Garbage recycling resource benefit dynamic evaluation method for urban steward

Abstract

The invention relates to the technical field of urban garbage management, and discloses a garbage recycling resource benefit dynamic evaluation method for urban large housekeepers. The method comprises the following steps: receiving real-time sensor data of a plurality of garbage collection points, determining garbage type distribution and processing priorities by using a preset garbage classification model, and dynamically generating a resource allocation strategy which comprises recycling vehicle path planning and processing equipment scheduling instructions. And predicting a garbage increment trend through a preset spatial-temporal distribution model, adjusting resource allocation strategy parameters, updating a processing priority through an iterative optimization algorithm, and outputting a resource scheduling scheme. According to the method, the condition of a recycling point can be accurately mastered in real time, accurate garbage classification and efficient resource allocation are achieved, garbage increment can be predicted, emergencies can be dealt with, garbage recycling resource benefits are effectively improved, urban garbage management work is optimized, and urban sustainable development is promoted.

Description

Technical Field

[0001] The present invention relates to the technical field of urban waste management, and in particular to a method for dynamically evaluating the benefits of waste recycling resources for urban managers. Background Art

[0002] With the acceleration of urbanization, the urban population has grown dramatically, and the amount of urban waste generated has also increased significantly. Waste recycling, as a key part of urban environmental management, is crucial for resource recycling, environmental protection, and the sustainable development of cities. However, current urban waste recycling management faces many challenges. Traditional waste recycling models often lack effective collection and analysis of real-time data from waste collection points. During the waste collection process, information such as waste capacity, recycling equipment status, and environmental parameters cannot be obtained in a timely manner, making it difficult for urban waste management departments to accurately grasp the actual conditions of each recycling point. For example, due to the lack of real-time knowledge of the waste capacity of a waste collection point, it is often the case that the waste collection vehicle arrives overflowing, or the waste at the collection point remains uncollected for a long time. This not only affects the aesthetics of the urban environment, but can also cause environmental pollution problems.

[0003] Existing waste sorting methods are crude and difficult to achieve accurate classification. In most cases, waste sorting relies solely on simple manual experience or limited classification symbols, failing to fully tap the potential resource value of waste. This results in some recyclable and non-recyclable waste being mixed together, resulting in wasted resources and increased waste disposal costs. Furthermore, there is a lack of scientific basis for determining the treatment priority of different types of waste, and the processing order cannot be rationally arranged according to the characteristics of the waste, which reduces waste disposal efficiency.

[0004] In terms of resource allocation, recycling vehicle routing and processing equipment scheduling are illogical. Recycling vehicles often follow fixed routes for waste collection, failing to consider real-time traffic conditions and the actual needs of each collection point. This results in long travel times, high fuel consumption, and low recycling efficiency. Furthermore, processing equipment scheduling lacks flexibility, failing to optimize configuration based on equipment load and waste type. This results in some equipment being overloaded while others remain idle, resulting in a waste of resources.

[0005] Inaccurate forecasts of waste growth trends are also a prominent issue. Urban waste generation is influenced by a variety of factors, such as seasonal variations, population mobility, and economic activity. Existing forecasting methods struggle to comprehensively account for these complex factors, making it difficult to accurately predict future waste growth. This leaves urban waste management departments without a scientific basis for resource reserves and allocation, making resource shortages and surpluses more likely.

[0006] In addition, traditional garbage collection management systems cannot be dynamically adjusted and optimized according to actual situations. Once an emergency occurs, such as a sudden significant increase in the amount of garbage at a certain collection point, the system cannot respond in a timely manner and adjust the resource allocation strategy, resulting in chaos in the garbage collection work and affecting the normal operation of the city. In summary, the current urban garbage collection management model urgently needs to be improved, and an efficient and intelligent dynamic evaluation method for the resource benefits of garbage collection is required to enhance the urban garbage collection management level, achieve the efficient utilization of resources, and the sustainable development of the city. Summary of the Invention

[0007] The purpose of the present invention is to provide a dynamic evaluation method for the resource benefits of garbage collection for the urban housekeeper to solve the problems raised in the above background technology.

