Garbage whole-process internet-of-things sensing system based on RFID technology

Abstract

The invention relates to the field of garbage IOT (Internet of Things) sensing, and particularly discloses a garbage whole-process IOT sensing system based on an RFID (Radio Frequency Identification Device) technology, which comprises the following steps of: firstly, acquiring garbage can filling state data, garbage generation time and place data and garbage clearance process data acquired by RFID equipment, and then utilizing a deep learning technology to acquire garbage can filling state data, garbage generation time and place data and garbage clearance process data; the method comprises the following steps: performing feature extraction and correlation analysis on the three, and finally generating a garbage disposal progress report through a generator, thereby adjusting a garbage clearance strategy and a resource allocation strategy according to the progress report to ensure that the garbage is timely and effectively disposed, and further optimizing the overall disposal efficiency.

Description

Technical Field

[0001] The present application relates to the field of garbage IoT perception, and more specifically, to a garbage full-process IoT perception system based on RFID technology. Background Art

[0002] Garbage refers to waste generated in people's daily life or production activities that is considered no longer useful. These wastes usually include food residues, packaging materials, old furniture, waste batteries and other materials. There are many types of garbage, from recyclables to hazardous waste, and the treatment methods are also different. Effective garbage management can not only reduce environmental pollution, but also promote resource recovery and reuse.

[0003] In recent years, with the improvement of people's living standards and environmental awareness, the timely removal of garbage has become increasingly important. Especially in hot weather, if the garbage in the community is stored for a long time and is not removed in time, it will have a great impact on property management and people's living environment.

[0004] Therefore, a garbage whole-process IoT sensing system based on RFID technology is desired. Summary of the invention

[0005] In order to solve the above technical problems, the present application is proposed. The embodiment of the present application provides a garbage whole process IoT perception system based on RFID technology, which first obtains the garbage bin filling status data, garbage generation time and location data and garbage removal process data collected by the RFID device, and then uses deep learning technology to perform feature extraction and correlation analysis on the three, and finally generates a garbage treatment progress report through a generator, so as to adjust the garbage removal strategy and resource allocation strategy according to the progress report to ensure that the garbage is handled in a timely and effective manner, thereby optimizing the overall processing efficiency.

[0006] According to one aspect of the present application, a garbage whole process IoT sensing system based on RFID technology is provided, which includes:

[0007] The whole process of garbage IoT data acquisition module is used to obtain the garbage bin filling status data, garbage generation time and location data and garbage removal process data collected by RFID equipment, among which the garbage removal process data includes the time and location of the garbage bin being put into the garbage collection point, the time and location of the garbage removal truck picking up the garbage bin at the collection point, and the time and location of the garbage removal truck entering the garbage disposal site;

[0008] A garbage whole process IoT data processing module is used to extract garbage multimodal information association feature vectors and garbage removal process text semantic association feature vectors from the garbage bin filling status data, the garbage generation time and location data, and the garbage removal process data;

[0009] The garbage disposal progress report generation module is used to generate a garbage disposal progress report based on the garbage multimodal information association feature vector and the garbage removal process text semantic association feature vector.

[0010] Compared with the prior art, the present application provides a garbage full-process IoT perception system based on RFID technology, which first obtains the garbage bin filling status data, garbage generation time and location data, and garbage removal process data collected by RFID equipment, and then uses deep learning technology to perform feature extraction and correlation analysis on the three, and finally generates a garbage disposal progress report through a generator, so as to adjust the garbage removal strategy and resource allocation strategy according to the progress report to ensure that the garbage is handled in a timely and effective manner, thereby optimizing the overall processing efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0011] By describing the embodiments of the present application in more detail in conjunction with the accompanying drawings, the above and other purposes, features and advantages of the present application will become more apparent. The accompanying drawings are used to provide a further understanding of the embodiments of the present application and constitute a part of the specification. Together with the embodiments of the present application, they are used to explain the present application and do not constitute a limitation of the present application. In the accompanying drawings, the same reference numerals generally represent the same components or steps.

[0012] Figure 1 This is a block diagram of a garbage full-process IoT perception system based on RFID technology according to an embodiment of the present application.

[0013] Figure 2 This is a block diagram of a garbage whole-process Internet of Things data processing module in a garbage whole-process Internet of Things perception system based on RFID technology according to an embodiment of the present application.

[0014] Figure 3 This is a block diagram of a garbage removal process feature extraction unit in a garbage full-process IoT perception system based on RFID technology according to an embodiment of the present application.

[0015] Figure 4 This is a block diagram of a garbage processing progress report generation module in a garbage full-process IoT perception system based on RFID technology according to an embodiment of the present application. DETAILED DESCRIPTION

[0016] Below, the exemplary embodiments according to the present application will be described in detail with reference to the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments of the present application, and it should be understood that the present application is not limited to the exemplary embodiments described here.

[0017] Figure 1 This is a block diagram of the garbage whole process IoT sensing system based on RFID technology in the embodiment of the present application. Figure 1As shown, according to the RFID technology-based garbage whole-process IoT perception system 100 of the embodiment of the present application, it includes: a garbage whole-process IoT data acquisition module 110, which is used to acquire garbage bin filling status data, garbage generation time and location data and garbage removal process data collected by RFID equipment, wherein the garbage removal process data includes the time and location when the garbage bin is put into the garbage collection point, the time and location when the garbage removal truck arrives at the collection point to collect the garbage bin and load it onto the truck, and the time and location when the garbage removal truck enters the garbage disposal site; a garbage whole-process IoT data processing module 120, which is used to extract garbage multimodal information association feature vectors and garbage removal process text semantic association feature vectors from the garbage bin filling status data, the garbage generation time and location data and the garbage removal process data; a garbage disposal progress report generation module 130, which is used to generate a garbage disposal progress report based on the garbage multimodal information association feature vectors and the garbage removal process text semantic association feature vectors.

