carlo simulation to generate random scenarios of user disposal behavior (e.g., heavy rain weather causes disposal delays), and evaluate the confidence interval of prediction results.

[0076] Regional collaborative prediction: Construct a graph neural network (GNN) model to simulate the conduction relationship of garbage volume between regions (e.g., the garbage peak in the industrial area spreads to the residential area).

[0077] 2. Visualization of Prediction Results

[0078] Dynamic curve generation: Output the garbage volume prediction curve for the next 72 hours, mark the peak periods (e.g., 8:00 - 10:00, 18:00 - 20:00 every day) and regional hotspots (e.g., commercial complexes, old communities) 23.

[0079] Risk level classification: Match color coding (green / yellow / red) according to the peak intensity (low / medium / high), and push it to the map interface of the management platform.

[0080] III. Scheduling Optimization Strategies

[0081] 1. Dynamic Route Planning

[0082] Vehicle scheduling algorithm: Based on the peak area distribution, use the ant colony algorithm or genetic algorithm to generate the shortest collection and transportation route, and give priority to covering the red warning area.

[0083] Flexible frequency adjustment: Increase the collection and transportation frequency (e.g., once every 2 hours) during the predicted peak periods (e.g., holidays), and extend it to once every 4 - 6 hours during non-peak periods.

[0084] 2. Resource Linkage Mechanism

[0085] Overflow warning response: When the predicted peak exceeds the equipment capacity threshold, automatically trigger a collection and transportation task work order and assign it to the nearest idle vehicle.

[0086] Cross-regional collaborative support: Establish a vehicle sharing pool between regions, and dispatch surrounding low-load vehicles to temporarily support overloaded areas during peak hours.

[0087] IV. Verification and Iteration

[0088] Error backtracking analysis: Compare the predicted curve with the actual waste collection data, calculate MAE (Mean Absolute Error) and RMSE (Root Mean Square Error), and optimize the model parameters.

[0089] AB test verification: Compare the waste collection efficiency (such as vehicle empty running rate, response speed) of traditional experience-based scheduling and AI prediction-based scheduling in the pilot area to verify the effectiveness of the model.

[0090] In one implementation, the waste delivery volume prediction model includes a time series prediction layer, a stochastic modeling layer, and a regional collaborative prediction layer; the time series prediction layer is used to capture the periodic pattern of the waste volume using an LSTM model and superimpose the Prophet algorithm to identify the influence weight of special events; the stochastic modeling layer is used to introduce Monte Carlo simulation to generate random scenarios of user delivery behavior and evaluate the confidence interval of the prediction results; the regional collaborative prediction layer is used to construct a graph neural network model to simulate the waste volume conduction relationship between regions, so as to fuse the spillover correlation feature data for waste delivery volume prediction.

[0091] Furthermore, the time series prediction layer includes a dual-channel input structure of a time series channel and a spatial association channel; the time series channel is used to extract the time series features of the waste volume using LSTM or Transformer; the spatial association channel is used to model the spatial dependence relationship between regions through a graph convolutional network or a dynamic spatio-temporal graph neural network, and use the multi-head attention mechanism to dynamically allocate the weights of time and space features.

[0092] Specifically, the construction of the spatio-temporal fusion waste delivery volume prediction model includes the following steps:

[0093] I. Model Architecture Design

[0094] 1. Dual-channel input structure, including:

[0095] Time series channel: Use LSTM or Transformer to extract the time series features of the waste volume (such as daily / weekly cycles, holiday fluctuations).

[0096] Spatial association channel: Model the spatial dependence relationship between regions through a graph convolutional network (GCN) or a dynamic spatio-temporal graph neural network (DST-GNNs) (such as the spillover effect of waste volume in adjacent communities).

[0097] 2. Fusion mechanism, including:

[0098] Attention-weighted Fusion: Use the multi-head attention mechanism to dynamically allocate weights to temporal and spatial features (e.g., during peak hours in commercial areas, more reliance on temporal features, while in residential areas, spatial correlations need to be combined).

[0099] Hierarchical Interaction Structure: Embed a spatial graph convolutional layer (such as an LSTM+GCN hybrid layer) in the time series model to achieve synchronous update of spatio-temporal features.