[0008] To achieve the above purpose, the present invention provides the following technical solution: A dynamic evaluation method for the resource benefits of garbage collection for the urban housekeeper, the method includes: Receiving real-time sensor data of multiple garbage collection points in the target area, the sensor data includes garbage capacity, recycling equipment status, and environmental parameters; Based on a preset garbage classification model, performing multi-dimensional feature extraction on the real-time sensor data to determine the garbage type distribution and processing priority of each garbage collection point; According to the processing priority, dynamically generate a resource allocation strategy, and the strategy includes recycling vehicle route planning and processing equipment scheduling instructions; Based on a preset spatio-temporal distribution model, predicting the garbage increment trend of each garbage collection point within a preset future time period, and adjusting the dynamic weight parameters in the resource allocation strategy; Under the adjusted resource allocation strategy, updating the processing priority of each collection point through an iterative optimization algorithm, and outputting the final resource scheduling plan to the urban garbage management system.

[0009] Preferably, the construction steps of the garbage classification model include: Collecting historical garbage treatment data to construct a training sample set including garbage composition, weight, and treatment time; Performing feature partitioning on the training sample set through an unsupervised clustering algorithm to generate initial classification labels; Combining manual annotation to correct the initial classification labels to obtain multi-level classification rules; Embedding the multi-level classification rules into a lightweight neural network to generate the dynamically updatable garbage classification model.

[0010] Preferably, the dynamic generation of the resource allocation strategy includes: According to the processing priority, assigning an initial weight coefficient to each collection point; Calculate the accessibility score of each collection point based on real-time traffic flow data and equipment failure status; Weightedly fuse the accessibility score and the initial weight coefficient to generate the path planning of the collection vehicle; Generate the scheduling instruction of the processing equipment according to the path planning result and in combination with the equipment load balancing principle.

[0011] Preferably, the construction steps of the spatio-temporal distribution model include: Collect historical garbage increment data and external environment data of the target area to construct a spatio-temporal sequence data set; Extract periodic features from the spatio-temporal sequence data set through a sliding window algorithm to determine the time correlation of garbage increment; Combine geographic information system data to establish a spatial association matrix and quantify the garbage transfer probability between collection points; Input the time correlation and the spatial association matrix into a long short-term memory network to generate the spatio-temporal distribution model.

[0012] Preferably, the method further includes: Identify collection points where the garbage increment exceeds a preset threshold according to the prediction result of the spatio-temporal distribution model; Assign an emergency response label to the identified collection points in the resource allocation strategy; Based on the emergency response label, preferentially dispatch redundant collection resources to the corresponding collection points and suspend non-emergency tasks.

[0013] Preferably, the calculation of the accessibility score includes: Obtain the real-time road congestion index and weather impact factor to construct a dynamic traffic evaluation matrix; Calculate the shortest path between collection points through a graph theory algorithm to generate a basic accessibility value; Perform a convolution operation on the dynamic traffic evaluation matrix and the basic accessibility value to obtain the accessibility score; Among them, the calculation formula of the accessibility score is:

[0014] In the formula, represents the real-time congestion index of the th road, represents the weather impact factor weight of the th road, represents the basic accessibility value.

[0015] Preferably, the correction of the multi-level classification rule includes: Evaluate the confidence of the initial classification label and screen out low-confidence samples; Completing the labels of the low-confidence samples through a semi-supervised learning algorithm; Reconstructing the hierarchical logic of the multi-level classification rules according to the completed labels.

[0016] Preferably, the execution of the iterative optimization algorithm includes: Setting an optimization objective function for the resource allocation strategy, the function including processing efficiency and energy consumption constraints; Performing multiple rounds of iterative adjustment on the dynamic weight parameters through a particle swarm optimization algorithm; In each round of iteration, dynamically pruning the particle swarm according to the degree of constraint violation and retaining the optimal solution set; Outputting the resource scheduling scheme that meets the preset convergence conditions; Among them, the optimization objective function is:

[0017] In the formula, represents the processing efficiency score, represents the energy consumption score, and are the normalized weight coefficients of efficiency and energy consumption respectively.

[0018] Preferably, the method further includes: After allocating the emergency response tag, real-time monitoring the processing progress of the corresponding collection point; If the processing progress does not reach the expected rate, triggering an adaptive scaling mechanism to dynamically adjust the number of collection vehicles and equipment power.

[0019] Preferably, the setting of the energy consumption constraint includes: Collecting the rated power and historical energy consumption data of each processing device to construct an energy consumption baseline model; Calculating the dynamic energy consumption upper limit according to the real-time device load rate; Taking the dynamic energy consumption upper limit as the constraint condition of the optimization objective function; Among them, the calculation formula of the dynamic energy consumption upper limit is:

[0020] In the formula, represents the rated power of the device, represents the real-time load rate, is the environmental temperature correction coefficient.