[0018] In the above-mentioned garbage whole process IoT perception system 100 based on RFID technology, the garbage whole process IoT data acquisition module 110 is used to obtain the garbage bin filling status data, garbage generation time and location data and garbage removal process data collected by the RFID device, wherein the garbage removal process data includes the time and location of the garbage bin into the garbage collection point, the time and location of the garbage collection truck to the collection point to collect the garbage bin, and the time and location of the garbage collection truck entering the garbage disposal site. It should be understood that garbage refers to waste generated by people in daily life or production process and is regarded as having lost its use value. Common garbage includes food residues, packaging materials, old furniture and discarded batteries. There are many types of garbage, and the treatment methods vary according to their categories, from recyclables to hazardous waste, each with different treatment strategies. Effective garbage management not only helps to reduce environmental pollution, but also promotes the recycling and reuse of resources. With the improvement of living standards and the enhancement of environmental awareness, the timely cleaning of garbage has become more and more important. Especially in hot weather, if the garbage in the community is not processed for a long time, it may have a significant adverse impact on property management and the living environment of residents. Therefore, in the technical solution of this application, the garbage bin filling status data, garbage generation time and location data, and garbage removal process data collected by the RFID device are obtained, and combined with deep learning technology, a garbage disposal progress report is generated to adjust the garbage removal strategy and resource allocation plan according to the report content. This adjustment is intended to ensure that garbage can be disposed of in a timely and effective manner, thereby improving the overall processing efficiency.

[0019] Specifically, the acquisition of garbage bin filling status data enables the garbage management system to grasp the use of garbage bins in real time. This data helps managers understand the actual filling volume of garbage bins and avoids the garbage bins being overfilled and unable to accommodate more garbage. The garbage generation time and location data provides the specific time and location of garbage generation. This information can not only help analyze the peak period and hot spots of garbage generation, but also optimize the schedule of garbage collection. Knowing the time and location of garbage generation can help better plan the removal tasks, thereby improving the efficiency of removal, reducing the retention time of garbage, and avoiding the negative impact on the environment caused by long-term storage of garbage. The garbage removal process data includes the time and location of the garbage bin placed in the garbage collection point, the time and location of the garbage collection truck arriving at the collection point, and the time and location of the garbage collection truck entering the garbage treatment site. These data provide a detailed timeline of the removal process to ensure that each link from garbage generation to treatment is carried out within the predetermined time frame. By collecting this data through RFID equipment, real-time monitoring and data-driven decision-making of the garbage management process can be achieved, thereby improving the efficiency of garbage removal and the overall intelligence level of the system. This comprehensive data collection and analysis method can significantly improve the timeliness and accuracy of garbage treatment, and provide strong support for creating a cleaner and more environmentally friendly urban environment.

[0020] In the above-mentioned garbage whole process IoT perception system 100 based on RFID technology, the garbage whole process IoT data processing module 120 is used to extract garbage multimodal information association feature vectors and garbage collection process text semantic association feature vectors from the garbage bin filling status data, the garbage generation time and location data, and the garbage collection process data. In this way, the garbage management system can achieve more accurate garbage disposal and optimization strategies. Among them, multimodal information provides a comprehensive perspective, and semantic features deeply analyze the details of the collection process. This comprehensive analysis capability can effectively improve the overall intelligent level of garbage management, reduce resource waste, improve processing efficiency, and ultimately improve the urban environment and the quality of life of residents.

[0021] Figure 2 FIG. 1 is a block diagram of a garbage whole process IoT data processing module in a garbage whole process IoT sensing system based on RFID technology according to an embodiment of the present application. Figure 2As shown, in a specific embodiment of the present application, the garbage whole process IoT data processing module 120 includes: a garbage bin filling state feature extraction unit 121, used to perform feature extraction on the garbage bin filling state data to obtain a garbage bin filling state semantic feature vector; a garbage generation time and place feature extraction unit 122, used to perform feature extraction on the garbage generation time and place data to obtain a global semantic feature vector of the garbage generation time and place data; a garbage multimodal information feature association unit 123, used to associate the garbage bin filling state semantic feature vector with the garbage generation time and place data global semantic feature vector to obtain the garbage multimodal information association feature vector; a garbage removal process feature extraction unit 124, used to perform feature extraction on the garbage removal process data to obtain the garbage removal process data.

[0022] It should be understood that feature extraction can transform the original trash can filling status data into a more abstract and meaningful semantic feature vector, which makes the data more suitable for machine learning and data mining technology. The semantic feature vector can capture the core characteristics of the trash can filling status, such as the changing trend of the garbage volume, filling speed and peak hours. These features may be implicit in the original data, but through feature extraction, the expressiveness of the data can be significantly improved, making it more operational.

[0023] Furthermore, the garbage generation time and location data provides information about the specific time and location of garbage generation, but the original data is usually scattered and lacks systematization. Through feature extraction, these data can be converted into useful global semantic feature vectors, thereby providing in-depth support for subsequent analysis and decision-making. Among them, by comprehensively considering time (such as peak hours, seasonal changes) and location (such as the amount of garbage generated in a specific area), the regularity of garbage generation can be revealed. For example, there may be a surge in garbage generation in certain time periods and regions. This information can help predict future garbage generation trends and adjust the frequency of collection and resource allocation.