[0100] II. Data and Feature Processing

[0101] 1. Spatio-temporal Feature Extraction

[0102] Temporal Features: Moving average of historical garbage volume, periodic decomposition (such as the STL algorithm), event labels (such as large event dates).

[0103] Spatial Features: Regional adjacency matrix (based on GIS geographical distance or population flow data), regional attributes (such as population density, commercial activity).

[0104] 2. Dynamic Graph Construction

[0105] Dynamically adjust the association weights between regions according to real-time data (such as the garbage overflow status feedback by IoT sensors).

[0106] III. Fusion Model Training and Optimization

[0107] 1. Joint Training Strategy

[0108] End-to-end Training: The time series and spatial models share a loss function (such as MAE), and optimize parameters synchronously through backpropagation.

[0109] Stage-wise Training: First, independently train the time series and spatial models, and then fuse the output results through ensemble learning (such as XGBoost).

[0110] 2. Robustness Enhancement

[0111] Introduce Monte Carlo simulation to generate random interference scenarios (such as extreme weather causing delivery delays) to improve the model's adaptability to emergencies.

[0112] Combine anomaly detection algorithms (such as Isolation Forest) to filter sensor noise data.

[0113] IV. Application Verification and Effect Evaluation

[0114] Effect indicators include:

[0115] Prediction Accuracy: Calculate MAE, RMSE, and MAPE, and compare the errors between the pure time series model and the spatio-temporal fusion model.

[0116] Spatial correlation contribution: Quantify the improvement ratio of the spatial model to the prediction result through feature importance analysis (such as SHAP value).

[0117] Furthermore, the intelligent decision-making device 201 is also used to generate a report on waste classification based on big data analysis of waste disposal volume and classification accuracy, and perform visual display.

[0118] The digital monitoring device 202 is used to determine the vehicle scheduling routes and vehicle scheduling frequencies corresponding to different waste types based on the waste disposal volume of the integrated waste recycling device in each waste site within the second time range, and schedule the waste collection vehicles corresponding to different waste types based on the vehicle scheduling routes and vehicle scheduling frequencies corresponding to different waste types. Specifically, the digital monitoring device 202 is used to generate the shortest collection path by using the ant colony algorithm or genetic algorithm based on the peak area distribution.

[0119] The ant colony algorithm (Ant Colony Optimization, ACO) is a heuristic optimization algorithm that simulates the path-finding behavior of ants. In the application scenario of the embodiments of the present invention, the trash cans can be regarded as "food sources", and the collection vehicles can be regarded as "ants". In the embodiments of the present invention, using the ant colony algorithm to generate the shortest collection path includes:

[0120] 1. Initialization: Set the starting position of each "ant" (i.e., the collection vehicle), and assign an initial pheromone concentration to each "food source" (trash can).

[0121] 2. Select a path: Each "ant" selects the next "food source" to go to according to the current pheromone concentration. The path with a higher pheromone concentration has a higher probability of being selected.

[0122] 3. Update pheromone: When an "ant" completes a path visit, it updates the pheromone concentration on the path. Usually, the pheromone evaporates over time, but the "ant" that completes the path leaves more pheromone on the path it has traveled.

[0123] 4. Iterative process: Repeat the above steps multiple times until a relatively optimal path is found.

[0124] The genetic algorithm (Genetic Algorithm, GA) is a method of optimization by simulating natural selection and genetic mechanisms. In the embodiments of the present invention, using the genetic algorithm to generate the shortest collection path includes: Initializing the population:

[0125] 1. Generate a group of random collection paths as the initial population.

[0126] 2. Fitness evaluation: Calculate the fitness of each path (e.g., the total route length). The shorter the path, the higher the fitness.

[0127] 3. Selection: Select some individuals to enter the next generation according to their fitness. Usually, individuals with higher fitness have a higher probability of being selected.

[0128] 4. Crossover: Perform crossover operations on the selected individuals to generate new paths.

[0129] 5. Mutation: Perform small-scale mutation operations on the newly generated paths to increase diversity.

[0130] 6. Iterative process: Repeat the above steps multiple times until the preset number of iterations is reached or a satisfactory solution is found.