[0021] Compared with the prior art, the beneficial effects of the present invention are: The dynamic evaluation method of garbage recycling resource benefits proposed by the present invention for the urban butler brings significant beneficial effects in many aspects to urban garbage recycling management. In terms of data acquisition and processing, by receiving real-time sensor data of multiple garbage recycling points in the target area, including garbage capacity, recycling equipment status, and environmental parameters, the urban garbage management department can accurately and in real-time grasp the actual situation of each recycling point. This avoids problems such as garbage overflow or long-term uncleaned due to information lag, and greatly improves the cleanliness of the urban environment. For example, when the garbage capacity of a certain recycling point approaches the upper limit, the system can issue a warning in time and arrange a recycling vehicle to go for cleaning, effectively maintaining the environmental hygiene of the city.

[0022] In the garbage classification link, multi-dimensional feature extraction is carried out based on a preset garbage classification model to determine the garbage type distribution and processing priority of each garbage recycling point. This precise classification method can fully explore the potential resource value in garbage and improve the resource recycling rate. Effectively separating different types of garbage enables recyclable garbage to be properly recycled and reused, reducing resource waste. At the same time, arranging the processing order according to the processing priority greatly improves the garbage processing efficiency. For garbage that needs to be processed in time, such as perishable garbage, it can be processed first, avoiding odor and environmental pollution caused by untimely processing.

[0023] The dynamic generation of resource allocation strategies is a major highlight of the present invention. According to the processing priority, combined with real-time traffic flow data and equipment failure status, an initial weight coefficient is assigned to each recycling point and the reachability score is calculated, so as to generate reasonable recycling vehicle route planning and processing equipment scheduling instructions. This enables the recycling vehicle to avoid congested roads, reduce driving time and fuel consumption, and improve the recycling efficiency. Moreover, according to the principle of equipment load balancing, the processing equipment is scheduled to avoid overuse or idleness of the equipment, extend the service life of the equipment, and reduce the equipment maintenance cost. For example, during peak traffic hours, the system will adjust the vehicle route according to the real-time road conditions, give priority to going to the recycling points with high reachability, and at the same time reasonably allocate garbage to the processing equipment with lower load.

[0024] By predicting the garbage increment trend of each garbage recycling point within a preset future time period through a preset spatio-temporal distribution model and adjusting the dynamic weight parameters in the resource allocation strategy, the urban garbage management department can make preparations for resource reserve and allocation in advance. When it is predicted that the garbage increment in a certain area increases due to special activities or seasonal changes, more recycling vehicles and processing equipment can be arranged in advance to avoid resource shortages. On the contrary, if it is predicted that the garbage volume decreases, the resource input can be reasonably reduced to avoid resource waste.

[0025] In the face of emergencies, the method of the present invention demonstrates strong adaptability. After identifying a recycling point where the garbage increment exceeds the preset threshold, an emergency response tag is assigned to it, redundant recycling resources are preferentially scheduled, non-emergency tasks are suspended, and the processing progress is monitored in real time. If the processing progress does not reach the expected rate, an adaptive scaling mechanism is triggered to dynamically adjust the number of recycling vehicles and equipment power. This series of measures ensures that the system can respond quickly when unexpected situations occur during garbage recycling, guarantee the smooth progress of garbage recycling work, and maintain the normal operation of the city.

[0026] In terms of optimizing resource scheduling, through an iterative optimization algorithm, an optimization objective function including processing efficiency and energy consumption constraints is set, and the particle swarm optimization algorithm is used for multiple rounds of iterative adjustment. It can achieve the best balance between processing efficiency and energy consumption on the premise of meeting energy consumption constraints. This not only improves the efficiency of garbage recycling but also reduces energy consumption, in line with the concept of sustainable development. For example, while ensuring the timely processing of garbage, unnecessary energy consumption of equipment is reduced, and operating costs are lowered. Brief Description of the Drawings

[0027] Figure 1 It is the working principle diagram of the dynamic evaluation method for the benefit of garbage recycling resources described in the present invention; Figure 2 It is the working principle diagram for dynamically generating resource allocation strategies; Figure 3 It is the flow chart for constructing the spatio-temporal distribution model; Figure 4 It is the flow chart for monitoring and adjusting the processing progress after emergency response. Detailed Embodiment

[0028] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without making creative efforts shall fall within the protection scope of the present invention.