[0024] Furthermore, the semantic feature vectors of trash bin filling status provide detailed information about the actual usage of trash bins, such as filling volume, changing trends, and full-load frequency. These feature vectors can reveal the real-time status and usage patterns of trash bins. The global semantic feature vectors of garbage generation time and location data provide a global perspective on the time and location of garbage generation, revealing the pattern and regional distribution of garbage generation. Combining the two can form a comprehensive understanding of the garbage management process, allowing the system to not only grasp the specific conditions of each trash bin, but also understand the macro trends of garbage generation. By integrating these feature vectors, association analysis can identify the relationship between garbage generation and filling status. For example, it can be analyzed which time periods and locations are more likely to reach full load status, and how these situations match the pattern of garbage generation. This analysis helps to predict and prevent the risk of trash bin overflow, thereby optimizing the collection plan, adjusting the collection frequency and route, and avoiding environmental pollution.

[0025] In particular, feature extraction converts the complex and diverse information in the garbage collection process data into concise and structured feature vectors, which helps to extract and summarize the key factors in the data. For example, by extracting features such as collection time, collection frequency, route optimization, and vehicle utilization, efficiency bottlenecks and potential problems in the collection process can be accurately identified. In this way, the system can more clearly understand the execution of each collection task and reveal the rules and anomalies in the collection process.

[0026] In a specific embodiment of the present application, the trash can filling state feature extraction unit 121 includes: segmenting the trash can filling state data and passing it through a word embedding encoder to obtain a sequence of trash can filling state word vectors; passing the sequence of trash can filling state word vectors through a trash can filling state semantic encoder to obtain the trash can filling state semantic feature vector.

[0027] It should be understood that, first of all, word segmentation is to decompose the information in the trash can filling status data into basic units. This process is very important for structuring unstructured data. Although the trash can filling status data is usually numerical data, it can be decomposed into different "words" or "fragments" by time, location, garbage volume, etc., so as to better understand various aspects of the data. Word segmentation can break down complex data into smaller, easier-to-handle units, making subsequent processing and analysis more detailed and accurate. The word embedding encoder converts these segmented data into word vectors. Word embedding technology maps each "word" or "fragment" to a high-dimensional vector space so that each data unit can be represented as a vector of fixed length. This vector not only contains the basic information of the vocabulary, but also captures the relationship and semantic similarity between words. In the scenario of trash can filling status data, this means that features such as filling amount, time and location can be converted into vectors, which is convenient for subsequent machine learning models to process. The trash can filling status word vector sequence obtained by the word embedding encoder can fully express the implicit information and complex relationships in the data. In this way, the time series characteristics, spatial distribution characteristics, etc. in the data can be effectively captured and represented, enabling the model to identify and learn important patterns and laws in waste management.

[0028] Furthermore, the word vector sequence converts the original data of the trash can filling state into numerical vectors through word embedding technology. These vectors capture the basic information and some simple semantic relationships in the data. However, although these word vectors provide a preliminary representation of the data, they fail to deeply capture the complex patterns and high-level semantic information in the trash can filling state. In order to solve this problem, the trash can filling state semantic encoder further processes these word vector sequences to extract the semantic feature vectors of the trash can filling state. Semantic encoders usually use deep learning techniques, such as recurrent neural networks (RNN), long short-term memory networks (LSTM) or transformers, to analyze and learn contextual information and deep relationships in the sequence. These encoders can capture the time series characteristics of the trash can filling state, filling pattern changes and other complex dynamic relationships, thereby generating richer and more useful feature vectors. Specifically, the transformer-based Bert model of the trash can filling state semantic encoder is used to perform global context semantic encoding on the sequence of the trash can filling state word vectors to obtain multiple filling state feature vectors; and the multiple filling state feature vectors are cascaded to obtain the trash can filling state semantic feature vector.

[0029] In a specific embodiment of the present application, the garbage generation time and place feature extraction unit 122 includes: passing the garbage generation time and place data through a garbage generation time and place data text understanding model based on a context encoder to obtain multiple garbage generation time and place data text feature vectors; arranging the multiple garbage generation time and place data text feature vectors in one dimension to obtain a garbage generation time and place data text splicing semantic feature vector; passing the garbage generation time and place data text splicing semantic feature vector through a garbage generation time and place data text multi-scale feature encoder to obtain the garbage generation time and place data global semantic feature vector.

[0030] It should be understood that the garbage generation time and location data usually include the specific time and location information of the garbage generation, which is crucial for garbage management. However, the original time and location data may be isolated, unstructured, and lack sufficient contextual relevance, which makes the direct use of these data for analysis limited. Models based on context encoders (such as long short-term memory networks LSTM, transformers Transformer, etc.) can handle long-distance dependencies and complex contextual information in time series data. Specifically, the garbage generation time and location data are segmented to obtain a garbage generation word sequence; the embedding layer of the garbage generation time and location data text understanding model based on the context encoder is used to map each garbage generation word in the garbage generation word sequence to a word embedding vector to obtain a sequence of garbage generation word embedding vectors; the converter-based Bert model of the garbage generation time and location data text understanding model based on the context encoder is used to perform global context semantic encoding on the sequence of garbage generation word embedding vectors to obtain multiple garbage generation time and location data text feature vectors.