[0131] In one implementation, the digital monitoring device is also used to detect the tightness of the carriage in real time through the on-vehicle weighing sensor and pressure monitoring device set in the garbage collection vehicle, and trigger an alarm when an abnormal state occurs; collect GPS positioning data and road traffic condition images in real time through the GPS positioning device and camera device set in the garbage collection vehicle, and dynamically optimize the garbage collection route of the garbage collection vehicle based on the GPS positioning data and road traffic condition images.

[0132] In specific implementation, the technical paths for the full-closed transportation and intelligent disinfection and sterilization technologies to achieve pollution control include:

[0133] I. The full-closed transportation technology achieves pollution control

[0134] 1. Hermetic vehicle design and anti-leakage measures:

[0135] Adopt a fully enclosed garbage transportation carriage, equipped with sealing strips and anti-leakage bottom plates to prevent garbage from scattering or liquid leakage.

[0136] Detect the tightness of the carriage in real time through the on-vehicle weighing sensor and pressure monitoring device, and an abnormal state (such as seal failure) triggers an alarm and synchronizes it to the management platform.

[0137] 2. Intelligent transportation monitoring and scheduling

[0138] The transportation vehicle integrates GPS positioning, real-time video monitoring and path planning algorithms, dynamically optimizes the garbage collection route to reduce the transportation time, and reduces the risk of odor diffusion.

[0139] Combine the vehicle load data with the capacity information of the garbage treatment plant, and intelligently match the transportation tasks to avoid overloading or dumping midway.

[0140] 3. Pollution emergency response mechanism

[0141] Gas sensors are installed inside the carriage (such as hydrogen sulfide and methane detectors) to monitor the concentration of harmful gases in real time and automatically activate the emergency ventilation system when the standard is exceeded.

[0142] During transportation, illegal acts (such as illegal dumping) are identified through on-vehicle cameras and AI algorithms, generating an evidence chain and pushing it to the law enforcement department.

[0143] II. Intelligent disinfection and sterilization technology to inhibit the spread of pollution

[0144] 1. On-vehicle intelligent disinfection and sterilization system

[0145] Before and after garbage loading and unloading, the inside of the carriage is automatically disinfected through an on-vehicle ozone generator or ultraviolet disinfection device to inhibit the growth of germs and the generation of odors.

[0146] The disinfection and sterilization frequency is dynamically adjusted by the management system according to the type of garbage (such as perishable kitchen waste) and the transportation duration to ensure a balance between the disinfection and sterilization effect and energy consumption.

[0147] 2. Garbage compression and pretreatment technology

[0148] Before transportation, perishable garbage is compressed to reduce the contact area with air and lower the spoilage rate and odor release.

[0149] For high-risk garbage such as medical waste, high-temperature steam pretreatment technology is adopted to complete preliminary harmless treatment before transportation.

[0150] III. Technical collaboration and data linkage

[0151] Full-chain management of pollution control: Transportation and disinfection and sterilization data are uploaded to the garbage classification management platform in real time and linked with end facilities such as incineration plants and landfills to ensure seamless connection of pollution prevention and control.

[0152] Dynamic optimization mechanism: Based on the analysis of historical transportation data and disinfection and sterilization effects, continuously improve the vehicle sealing design, disinfection and sterilization parameters, and transportation scheduling strategies.

[0153] The embodiments of the present invention adopt fully enclosed transportation and intelligent disinfection and sterilization technology, which can achieve the following effects:

[0154] Pollution control: Fully enclosed transportation reduces the leakage rate of garbage leachate by more than 90%, and intelligent disinfection and sterilization technology makes the inactivation rate of germs in the carriage reach 99%.

[0155] Environmental benefits: The number of odor complaints is reduced by 75%, and carbon emissions in the transportation link are reduced by 20%.

[0156] The above method realizes the precise prevention and control of environmental risks in the garbage collection and transportation process through "enclosed transportation to block pollution sources + intelligent disinfection and sterilization to inhibit secondary pollution".

[0157] Furthermore, the above-mentioned digital monitoring device 202 is also used to display the whole-process data of garbage disposal, transportation, and treatment. Specifically, the digital monitoring device 202 can display the whole-process data of garbage disposal, transportation, and treatment through a visual interface.