[0029] Please refer to Figures 1-4 , the present invention provides a dynamic evaluation method for the benefit of garbage recycling resources for the urban butler, aiming to efficiently manage urban garbage recycling work and improve resource utilization efficiency. The specific implementation steps are as follows: Through sensors deployed at multiple garbage recycling points in the target area, relevant data of the garbage recycling points are collected in real time. These sensors include sensors for monitoring garbage capacity, sensors for detecting the status of recycling equipment, and sensors for obtaining environmental parameters (such as temperature, humidity, etc.). The system continuously receives these real-time sensor data to provide a basis for subsequent analysis.

[0030] Process the collected real-time sensor data using a preset waste classification model. This model can extract multi-dimensional features from the data to determine the waste type distribution at each waste collection point. At the same time, determine the processing priority for each waste collection point according to the type, volume, and other relevant factors of the waste. For example, for hazardous waste, the processing priority is usually high; for collection points with a volume approaching the upper limit, their processing priority will also be increased.

[0031] Based on the determined processing priorities, dynamically generate a resource allocation strategy. This strategy covers the route planning of recycling vehicles and the scheduling instructions for processing equipment. When planning the routes of recycling vehicles, consider factors such as real-time traffic conditions, the locations of each collection point, and the processing priorities to ensure that the vehicles can efficiently travel to each collection point for waste collection. When scheduling processing equipment, reasonably arrange the equipment to process the collected waste according to the load of the equipment and the processing requirements of different types of waste.

[0032] With the help of a preset spatio-temporal distribution model, predict the waste increment trend at each waste collection point within a preset future time period based on historical waste increment data and current real-time data. Based on the prediction results, adjust the dynamic weight parameters in the resource allocation strategy. If a certain collection point is predicted to have a large waste increment, then its weight will be appropriately increased during resource allocation to increase the recycling resources allocated to it.

[0033] Based on the adjusted resource allocation strategy, use an iterative optimization algorithm to update the processing priorities of each collection point. Through continuous iteration, make the resource allocation more reasonable. Set an optimization objective function, consider factors such as processing efficiency and energy consumption, and after multiple rounds of iterative calculations, finally output a resource scheduling plan that meets the preset convergence conditions, and transmit this plan to the urban waste management system to achieve scientific and efficient management of urban waste collection work.

[0034] The following further illustrates the implementation of the present invention in combination with Embodiments 1 to 5.

[0035] Embodiment 1: When constructing the waste classification model, first carry out the collection work of the training sample set. Collect a large amount of historical waste treatment data from the historical database of the urban waste treatment system. These data detail the composition of the waste, such as the specific proportions of paper, plastic, and metal in recyclables, the types and quantities of batteries, expired drugs, etc. in hazardous waste, as well as the specific compositions of other waste and kitchen waste. At the same time, record the weight information of the waste, accurate to the actual weight value of each batch of waste. Also collect the processing time data from the collection point to the completion of processing, and count the time spent in different types of waste and different processing processes. Organize these data to construct a training sample set containing waste composition, weight, and processing time.

[0036] Use an unsupervised clustering algorithm to perform feature partitioning on the training sample set. For example, the K-Means clustering algorithm is adopted. This algorithm will automatically divide similar data into the same category according to the characteristics of the data, generating initial classification labels. However, the classification results obtained by the unsupervised clustering algorithm may not be completely accurate. Therefore, it is necessary to combine manual annotation to correct the initial classification labels. Organize professional waste classification personnel to check and annotate the clustering results one by one, and reclassify the samples with inaccurate classification.

[0037] During the correction process, perform confidence evaluation on the initial classification labels. The confidence level can be determined by calculating the certainty degree of the clustering algorithm for classifying each sample, and filter out the samples with low confidence. For these samples with low confidence, use a semi-supervised learning algorithm to complete the label complementation. The semi-supervised learning algorithm uses a small number of labeled samples and a large number of unlabeled samples for learning, and improves the classification labels by mining the potential information in the unlabeled samples. According to the complemented labels, reconstruct the hierarchical logic of the multi-level classification rules. For example, according to the attributes such as recyclability and harmfulness of the garbage, divide the garbage into different major categories, and then further subdivide the small categories under the major categories to form a clear multi-level classification system.

[0038] Finally, embed the constructed multi-level classification rules into a lightweight neural network. Select a suitable lightweight neural network architecture, such as MobileNet, etc., transform the classification rules into the parameters and structure of the neural network, and generate a garbage classification model that can be dynamically updated. As new garbage disposal data is continuously collected, the model can be further optimized and updated to adapt to the changing garbage composition and disposal requirements.