[0031] Furthermore, the text feature vectors of the garbage generation time and location data usually represent different dimensions of garbage generation, such as time, location, and amount of garbage generated. Each feature vector contains specific information of these dimensions, but if these feature vectors are directly used as input, they may not be able to fully express the complex relationship between time and location and their comprehensive impact on garbage generation. Here, the feature vectors are arranged in one dimension to form a concatenated semantic feature vector of the garbage generation time and location data, so that the structure is more consistent when the data is input into the model, reducing the model's complex processing of the input format and improving training efficiency.

[0032] Furthermore, although the concatenated semantic feature vector can integrate multiple features into a unified representation, this representation may still be difficult to capture the subtle differences and complex relationships in the data. The multi-scale feature encoder can analyze the characteristics of garbage generation from micro to macro scales by performing multi-level processing on the concatenated vector. This processing method allows the model to not only understand the local detailed features (such as the changes in garbage generation in a specific time period), but also grasp the global trends and patterns (such as the long-term changes in the overall garbage generation). Specifically, the multi-scale feature encoder usually includes a multi-level network structure, each layer can encode the data from a different perspective. For example, the low-level encoder may focus on local time changes, while the high-level encoder may focus on the overall time trend and long-term pattern. In this way, the encoder can combine the short-term fluctuations and long-term trends in the garbage generation data to generate a global semantic feature vector. This global perspective helps the model better understand the overall picture of garbage generation, so that multiple influencing factors can be considered when making predictions and optimizations. Specifically, the multi-scale feature encoder for the text of data on the time and place of garbage generation includes: a first convolution layer, a second convolution layer in parallel with the first convolution layer, and a cascade layer connected to the first convolution layer and the second convolution layer, wherein the first convolution layer uses a one-dimensional convolution kernel with a first scale, and the second convolution layer uses a one-dimensional convolution kernel with a second scale. More specifically, the first convolution layer of the multi-scale feature encoder for the text of data on the time and place of garbage generation is used to perform one-dimensional convolution encoding on the concatenated semantic feature vector of the text of data on the time and place of garbage generation to obtain a first-scale feature vector; the second convolution layer of the multi-scale feature encoder for the text of data on the time and place of garbage generation is used to perform one-dimensional convolution encoding on the concatenated semantic feature vector of the text of data on the time and place of garbage generation to obtain a second-scale feature vector; the first-scale feature vector and the second-scale feature vector are cascaded to obtain the global semantic feature vector of the data on the time and place of garbage generation.

[0033] Figure 3 FIG. 1 is a block diagram of a garbage collection process feature extraction unit in a garbage collection process IoT sensing system based on RFID technology according to an embodiment of the present application. Figure 3 As shown, in a specific embodiment of the present application, the garbage removal process feature extraction unit 124 includes: a garbage removal process convolutional encoding subunit 1241, used to pass the garbage removal process data through a converter-based garbage removal process text encoder to obtain multiple garbage removal process text feature vectors; a garbage removal process text cascade subunit 1242, used to cascade the multiple garbage removal process text feature vectors into a garbage removal process text semantic association feature vector.

[0034] It should be understood that garbage collection process data usually contains a large amount of text information, such as collection records, processing reports, and transportation routes. This text information contains multi-dimensional data about the time, location, quantity, and status of garbage collection, and there are complex relationships between these data. For example, the garbage collection situation at a specific location on a certain day may be related to multiple factors such as the amount of garbage generated at that location, the frequency of collection, and weather conditions. To extract useful information from this data, an encoder that can understand and integrate this multi-level and multi-dimensional information is needed. The self-attention mechanism in the transformer model can calculate the dependencies between the various parts of the sequence, so as to better understand the associations between different times, locations, and events in the garbage collection data. This mechanism can capture long-term dependencies in the data and help extract more comprehensive feature representations. Compared with traditional sequence models, transformers can process long sequence data more efficiently and support parallel computing. This can significantly improve the processing efficiency and speed when processing large-scale garbage collection data. Specifically, the garbage removal process data is converted into an embedding vector using the embedding layer of the converter-based garbage removal process text encoder to obtain a sequence of garbage removal embedding vectors; and the sequence of garbage removal embedding vectors is encoded based on global contextual semantics using the converter-based Bert model of the converter-based garbage removal process text encoder to obtain multiple garbage removal process text feature vectors.

[0035] Furthermore, feature vector concatenation is a method of combining multiple feature vectors into one long vector. Each feature vector usually encodes the data from a different perspective and dimension. For example, one feature vector may capture the time information of the garbage collection process, while another may focus on the location and amount of collection. When these feature vectors are concatenated together, they can form a comprehensive representation that contains richer information.

[0036] In a specific embodiment of the present application, the garbage collection process text cascading subunit 1242 includes: creating a GarbageCollectionJob task class for implementing the Job interface of Quartz, and defining the logic of cascading multiple garbage collection process text feature vectors into a garbage collection process text semantic association feature vector in the execute method; in the Quartz configuration file, defining a scheduler, a job, and a trigger, wherein the scheduler is responsible for managing the execution of the job, the job is the task actually executed, named garbageCollectionJob, belongs to group1, and the trigger defines when the job is executed, named garbageCollectionTrigger, belongs to group1, and is set to execute every 10 seconds; starting the Quartz scheduler to start executing the defined task; in the Gar In the bageCollectionJob class, the getTextFeatureVectors method is defined to obtain multiple garbage collection process text feature vectors; in the GarbageCollectionJob class, the concatenateFeatureVectors method is defined to concatenate multiple garbage collection process text feature vectors into a garbage collection process text semantic association feature vector; in the GarbageCollectionJob class, the processConcatenatedFeatureVector method is defined to process the concatenated garbage collection process text semantic association feature vector; in the GarbageCollectionJob class, the main method is defined to create and start the Quartz scheduler, define jobs and triggers, and schedule them for execution.