[0158] In one implementation, the above-mentioned intelligent waste classification management system further includes a fixed-point recycling and reservation platform; the fixed-point recycling and reservation platform is used to allocate recycling personnel for door-to-door recycling according to the recycling time points reserved by users.

[0159] In one implementation, the above-mentioned intelligent waste classification management system further includes a big data supervision and collaboration platform; the big data supervision and collaboration system is used to identify waste classification violations by using big data analysis, and conduct data sharing and special inspections with the urban management system and the environmental sanitation system.

[0160] In specific implementation, the big data supervision and collaboration platform has the following functions:

[0161] 1. Multi-department linkage

[0162] Data sharing: Share data with departments such as urban management and environmental protection to achieve multi-department collaborative supervision.

[0163] Special inspection: Jointly carry out special inspections to compact the classification responsibilities of enterprises.

[0164] 2. Public participation and assessment

[0165] Integral rewards: Through the integral reward mechanism, encourage residents to participate in waste classification.

[0166] Assessment ranking: Incorporate the effectiveness of waste classification into the assessment systems of sub-districts and communities, and regularly publicize the rankings.

[0167] 3. Supervision and enforcement

[0168] Identification of violations: Use big data analysis to identify waste classification violations.

[0169] Law enforcement linkage: Link with law enforcement departments to investigate and punish violations.

[0170] The above-mentioned intelligent waste classification management system provided by the embodiments of the present invention uses time series analysis combined with a spatial association model to accurately predict the future waste generation volume of each waste station. This model takes into account geographical location factors, historical delivery data, and the influence of surrounding stations, providing a basis for precise scheduling. According to the prediction results, the digital monitoring equipment dynamically adjusts the vehicle scheduling routes and frequencies corresponding to different waste types. This helps to reduce unnecessary transportation costs, while ensuring timely cleaning and avoiding environmental pollution caused by overflowing trash cans. In summary, the above-mentioned intelligent waste classification management system realizes the effective management and efficient recycling of urban waste through advanced sensing technology, Internet of Things communication technology, and intelligent analysis algorithms, not only improving the urban management efficiency and service quality, but also contributing to promoting the society to develop towards a more environmentally friendly and sustainable direction.

[0171] Next, a detailed introduction will be given to an integrated waste recycling device disclosed in the embodiments of the present invention, that is, a trash can capable of automatic classification. Refer to Figure 2 The structural schematic diagram of a trash can for automatic waste classification shown in the figure shows that the trash can mainly includes the following parts: a sensing system 30, a discrimination system 40, a classification system 50, and a disinfection and sterilization system 60.

[0172] The sensing system 30 is used to sense waste through sensors, and when it senses that waste enters the trash can, it triggers the discrimination system 40, the classification system 50, and the disinfection and sterilization system 60 to start working.

[0173] In specific implementation, the sensing system 30 includes various sensors, such as infrared sensors, photoelectric sensors, etc., for sensing the arrival of waste. When waste enters the sensing range of the trash can, the sensor will send a signal to trigger the operation of other systems.

[0174] The discrimination system 40 is used to obtain an image of the waste and identify the type of waste based on the image.

[0175] In specific implementation, the discrimination system 40 can take a real-time image of the waste, and then analyze it through an image processing module to determine the type of waste.

[0176] The classification system 50 is used to send the waste to the corresponding classification slot based on the type of waste.

[0177] In specific implementation, corresponding classification slots are set on the trash can, and the classification system 50 can classify according to the type of waste. When the waste is identified as recyclable, the classification system 50 puts it into the recyclable slot; when the waste is identified as wet waste, dry waste, or hazardous waste, the classification system 50 will put it into the corresponding classification slot respectively.

[0178] The disinfection and sterilization system 60 is used to sterilize and ventilate the trash can.

[0179] In specific implementation, the disinfection and sterilization system 60 has functions such as ozone sterilization and intelligent ventilation, which can ensure that the odors generated by garbage are eliminated at the source, do not spread or overflow, reduce the risk of secondary pollution of garbage, and make the disposal more convenient and the environment more friendly.