[0039] In this embodiment, by screening out samples with low confidence through confidence evaluation, it is possible to accurately locate the samples that need further processing, avoid repeated operations on a large number of accurately classified samples, and improve the correction efficiency. The semi-supervised learning algorithm makes full use of the information of the unlabeled samples and improves the accuracy of the classification labels as much as possible under the condition of limited manual annotation workload. Reconstructing the hierarchical logic of the multi-level classification rules makes the classification system more scientific and reasonable, providing strong support for accurately determining the garbage type distribution and treatment priority in the future.

[0040] Embodiment 2: During the process of dynamically generating a resource allocation strategy, assign an initial weight coefficient to each recycling point according to its processing priority. For recycling points with a high processing priority, assign a relatively large initial weight coefficient. For example, for a recycling point with a first-level processing priority, the initial weight coefficient is set to 0.8; for a recycling point with a second-level processing priority, the initial weight coefficient is set to 0.6; for a recycling point with a third-level processing priority, the initial weight coefficient is set to 0.4. This can ensure that high-priority recycling points can receive more attention during resource allocation.

[0041] To calculate the accessibility scores of each recycling point, first obtain the real-time road congestion index and weather impact factor, and construct a dynamic traffic evaluation matrix. The real-time road congestion index can be obtained from the system of the traffic management department, which reflects the current congestion level of the road, with a value range of 0 - 1. The larger the value, the more congested the road. The weather impact factor is determined according to the degree of impact of different weather conditions on traffic. For example, on rainy days, the road is slippery and the vehicle driving speed is affected, and the weather impact factor for rainy days is set to 0.8; on sunny days, the weather impact factor is set to 1. A dynamic traffic evaluation matrix is constructed based on the real-time congestion index of different roads and the corresponding weather impact factor weights.

[0042] Calculate the shortest paths between each recycling point through graph theory algorithms to generate the basic accessibility values. Graph theory algorithms such as Dijkstra's algorithm can be used to calculate the shortest path lengths from one recycling point to other recycling points, which are used as the basic accessibility values.

[0043] Perform a convolution operation on the dynamic traffic evaluation matrix and the basic accessibility values to obtain the accessibility scores. The calculation formula for the accessibility scores is:

[0044] where, represents the real-time congestion index of the th road, reflecting the current congestion status of the road; represents the weather impact factor weight of the th road, reflecting the degree of impact of the weather on the driving conditions of this road; represents the basic accessibility value, which is the shortest path length between recycling points calculated through graph theory algorithms.

[0045] Perform a weighted fusion of the accessibility scores and the initial weight coefficients to generate the path planning for the recycling vehicles. For example, for two recycling points A and B, the initial weight coefficient of A is 0.8 and the accessibility score is 5; the initial weight coefficient of B is 0.6 and the accessibility score is 8. Through weighted fusion calculation, considering both the weights and accessibility, determine the path planning for the vehicle to go to point B first and then to point A.

[0046] According to the path planning results, combined with the principle of equipment load balancing, generate the processing equipment scheduling instructions. If a certain processing equipment has a relatively low current load and the type of garbage to be recycled is suitable for this equipment to process, then this equipment is preferentially scheduled to process these garbage to ensure that the loads of each processing equipment are relatively balanced and improve the overall processing efficiency.

[0047] In this embodiment, by comprehensively considering factors such as processing priority, real-time traffic conditions, and weather, a resource allocation strategy is generated, making the recycling vehicle route planning more reasonable and the processing equipment scheduling more scientific. The calculation method of the reachability score fully considers various factors in actual traffic, providing a more accurate basis for route planning and helping to improve the efficiency of garbage recycling and the resource utilization benefit.

[0048] Embodiment 3: When constructing the spatio-temporal distribution model, first collect the historical garbage increment data and external environment data of the target area to construct a spatio-temporal sequence data set. The historical garbage increment data can be obtained from the database of the urban garbage management department, which details the garbage increment situation at each garbage recycling point at different times, accurate to the daily or even hourly garbage increase. The external environment data includes meteorological data such as temperature, precipitation, and wind speed, as well as urban activity data, such as holiday arrangements and large event holding situations. These data are sorted according to the time sequence and the geographical location of the recycling points to construct a spatio-temporal sequence data set.

[0049] Use the sliding window algorithm to extract periodic features from the spatio-temporal sequence data set to determine the time correlation of the garbage increment. The sliding window algorithm slides a fixed-length window on the time series and analyzes the data within the window. For example, set the window length to one week. By statistically analyzing the data of each week, it is found that the garbage increment shows different patterns on weekdays and weekends. The garbage increment is relatively stable on weekdays and increases on weekends, thus determining the time correlation of the garbage increment.