[0037] It should be understood that the implementation steps of the garbage collection process text feature vector cascade system based on the Quartz mechanism include installing the Quartz framework, defining the task class, configuring the scheduler, starting the scheduler, obtaining the text feature vector, cascading the feature vector, processing the cascaded feature vector, and the main program entry. These steps ensure the reliability and efficiency of the system.

[0038] Some sample codes are as follows.

[0039]

[0040]

[0041]

[0042]

[0043] It should be understood that the above deployment code shows how to use the Quartz framework to implement the task of cascading multiple garbage collection process text feature vectors into a garbage collection process text semantic association feature vector. Through the above deployment code, a garbage collection process text feature vector cascading system based on the Quartz framework is implemented. The Quartz framework provides a powerful task scheduling function that can automatically execute defined tasks without manual intervention. In this example, the task is executed every 10 seconds, ensuring the real-time processing and updating of garbage collection process data. By cascading multiple garbage collection process text feature vectors into a garbage collection process text semantic association feature vector, the system can efficiently process and analyze the data in the garbage collection process and extract valuable information. The system can monitor the garbage collection process in real time and generate a processing progress report. This helps to improve the transparency and efficiency of garbage disposal and provide real-time data support for managers. The system adopts a modular design, and each functional module (such as obtaining text feature vectors, cascading feature vectors, and processing cascaded feature vectors) is independently implemented, which is easy to maintain and expand. The flexibility and scalability of the Quartz framework enable the system to adapt to different business needs. By adjusting the scheduler configuration, you can easily change the execution frequency and trigger conditions of the task.

[0044] In this way, the garbage collection process text feature vector cascade system based on the Quartz framework realizes comprehensive monitoring and data analysis of the garbage collection process through automated task scheduling, efficient data processing, real-time monitoring and reporting, modular design, flexibility and scalability, and improves the efficiency and transparency of garbage disposal.

[0045] GarbageCollectionJob is a task class that implements the Job interface of the Quartz framework and is responsible for performing the cascade operation of the text feature vectors of the garbage collection process in the execute method. garbageCollectionJob is the job name of the task registered in the Quartz scheduler and belongs to group1. garbageCollectionTrigger is the trigger name, which also belongs to group1 and is set to execute every 10 seconds. The getTextFeatureVectors method is used to obtain text feature vectors, the concatenateFeatureVectors method is used to concatenate feature vectors, and the processConcatenatedFeatureVector method is used to process the concatenated feature vectors. The main method is the main program entry, which is used to start the Quartz scheduler and schedule jobs and triggers.

[0046] Specifically, execute is a method of the Job interface in the Quartz framework. All task classes that implement the Job interface must implement this method. The execute method defines the specific execution logic of the task. When the Quartz scheduler triggers the task execution, this method will be called. In the GarbageCollectionJob class, the execute method is responsible for cascading multiple garbage collection process text feature vectors into a garbage collection process text semantic association feature vector and performing subsequent processing. By implementing the execute method, the task class can respond to the call of the scheduler and execute the predetermined task logic.

[0047] More specifically, group1 is a logical grouping used in the Quartz scheduler to organize and manage jobs and triggers. In the Quartz framework, jobs and triggers can be assigned to different groups to better manage and classify tasks. In this example, both the garbageCollectionJob job and the garbageCollectionTrigger trigger are assigned to group1. By grouping jobs and triggers, batch operations and management can be more convenient, such as starting, stopping, or modifying multiple tasks in the same group. The use of group1 improves the flexibility and maintainability of task management.

[0048] In the above-mentioned garbage whole process IoT perception system 100 based on RFID technology, the garbage treatment progress report generation module 130 is used to generate a garbage treatment progress report based on the garbage multimodal information association feature vector and the garbage removal process text semantic association feature vector. It should be understood that garbage multimodal information usually includes data from different sources, such as sensor data, images, videos, geographic location data, etc. These multimodal information can provide detailed data on garbage volume, garbage type, removal frequency, equipment status, etc. The garbage removal process text semantic association feature vector is a feature vector generated based on text data, which includes removal records, operation logs, reports and notifications. Through in-depth analysis of these text data, the specific details of garbage removal, operational problems, personnel feedback and other semantic information can be understood. These feature vectors can capture the context and semantic information in the garbage treatment process, which helps to deeply understand the dynamics and problems of garbage treatment. By combining these features, a comprehensive progress report can be created, which includes both data-driven objective indicators and specific situations and problems in the process.

[0049] Figure 4 The block diagram of the garbage processing progress report generation module in the garbage whole process IoT perception system based on RFID technology according to the embodiment of the present application. Figure 4As shown, in a specific embodiment of the present application, the garbage disposal progress report generation module 130 includes: a garbage disposal feature fusion unit 131, which is used to fuse the garbage multimodal information association feature vector and the garbage removal process text semantic association feature vector to obtain a garbage disposal progress generation feature vector; a garbage disposal feature optimization unit 132, which is used to perform gated pruning adjustment on the garbage disposal progress generation feature vector based on sparse optimization to obtain a sparse optimized garbage disposal progress generation feature vector; a progress report generation unit 133, which is used to pass the sparse optimized garbage disposal progress generation feature vector through a generator to generate a garbage disposal progress report.