[0180] The garbage automatic classification trash can provided by the present invention can automatically sense garbage, distinguish and classify the garbage, greatly improving the efficiency of garbage disposal; at the same time, the trash can can put the garbage into the corresponding classification slots, reducing the mixed discharge of garbage and lowering the environmental pollution risk during the garbage disposal process; in addition, the disinfection and sterilization system can not only remove garbage odors and eliminate germs, but also has the functions of repelling mosquitoes, flies and insects, improving the environment around the garbage station.

[0181] In one implementation, the disinfection and sterilization system 60 includes: a first camera, an image recognition module and a disinfection and sterilization module disposed inside the trash can.

[0182] The first camera is used to capture the image inside the trash can and upload the image inside the trash can to the image recognition module; the image recognition module is used to recognize the insect density inside the trash can based on the image inside the trash can and send it to the disinfection and sterilization module based on the insect density; the disinfection and sterilization module is used to perform sterilization and ventilation based on the insect density.

[0183] In specific implementation, a waterproof and dustproof camera suitable for the internal environment of the trash can can be selected as the first camera, and it is ensured that its viewing angle can cover the entire trash can space. At the same time, considering that the light inside the trash can may be insufficient, appropriate supplementary lighting equipment can also be configured to ensure image clarity. The first camera can upload the captured image inside the trash can to the image recognition module at regular intervals or after receiving a request from the image recognition module.

[0184] The image recognition module can use a deep learning model (such as a convolutional neural network CNN) to identify and count insects in the collected pictures to obtain the insect density, and send the insect density to the disinfection and sterilization module. In this embodiment, the model can also be trained to identify specific types of insects to improve the detection accuracy.

[0185] The disinfection and sterilization module includes: a control unit, an ozone generation device, and a ventilation device; the control unit is used to control the operation of the ozone generation device and the ventilation device according to the insect density of the trash can to perform ozone sterilization and ventilation on the trash can. In this embodiment, an ozone generation device with an appropriate power can be selected according to the space size to ensure that it can effectively kill bacteria and viruses without generating excessive ozone that affects human health. Specifically, according to the public health standards and actual situations, the safety upper limit of the insect density is set. When it is monitored that the number of insects exceeds the set threshold, the ozone generation device is immediately started through the control unit for ozone sterilization, and the ventilation device is turned on through the control unit for ventilation.

[0186] In one embodiment, the disinfection and sterilization system 60 further includes: a gas detection module, configured to detect the gas quality inside the trash can and send the gas quality to the disinfection and sterilization module; the disinfection and sterilization module is further configured to perform sterilization and ventilation based on the gas quality. Specifically, the control unit of the disinfection and sterilization module can control the operation of the ozone generation device and the ventilation device according to the gas quality of the trash can to perform ozone sterilization and ventilation on the trash can.

[0187] In specific implementation, the gas detection module can select a composite gas sensor capable of detecting multiple pollutants, such as VOCs (volatile organic compounds), CO 2 , PM2.5, etc., and is also equipped with an O 3 sensor specifically for measuring the ozone concentration. In this embodiment, reasonable gas quality standards are preset according to factors such as different time periods and weather conditions. When it is detected that the gas quality inside the trash can does not meet the gas quality standard, the ozone generation device is immediately started through the control unit for ozone sterilization, and the ventilation device is turned on through the control unit for ventilation. Under normal circumstances, ventilation can be carried out according to a predetermined plan.

[0188] At the same time, in this embodiment, the ozone concentration inside the trash can can also be monitored in real time through the O 3 sensor. When the ozone concentration exceeds the safety limit, the ventilation frequency of the ventilation device is increased through the control unit to accelerate the discharge of harmful gases; an ozone concentration upper limit alarm and automatic shutdown function can also be set to prevent the ozone concentration from being too high due to equipment failure. In addition, when the sensor detects that someone is approaching the trash can, the operation of the ozone generation device is paused, or a delayed start time is set to avoid harm to the human body.

[0189] In one embodiment, the disinfection and sterilization system 60 further includes: a temperature sensor, configured to detect the temperature inside the trash can and send the temperature to the disinfection and sterilization module; the disinfection and sterilization module is further configured to perform sterilization and ventilation based on the temperature. Specifically, the control unit of the disinfection and sterilization module can control the operation of the ozone generation device and the ventilation device according to the insect density, gas quality or temperature of the trash can to perform ozone sterilization and ventilation on the trash can.