[0050] Combined with the geographic information system data, establish a spatial association matrix to quantify the garbage transfer probability between each recycling point. The geographic information system data includes information such as the geographical location coordinates of the recycling points. By analyzing factors such as the distance between recycling points and traffic connection conditions, a spatial association matrix is established. If two recycling points are close and have convenient transportation, the garbage transfer probability between them is relatively high; otherwise, it is relatively low. For example, the distance between recycling point A and recycling point B is close and the middle road is unobstructed, and the garbage transfer probability between them is set to 0.6; the distance between recycling point C and recycling point D is far and the traffic is inconvenient, and the garbage transfer probability between them is set to 0.2.

[0051] Input the time correlation and the spatial association matrix into the long short-term memory network (LSTM) to generate the spatio-temporal distribution model. LSTM can effectively process time series data and capture the long-term dependence relationships in the data. By using the time correlation and the spatial association matrix as inputs, the LSTM model learns the changing patterns of the garbage increment in time and space, thus generating a spatio-temporal distribution model that can accurately predict the future garbage increment trend.

[0052] In this embodiment, the construction of the spatio-temporal distribution model fully considers the factors of both time and space dimensions. The application of the sliding window algorithm helps to discover the temporal pattern of garbage increment. The establishment of geographical information system data and spatial correlation matrix quantifies the spatial relationship between recycling points. The long short-term memory network integrates these factors to achieve accurate prediction of the garbage increment trend, providing a reliable basis for subsequent adjustment of resource allocation strategies.

[0053] Embodiment 4: According to the prediction results of the spatio-temporal distribution model, identify the recycling points where the garbage increment exceeds the preset threshold. The preset threshold can be set according to historical data and practical experience. For example, set the threshold that the garbage increment exceeds 10 cubic meters within the next 24 hours. When the model predicts that the garbage increment of a certain recycling point exceeds this threshold, assign an emergency response flag to it in the resource allocation strategy.

[0054] Based on the emergency response flag, preferentially dispatch redundant recycling resources to the corresponding recycling points and suspend non-emergency tasks. Redundant recycling resources include spare recycling vehicles and processing equipment, etc. For example, when recycling point X is marked as an emergency response, immediately dispatch a spare recycling vehicle to this recycling point, and at the same time suspend the recycling tasks of some other recycling points with lower processing priorities to ensure that the emergency recycling tasks can be completed first.

[0055] After assigning the emergency response flag, monitor the processing progress of the corresponding recycling point in real time. Through the sensors installed on the recycling vehicles and processing equipment, obtain the progress information of garbage recycling and processing in real time, such as the loading capacity of the recycling vehicle, the running time of the processing equipment, etc. If the processing progress does not reach the expected rate, for example, the loading capacity of the recycling vehicle does not reach 80% within the specified time, or the processing efficiency of the processing equipment is lower than 90% of the designed efficiency, then trigger the adaptive scaling mechanism.

[0056] The adaptive scaling mechanism will dynamically adjust the number of recycling vehicles and the power of equipment. If the processing progress is slow, increase the number of recycling vehicles and allocate vehicles from other non-emergency tasks for support; at the same time, increase the power of the processing equipment to speed up the garbage processing speed. On the contrary, if the processing progress is fast and the garbage volume at the recycling point gradually decreases, the number of recycling vehicles can be appropriately reduced and the power of the processing equipment can be lowered to save energy.

[0057] In this embodiment, through the prediction of garbage increment and the setting of the emergency response mechanism, it is possible to timely respond to emergencies in the garbage recycling process and ensure the smooth progress of the garbage recycling work. The real-time monitoring of the processing progress and the application of the adaptive scaling mechanism further improve the efficiency of garbage recycling and the resource utilization benefit, ensuring the efficient and stable operation of the urban garbage management system.

[0058] Example 5: When implementing the iterative optimization algorithm, first set the optimization objective function of the resource allocation strategy, which includes processing efficiency and energy consumption constraints. The optimization objective function is:

[0059] where represents the processing efficiency score, which is used to measure the efficiency of garbage collection and processing. For example, it can be determined by calculating the amount of garbage recycled per unit time or the number of garbage batches processed. represents the energy consumption score, which reflects the energy consumption during the garbage collection and processing process. It is calculated by statistically analyzing the driving mileage of the recycling vehicle, the operating time of the processing equipment, and the energy consumption data. and are the normalized weight coefficients of efficiency and energy consumption respectively, which are used to adjust the importance of processing efficiency and energy consumption in the optimization objective, and satisfy , and their values can be adjusted according to actual needs. For example, when more emphasis is placed on processing efficiency, is set to 0.7, is set to 0.3; when more emphasis is placed on energy consumption, is set to 0.4, is set to 0.6.