[0050] It should be understood that by fusing data from different sources, the biases and deficiencies that may exist in a single data source can be eliminated. For example, sensor data provides real-time information on the amount of garbage, while text data can explain the background and context of this data, thereby improving the accuracy of progress descriptions.

[0051] In particular, in the technical solution of the present application, the feature vector generated by the progress of garbage disposal is formed by the fusion of multimodal information of the filling state of the garbage bin, the time and place of garbage generation, and the garbage removal process, and contains rich contextual information and diverse features, such as different times, places and filling states. This high-dimensional and diverse information has advantages in expressing the global state of garbage disposal, but if the dimension design of the learning parameter matrix of the generator is inconsistent with the dimension of the feature vector generated by the progress of garbage disposal, or the parameter setting fails to cover the diversity of the feature vector, it may cause the model to focus only on certain specific features and ignore other equally important aspects. This deviation will aggravate the information confusion between features, resulting in unnecessary noise or repeated features in the generated garbage disposal progress report, thereby further causing the overfitting of the model, resulting in the final generated report may not truly reflect the actual garbage disposal situation. Therefore, in the technical solution of the present application, the feature vector generated by the progress of garbage disposal is adjusted by gated pruning based on sparse optimization to obtain a sparse optimized feature vector generated by the progress of garbage disposal.

[0052] Among them, the garbage disposal progress generation feature vector is subjected to gated pruning adjustment based on sparse optimization to obtain a sparse optimized garbage disposal progress generation feature vector, including: extracting each row vector of the learning parameter matrix in the generator as a set of garbage disposal progress learning parameter row vectors; calculating the garbage disposal progress entropy degradation maximization simulation value between the garbage disposal progress generation feature vector and each garbage disposal progress learning parameter row vector in the set of garbage disposal progress learning parameter row vectors to obtain a set of garbage disposal progress entropy degradation maximization simulation values; based on the comparison between each garbage disposal progress entropy degradation maximization simulation value in the set of garbage disposal progress entropy degradation maximization simulation values ​​and a preset threshold, the learning parameter matrix is ​​subjected to gated pruning processing to obtain a garbage disposal progress learning parameter entropy degradation optimization matrix; matrix multiplication of the garbage disposal progress generation feature vector and the garbage disposal progress learning parameter entropy degradation optimization matrix is ​​performed to obtain the sparse optimized garbage disposal progress generation feature vector.

[0053] The optimization steps are specifically expressed as follows:

[0054] P={p1,p2,...,p i ,...,p n}

[0055]

[0056] w i =mask(D i )

[0057]

[0058] P s ={w1·p1,w2·p2,...,w i ·p i ,...,w n ·p n}

[0059]

[0060] Among them, P is the set of row vectors of garbage disposal progress learning parameters, {…} is a set operation, p1, p2, p i 、p n are the eigenvalues ​​of the first, second, i-th, and n-th garbage disposal progress learning parameter row vectors in the set of garbage disposal progress learning parameter row vectors, respectively. i is the ith garbage disposal progress entropy degradation maximization simulation value in the set of garbage disposal progress entropy degradation maximization simulation values, max is the maximum value function, D(v1,p i ) represents the calculation of vectors v1 and p iThe Euclidean distance between them, v1 is the feature vector generated by the garbage disposal progress, p i T is the transposed vector of the row vector of the learning parameter of the i-th garbage disposal progress, L is the length of the feature vector, is matrix multiplication, w1, w2, w i 、w n are the first, second, i-th, and n-th learning parameter mask values, respectively. Mask is the mask operation, θ represents the preset threshold, and P s The entropy degradation optimization matrix for learning parameters of garbage disposal schedule is presented, and v1' is the feature vector generated for sparse optimization of garbage disposal schedule.

[0061] In the technical solution of the present application, it is taken into consideration that in the process of inputting the garbage disposal progress generating feature vector into the learning parameter matrix of the generator to obtain the garbage disposal progress report, there is a mismatch between the structure of the learning parameter matrix itself and the garbage disposal progress generating feature vector, that is, the learning parameter matrix has structural redundancy and information confusion relative to the garbage disposal progress generating feature vector, which may cause feature redundancy or noise in the garbage disposal progress report, resulting in model overfitting.

[0062] Based on this, in the technical solution of the present application, the garbage disposal progress generation feature vector is subjected to gated pruning adjustment based on sparse optimization, which uses the garbage disposal progress generation feature vector as a wandering node, uses the feature distribution between the garbage disposal progress generation feature vector and each row vector in the learning parameter matrix to simulate the wandering pattern, and uses the distance topology between the garbage disposal progress generation feature vector and each row vector in the learning parameter matrix as a wandering map. In this way, the wandering result of the wandering node in a specific wandering pattern on the wandering map is used to obtain the garbage disposal progress entropy degradation maximization simulation value. Furthermore, based on the comparison between each garbage disposal progress entropy degradation maximization simulation value in the set of garbage disposal progress entropy degradation maximization simulation values ​​and a preset threshold, the part of the learning parameter matrix that does not contribute significantly to the input data representation and may even introduce noise or redundant information is screened out. Specifically, based on the comparison result, the learning parameter matrix is ​​subjected to gated pruning processing to obtain the garbage disposal progress learning parameter entropy degradation optimization matrix. Finally, the garbage disposal progress generation feature vector is linearly modulated using the structure-optimized learning parameter matrix as a feature modulation unit to obtain the sparse optimized garbage disposal progress generation feature vector.