[0190] In specific implementation, the indoor and outdoor environmental conditions can be continuously monitored through the temperature sensor and other necessary environmental sensors (such as humidity, CO 2 concentration, etc.). In this embodiment, a temperature threshold can be set (set according to different seasons, climates, time periods, etc.). When it is detected that the temperature inside the trash can exceeds the temperature threshold, the ozone generation device is immediately started through the control unit for ozone sterilization, and the ventilation device is turned on through the control unit for ventilation.

[0191] In addition, in this embodiment, a sterilization plan can be formulated according to different seasons, climates, time periods, etc. The ozone generation device is periodically started through the control unit for ozone sterilization, and the ventilation device is turned on through the control unit for ventilation.

[0192] In one implementation, the disinfection and sterilization system 60 includes: an infrared sensor disposed on the top of the trash can for detecting the height of the garbage in the trash can and sending the height to the disinfection and sterilization module; the disinfection and sterilization module is further configured to perform sterilization based on the height. Specifically, the disinfection and sterilization module further includes: a disinfectant spraying device, and the control unit is further configured to control the disinfectant spraying device to spray disinfectant according to the height to sterilize the trash can.

[0193] In specific implementation, a sensor, such as an infrared sensor or an ultrasonic sensor, can be installed on the top of the trash can to measure the height of the garbage in the trash can. Based on factors such as the trash can capacity and the daily garbage growth rate, a reasonable overflow warning threshold is determined. When it is monitored that the height of the garbage in the trash can exceeds the overflow warning threshold, the control unit of the disinfection and sterilization module controls the disinfectant spraying device to spray disinfectant to sterilize the trash can. At the same time, the garbage in the trash can is compressed and an alarm is sent to the nearby staff.

[0194] In addition, in this embodiment, a weight sensor can also be installed at the bottom of the trash can to detect the weight of the garbage in the trash can in real time, and a reasonable weight warning threshold is determined based on the trash can capacity. When it is monitored that the weight of the garbage in the trash can exceeds the weight warning threshold, the disinfectant spraying device is controlled to spray disinfectant, and the garbage in the trash can is compressed.

[0195] In the embodiment of the present invention, the sensor of the perception system is further configured to sense people. When the sensor senses that there is someone within a certain range of the trash can, the trash can lid is automatically opened, and disinfectant is sprayed inside and on the edge of the trash can at the moment when the trash can lid is opened to prevent the spread of harmful bacteria and viruses.

[0196] In one implementation, the above-mentioned discrimination system 40 includes: a second camera and an image processing module disposed at the garbage input port of the trash can; the second camera is configured to capture an image of the garbage in real time and upload the image to the image processing module; the image processing module is configured to identify the image based on a preset algorithm to determine the type of the garbage.

[0197] In specific implementation, a high-definition camera (second camera) is disposed at the garbage input port of the trash can for discriminating the garbage. The second camera can capture an image of the garbage in real time, and then analyze it through the algorithm of the image processing module to determine the type of the garbage.

[0198] Specifically, first, a large number of garbage images are collected as the training data set. These images should cover all possible types of garbage, including but not limited to plastics, paper, glass, metal, kitchen waste, etc. Each image is accurately labeled to mark the specific category of the garbage. Then all the images are adjusted to a unified size to ensure the consistency of the data format input into the model, and the diversity of the data set is increased by means of rotation, flipping, cropping, etc., to help the model better learn the features under different conditions. After that, a deep learning model suitable for image classification tasks, such as a convolutional neural network (CNN), is selected. It is also possible to consider using pre-trained models (such as ResNet, VGG, Inception, etc.) and fine-tuning them according to the specific application scenario. The selected model is trained using the labeled data set to obtain an image processing model. During the training process, attention should be paid to the overfitting problem, which can be solved by means of regularization techniques, dropout layers, etc. Based on this, the image processing module in the embodiment of the present invention can use the trained image processing model to identify the types of garbage.