[0060] Collect the rated power and historical energy consumption data of each processing device to construct an energy consumption baseline model. The rated power can be obtained from the technical parameter manual of the device, and the historical energy consumption data is collected from the operation records of the device. By analyzing these data, an energy consumption baseline model is established. For example, a linear regression model is used to predict the energy consumption of the device under different load conditions.

[0061] Calculate the dynamic energy consumption upper limit according to the real-time device load rate. The calculation formula for the dynamic energy consumption upper limit is:

[0062] where represents the rated power of the device, which is the maximum power of the device under normal working conditions; represents the real-time load rate, which reflects the current working load degree of the device, and the value range is 0 - 1; is the environmental temperature correction coefficient, which considers the influence of environmental temperature on the energy consumption of the device. For example, in a high-temperature environment, it is difficult for the device to dissipate heat, and the energy consumption will increase. At this time, is set to 1.1; in a low-temperature environment, the energy consumption of the device is relatively low, and is set to 0.9. The dynamic energy consumption upper limit is used as a constraint condition for the optimization objective function to ensure that the energy consumption does not exceed a reasonable range when optimizing the resource allocation strategy.

[0063] The dynamic weight parameters are adjusted through multiple rounds of iteration by the particle swarm optimization algorithm. The particle swarm optimization algorithm simulates the foraging behavior of a bird flock. Each possible resource allocation strategy is regarded as a particle. Through information sharing among particles and their own search experience, the position of the particle (i.e., the dynamic weight parameter) is continuously adjusted to find the optimal solution. In each round of iteration, the particle swarm is dynamically pruned according to the degree of constraint violation, and the optimal solution set is retained. For example, if the resource allocation strategy corresponding to a certain particle results in energy consumption exceeding the dynamic energy consumption upper limit or the processing efficiency is too low and does not meet the actual requirements, the particle is removed from the particle swarm, and only the particles that meet the constraint conditions and have better objective function values are retained.

[0064] After multiple rounds of iterative calculation, when the preset convergence condition is met, the resource scheduling scheme is output. The preset convergence condition can be that the objective function value changes very little in several consecutive rounds of iteration. For example, the change rate is less than 0.01. At this time, it is considered that the algorithm has converged, and the resource scheduling scheme that meets the requirements is output.

[0065] In this embodiment, by setting reasonable optimization objective functions and energy consumption constraints, and combining the particle swarm optimization algorithm for iterative optimization, a balance can be found between processing efficiency and energy consumption, and efficient allocation of resources can be achieved. The setting of the dynamic pruning mechanism and the preset convergence condition ensures the convergence speed and solution quality of the algorithm, making the output resource scheduling scheme more scientific and reasonable, and improving the overall efficiency of urban waste recycling.

[0066] It should be noted that in this article, relational terms such as first and second are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variant thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements not only includes those elements, but also includes other elements not expressly listed, or also includes elements inherent to such process, method, article or device.

[0067] Although the embodiments of the present invention have been shown and described, for those of ordinary skill in the art, it can be understood that various changes, modifications, substitutions and variations can be made to these embodiments without departing from the principle and spirit of the present invention. The scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A dynamic evaluation method for the resource benefits of garbage collection for urban general managers, characterized in that, Including: Receiving real-time sensor data of multiple waste collection points in the target area, where the sensor data includes waste capacity, recycling equipment status, and environmental parameters; Based on a preset waste classification model, extracting features from the real-time sensor data to determine the waste type distribution and processing priority of each waste collection point; According to the processing priority, dynamically generating a resource allocation strategy, where the strategy includes recycling vehicle route planning and processing equipment scheduling instructions; Based on a preset spatio-temporal distribution model, predicting the waste increment trend of each waste collection point within a preset future time period and adjusting the dynamic weight parameters in the resource allocation strategy; Under the adjusted resource allocation strategy, updating the processing priority of each collection point through an iterative optimization algorithm and outputting the final resource scheduling plan to the urban waste management system.