[0063] Furthermore, the sparse optimization of garbage disposal progress generates feature vectors that incorporate features from multiple data sources, such as garbage volume, removal frequency, equipment status, and operation records. These features provide detailed data on the garbage disposal process, but directly analyzing these vectors may be complex and difficult to intuitively understand. The role of the generator is to convert the sparse optimization of garbage disposal progress generation feature vector representation into a structured report to facilitate analysis and decision-making by relevant personnel. The generator can integrate the information in the feature vector and present it in the form of charts, tables, or text, making the garbage disposal progress report more intuitive and easy to understand. This visualization can help users quickly grasp the processing progress, identify problem points, and focus on key points.

[0064] In another embodiment of the present application, an RFID device is installed at a garbage collection point, and an RFID tag is installed at a garbage bin. When the garbage bin is placed at the garbage collection point, the RFID device recognizes the tag and records the time and location. An RFID device is installed on the garbage truck. When the truck arrives at the collection point and collects the garbage bin, the RFID device on the truck recognizes the tag and records the time and location of the pickup. The garbage truck records the time and location after entering the garbage disposal site. The entire system uses RFID recognition technology to realize the IoT perception of the entire garbage process and present data in a timely manner.

[0065] In summary, the embodiment of the present application first obtains the garbage bin filling status data, garbage generation time and location data, and garbage removal process data collected by the RFID device, and then uses deep learning technology to perform feature extraction and correlation analysis on the three, and finally generates a garbage disposal progress report through a generator, so as to adjust the garbage removal strategy and resource allocation strategy according to the progress report to ensure that the garbage is handled in a timely and effective manner, thereby optimizing the overall processing efficiency.

[0066] As described above, the garbage whole process IoT perception system 100 based on RFID technology according to the embodiment of the present application can be implemented in various terminal devices. In one example, the garbage whole process IoT perception system 100 based on RFID technology can be integrated into the terminal device as a software module and / or hardware module. For example, the garbage whole process IoT perception system 100 based on RFID technology can be a software module in the operating system of the terminal device, or can be an application developed for the terminal device; of course, the garbage whole process IoT perception system 100 based on RFID technology can also be one of the many hardware modules of the terminal device.

[0067] Alternatively, in another example, the garbage full-process IoT perception system 100 based on RFID technology and the terminal device may also be separate devices, and the garbage full-process IoT perception system 100 based on RFID technology may be connected to the terminal device through a wired and / or wireless network, and transmit interactive information in accordance with an agreed data format.

[0068] The basic principles of the present disclosure are described above in conjunction with specific embodiments. However, it should be noted that the advantages, strengths, effects, etc. mentioned in the present disclosure are only examples and not limitations, and it cannot be considered that these advantages, strengths, effects, etc. are required by each embodiment of the present disclosure. In addition, the specific details disclosed above are only for the purpose of illustration and ease of understanding, and are not limitations. The above details do not limit the present disclosure to the necessity of adopting the above specific details to be implemented.

[0069] The block diagrams of the devices, apparatuses, equipment, and systems involved in this disclosure are only illustrative examples and are not intended to require or imply that they must be connected, arranged, and configured in the manner shown in the block diagrams. As will be appreciated by those skilled in the art, these devices, apparatuses, equipment, and systems can be connected, arranged, and configured in any manner. Words such as "including," "comprising," "having," and the like are open words, referring to "including but not limited to," and can be used interchangeably therewith. The words "or" and "and" used herein refer to the words "and / or," and can be used interchangeably therewith, unless the context clearly indicates otherwise. The word "such as" used herein refers to the phrase "such as but not limited to," and can be used interchangeably therewith.

[0070] Additionally, as used herein, "or" used in a list of items beginning with "at least one" indicates a separate list, so that, for example, a list of "at least one of A, B, or C" means A or B or C, or AB or AC or BC, or ABC (i.e., A and B and C). Furthermore, the word "exemplary" does not mean that the example described is preferred or better than other examples.

[0071] It should also be noted that in the system and method of the present disclosure, each component or each step can be decomposed and / or recombined. Such decomposition and / or recombination should be regarded as equivalent solutions of the present disclosure.

[0072] Various changes, substitutions, and modifications of the techniques described herein may be made without departing from the teachings defined by the appended claims. Furthermore, the scope of the claims of the present disclosure is not limited to the specific aspects of the processes, machines, manufactures, compositions of events, means, methods, and actions described above. Currently existing or later to be developed processes, machines, manufactures, compositions of events, means, methods, or actions that perform substantially the same functions or achieve substantially the same results as the corresponding aspects described herein may be utilized. Thus, the appended claims include such processes, machines, manufactures, compositions of events, means, methods, or actions within their scope.

[0073] The above description of the disclosed aspects is provided to enable any person skilled in the art to make or use the present disclosure. Various modifications to these aspects will be readily apparent to those skilled in the art, and the general principles defined herein may be applied to other aspects without departing from the scope of the present disclosure. Therefore, the present disclosure is not intended to be limited to the aspects shown herein, but rather to the widest scope consistent with the principles and novel features disclosed herein.

[0074] The above description has been given for the purpose of illustration and description. In addition, this description is not intended to limit the embodiments of the present disclosure to the forms disclosed herein. Although multiple example aspects and embodiments have been discussed above, those skilled in the art will recognize certain variations, modifications, changes, additions and sub-combinations thereof.