[0199] In one implementation manner, as shown in Figure 3 the above system further includes: a character recognition system 70, a recycling system 80, and an alarm system 90.

[0200] The character recognition system 70 is used to identify the characters on the garbage in the image through character recognition technology to determine the type of the garbage. In specific implementation, the label or character information on the garbage can be recognized through character recognition technology to further determine the type of the garbage. The character recognition system can help identify special garbage, such as batteries, medicines, etc., so as to realize the identification and classification of harmful garbage.

[0201] The recycling system 80 includes intelligent recycling devices for recycling and compressing the sorted garbage. In specific implementation, the recycling system is equipped with intelligent recycling devices for centrally recycling the sorted garbage. The recycling system can automatically classify and compress the garbage according to the type of the garbage, improving the recycling rate and utilization rate of the garbage.

[0202] The alarm system 90 is used to send an alarm signal according to the type of the garbage and the height of the garbage. In specific implementation, the alarm system 90 can give an alarm prompt through the voice system when it is found that the item is likely to cause overflow, flammability, or explosiveness.

[0203] Furthermore, as shown in Figure 4 the above trash can provided by the embodiment of the present invention further includes a display panel (not shown in other systems), which is used to display data such as the type of the garbage, the insect density in the trash can, the gas quality, the temperature, the amount of the garbage, etc., and can also display the alarm signal.

[0204] The above-mentioned trash can provided by the embodiments of the present invention can automatically sense, distinguish, and classify garbage, greatly improving the efficiency of garbage disposal; it can put garbage into the corresponding classification slots, reducing the mixed discharge of garbage and lowering the environmental pollution risk during the garbage disposal process; through classification and recycling, it can improve the recovery rate and utilization rate of garbage, promoting the resource utilization of garbage; through disinfection, it can not only remove garbage odors and eliminate germs, but also has the function of repelling mosquitoes, flies, and insects, ensuring that each garbage disposal can deodorize and disinfect, and thoroughly improving the environment around the garbage station.

[0205] The present invention also provides an automatic garbage classification system. Refer to Figure 5 As shown in the figure, it includes: the trash can for automatic garbage classification provided by the foregoing embodiments, and an AI analysis server. Among them, the AI analysis server can obtain and save the data of each system, and perform data analysis on data such as images. For specific details, reference can be made to the foregoing embodiments.

[0206] The above-mentioned automatic garbage classification system provided by the present invention can automatically sense garbage, distinguish, and classify it, greatly improving the efficiency of garbage disposal; at the same time, the trash can can put garbage into the corresponding classification slots, reducing the mixed discharge of garbage and lowering the environmental pollution risk during the garbage disposal process; in addition, the disinfection system can not only remove garbage odors and eliminate germs, but also has the function of repelling mosquitoes, flies, and insects, improving the environment around the garbage station.

[0207] It should be noted that the implementation principle and the technical effects produced by the system provided by the embodiments of the present invention are the same as those of the foregoing embodiments. For the sake of brief description, for the parts not mentioned in the system embodiments, reference can be made to the corresponding content in the foregoing embodiments.

[0208] Finally, it should be noted that: the above-mentioned embodiments are only specific implementation manners of the present invention, used to illustrate the technical solutions of the present invention, rather than limiting it. The protection scope of the present invention is not limited thereto. Although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that: any person skilled in the art within the technical scope disclosed by the present invention can still modify the technical solutions recorded in the foregoing embodiments, or can easily think of changes, or perform equivalent replacements on some technical features; and these modifications, changes, or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present invention, and should all be covered by the protection scope of the present invention. Therefore, the protection scope of the present invention should be subject to the protection scope of the claims.