2. The dynamic evaluation method for the resource benefit of garbage recycling according to claim 1, wherein The construction steps of the waste classification model include: Collecting historical waste treatment data and constructing a training sample set containing waste composition, weight, and treatment timeliness; Performing feature partitioning on the training sample set through an unsupervised clustering algorithm to generate initial classification labels; Combining manual annotation to correct the initial classification labels to obtain multi-level classification rules; Embedding the multi-level classification rules into a lightweight neural network to generate the dynamically updatable waste classification model.

3. The dynamic evaluation method for the resource benefit of waste recycling according to claim 1, characterized in that The dynamically generating resource allocation strategy includes: According to the processing priority, assigning an initial weight coefficient to each collection point; Based on real-time traffic flow data and equipment failure status, calculating the accessibility score of each collection point; Performing weighted fusion of the accessibility score and the initial weight coefficient to generate the recycling vehicle route planning; According to the route planning result, combining the equipment load balancing principle to generate the processing equipment scheduling instructions.

4. The dynamic evaluation method for the resource benefit of garbage recycling according to claim 1, wherein The construction steps of the spatio-temporal distribution model include: Collecting historical waste increment data and external environmental data of the target area and constructing a spatio-temporal sequence data set; Performing periodic feature extraction on the spatio-temporal sequence data set through a sliding window algorithm to determine the time correlation of waste increment; Combining geographic information system data to establish a spatial association matrix and quantifying the waste transfer probability between each collection point; Inputting the time correlation and the spatial association matrix into a long short-term memory network to generate the spatio-temporal distribution model.

5. The dynamic evaluation method for the resource benefit of garbage recycling according to claim 4, wherein Also including: According to the prediction result of the spatio-temporal distribution model, identifying the collection points where the waste increment exceeds the preset threshold; Assigning an emergency response flag to the identified collection points in the resource allocation strategy; Based on the emergency response flag, preferentially dispatching redundant recycling resources to the corresponding collection points and suspending non-emergency tasks.

6. The dynamic evaluation method for the resource benefit of garbage recycling according to claim 3, characterized in that, The calculation of the accessibility score includes: Obtaining the real-time road congestion index and weather impact factor and constructing a dynamic traffic assessment matrix; Calculating the shortest path between each collection point through a graph theory algorithm to generate a basic accessibility value; Performing a convolution operation on the dynamic traffic assessment matrix and the basic accessibility value to obtain the accessibility score; Among them, the calculation formula of the reachability score is as follows: , where in the formula represents the real-time congestion index of the th road, represents the weather impact factor weight of the th road, represents the basic reachability value.

7. The dynamic evaluation method for the resource benefit of garbage recycling according to claim 2, wherein The correction of the multi-level classification rules includes: Performing confidence evaluation on the initial classification labels and screening low-confidence samples; Completing the labels of the low-confidence samples through a semi-supervised learning algorithm; Reconstruct the hierarchical logic of the multi-level classification rule according to the completed label.

8. The dynamic evaluation method for the resource benefit of garbage recycling according to claim 1, wherein The execution of the iterative optimization algorithm includes: Setting an optimization objective function for the resource allocation strategy, the function including processing efficiency and energy consumption constraints; Performing multiple rounds of iterative adjustment on the dynamic weight parameter through a particle swarm optimization algorithm; In each round of iteration, dynamically pruning the particle swarm according to the degree of constraint violation and retaining the optimal solution set; Outputting the resource scheduling scheme that meets the preset convergence condition; Among them, the optimized objective function is as follows: , where in the formula represents the processing efficiency score, represents the energy consumption score, and are the normalized weight coefficients of efficiency and energy consumption respectively.

9. The dynamic evaluation method for the resource benefit of garbage recycling according to claim 5, wherein It further includes: After allocating the emergency response label, monitoring the processing progress of the corresponding recovery point in real time; If the processing progress does not reach the expected rate, triggering an adaptive scaling mechanism to dynamically adjust the number of recovery vehicles and equipment power.

10. The dynamic evaluation method for the resource benefit of garbage recycling according to claim 8, characterized in that The setting of the energy consumption constraint includes: Collecting the rated power and historical energy consumption data of each processing device to construct an energy consumption baseline model; Calculating the dynamic energy consumption upper limit according to the real-time device load rate; Taking the dynamic energy consumption upper limit as the constraint condition of the optimization objective function; Among them, the calculation formula for the dynamic energy consumption upper limit is: , in the formula, represents the rated power of the device, represents the real-time load rate, is the environmental temperature correction coefficient.

Images