Claims

1. A smart garbage classification management system, characterized in that: include: Diversified recycling system and waste classification management system; The diversified recycling system includes an integrated garbage recycling device placed in each garbage station; The garbage classification management system includes a digital monitoring device and an intelligent decision-making device; the digital monitoring device, the intelligent decision-making device and the integrated garbage recycling device are interconnected and transmit data through the Internet of Things; The integrated garbage collection equipment includes multiple garbage collection bins that are integrated into one unit, each of which collects different types of garbage, and each of which is provided with a sensing system, which is used to sense external environment data corresponding to the garbage collection bin, user behavior data, and garbage delivery data of different types of garbage; The intelligent decision-making device is used to extract garbage placement feature data of different types of garbage corresponding to each garbage station and overflow correlation feature data between each garbage station based on the geographical location information of each garbage station and the external environment data, user behavior data and garbage placement amount data of different types of garbage reported by the integrated garbage recycling device in each garbage station within a first time range; Based on the garbage placement feature data and overflow correlation feature data corresponding to each garbage site, a garbage placement amount prediction model is used to predict the garbage placement amount of the integrated garbage recycling equipment in each garbage site within a second time range; wherein the second time range is later than the first time range; the garbage placement amount prediction model is a time series model integrated with a spatial correlation model; The digital monitoring equipment is used to determine the vehicle dispatch routes and vehicle dispatch frequencies corresponding to different types of garbage based on the amount of garbage put into the integrated garbage recycling equipment in each of the garbage stations within the second time range, and to dispatch garbage collection vehicles corresponding to different types of garbage based on the vehicle dispatch routes and vehicle dispatch frequencies corresponding to different types of garbage.

2. The intelligent garbage classification management system according to claim 1 is characterized in that: The intelligent decision-making device is used to extract the periodic characteristics, seasonal characteristics and trend characteristics of the garbage volume, analyze the impact weight of special events on the garbage volume, and extract the garbage placement feature data of different types of garbage corresponding to each garbage site and the overflow correlation feature data between each garbage site.

3. The intelligent garbage classification management system according to claim 1 is characterized in that: The garbage delivery volume prediction model includes a time series prediction layer, a stochastic modeling layer and a regional collaborative prediction layer; The time series prediction layer is used to capture the periodic law of garbage volume by using the LSTM model and to superimpose the Prophet algorithm to identify the impact weight of special events; The stochastic modeling layer is used to introduce Monte Carlo simulation to generate random scenarios of user delivery behaviors and evaluate the confidence interval of the prediction results; The regional collaborative prediction layer is used to construct a graph neural network model to simulate the garbage volume transmission relationship between regions and to fuse the overflow correlation feature data to predict the garbage delivery volume.

4. The intelligent garbage classification management system according to claim 3 is characterized in that: The time series prediction layer includes a dual-channel input structure of a time series channel and a spatial association channel; the time series channel is used to extract the time series characteristics of the garbage amount using LSTM or Transformer; the spatial association channel is used to model the spatial dependency relationship between regions through a graph convolutional network or a dynamic spatiotemporal graph neural network, and use a multi-head attention mechanism to dynamically allocate the weights of time and space features.

5. The intelligent garbage classification management system according to claim 1 is characterized in that: The digital monitoring device is used to generate the shortest removal path based on the peak area distribution by using an ant colony algorithm or a genetic algorithm.

6. The intelligent garbage classification management system according to claim 1 is characterized in that: The digital monitoring equipment is also used to detect the airtightness of the vehicle compartment in real time through the on-board weighing sensor and pressure monitoring device installed in the garbage collection vehicle, and trigger an alarm when an abnormal state is determined; collect GPS positioning data and road traffic condition images in real time through the GPS positioning device and camera device installed in the garbage collection vehicle, and dynamically optimize the garbage collection route of the garbage collection vehicle based on the GPS positioning data and road traffic condition images.

7. The intelligent garbage classification management system according to claim 1 is characterized in that: The intelligent decision-making device is also used to analyze the amount of garbage put in and the classification accuracy based on big data, generate a report on garbage classification and display it visually.

8. The intelligent garbage classification management system according to claim 1 is characterized in that: The digital monitoring equipment is also used to display the full process data of garbage placement, transportation and treatment.

9. The intelligent garbage classification management system according to claim 1 is characterized in that: It also includes a designated recycling and reservation platform; The fixed-point recycling and reservation platform is used to assign recycling personnel to carry out door-to-door recycling according to the recycling time point reserved by the user.

10. The intelligent garbage classification management system according to claim 1 is characterized in that: It also includes a big data supervision and collaboration platform; The big data supervision and coordination system is used to use big data analysis to identify violations in garbage classification, and to share data and conduct special inspections with the urban management system and the sanitation system.

Images