Claims

1. A garbage whole process IoT sensing system based on RFID technology, characterized by: include: The whole process of garbage IoT data acquisition module is used to obtain the garbage bin filling status data, garbage generation time and location data and garbage removal process data collected by RFID equipment, among which the garbage removal process data includes the time and location of the garbage bin being put into the garbage collection point, the time and location of the garbage removal truck picking up the garbage bin at the collection point, and the time and location of the garbage removal truck entering the garbage disposal site; A garbage whole process IoT data processing module is used to extract garbage multimodal information association feature vectors and garbage removal process text semantic association feature vectors from the garbage bin filling status data, the garbage generation time and location data, and the garbage removal process data; The garbage disposal progress report generation module is used to generate a garbage disposal progress report based on the garbage multimodal information association feature vector and the garbage removal process text semantic association feature vector.

2. The garbage whole process IoT sensing system based on RFID technology according to claim 1 is characterized by: The garbage whole process IoT data processing module includes: A trash can filling state feature extraction unit, used for extracting features from the trash can filling state data to obtain a trash can filling state semantic feature vector; A garbage generation time and place feature extraction unit, used for extracting features from the garbage generation time and place data to obtain a global semantic feature vector of the garbage generation time and place data; A garbage multimodal information feature association unit, used for associating the garbage bin filling state semantic feature vector with the garbage generation time and location data global semantic feature vector to obtain the garbage multimodal information association feature vector; The garbage removal process feature extraction unit is used to extract features from the garbage removal process data to obtain the garbage removal process data.

3. The garbage whole process IoT sensing system based on RFID technology according to claim 2 is characterized by: The trash can filling state feature extraction unit comprises: The trash can filling state data is segmented and passed through a word embedding encoder to obtain a sequence of trash can filling state word vectors; The sequence of the trash can filling state word vectors is passed through a trash can filling state semantic encoder to obtain the trash can filling state semantic feature vector.

4. The garbage whole process IoT sensing system based on RFID technology according to claim 3 is characterized by: The garbage generation time and location feature extraction unit comprises: The garbage generation time and location data are passed through a garbage generation time and location data text understanding model based on a context encoder to obtain a plurality of garbage generation time and location data text feature vectors; Arrange the plurality of garbage generation time and location data text feature vectors in one dimension to obtain a garbage generation time and location data text splicing semantic feature vector; The semantic feature vector of the garbage generation time and place data text concatenation is passed through a garbage generation time and place data text multi-scale feature encoder to obtain a global semantic feature vector of the garbage generation time and place data.

5. The garbage whole process IoT sensing system based on RFID technology according to claim 4 is characterized in that: The garbage removal process feature extraction unit comprises: A garbage removal process convolutional encoding subunit, used for passing the garbage removal process data through a converter-based garbage removal process text encoder to obtain a plurality of garbage removal process text feature vectors; The garbage removal process text cascading subunit is used to cascade the multiple garbage removal process text feature vectors into a garbage removal process text semantic association feature vector.

6. The garbage whole process IoT sensing system based on RFID technology according to claim 5 is characterized by: The garbage removal process text cascade sub-unit includes: Create a GarbageCollectionJob task class to implement the Quartz Job interface, and define the logic of concatenating multiple garbage collection process text feature vectors into a garbage collection process text semantic association feature vector in the execute method; In the Quartz configuration file, define the scheduler, job, and trigger. The scheduler is responsible for managing the execution of the job. The job is the actual task that is executed. It is named garbageCollectionJob and belongs to group1. The trigger defines when the job is executed. It is named garbageCollectionTrigger, belongs to group1, and is set to execute every 10 seconds. Start the Quartz scheduler to start executing the defined tasks; In the GarbageCollectionJob class, define the getTextFeatureVectors method to obtain multiple garbage collection process text feature vectors; In the GarbageCollectionJob class, define the concatenateFeatureVectors method to concatenate multiple garbage collection process text feature vectors into a garbage collection process text semantic association feature vector; In the GarbageCollectionJob class, define The processConcatenatedFeatureVector method is used to process the concatenated garbage removal process text semantic association feature vector; In the GarbageCollectionJob class, define the main method to create and start the Quartz scheduler, define jobs and triggers, and schedule them for execution.

7. The garbage whole process IoT sensing system based on RFID technology according to claim 6 is characterized by: The garbage disposal progress report generating module comprises: A garbage disposal feature fusion unit, used to fuse the garbage multimodal information association feature vector and the garbage removal process text semantic association feature vector to obtain a garbage disposal progress generation feature vector; A garbage disposal feature optimization unit, used for performing gated pruning adjustment based on sparse optimization on the garbage disposal progress generation feature vector to obtain a sparse optimized garbage disposal progress generation feature vector; The progress report generating unit is used to generate a garbage disposal progress report by passing the sparse optimized garbage disposal progress generating feature vector through a generator.

8. The garbage whole process IoT sensing system based on RFID technology according to claim 7 is characterized in that: The garbage disposal feature optimization unit comprises: Extracting each row vector of the learning parameter matrix in the generator as a set of garbage disposal progress learning parameter row vectors; Calculating the garbage disposal progress entropy degradation maximization simulation value between the garbage disposal progress generation feature vector and each garbage disposal progress learning parameter row vector in the set of garbage disposal progress learning parameter row vectors to obtain a set of garbage disposal progress entropy degradation maximization simulation values; Based on the comparison between each garbage disposal progress entropy degradation maximization simulation value in the set of garbage disposal progress entropy degradation maximization simulation values ​​and a preset threshold, the learning parameter matrix is ​​gated and pruned to obtain a garbage disposal progress learning parameter entropy degradation optimization matrix; The garbage disposal progress generation feature vector is matrix-multiplied with the garbage disposal progress learning parameter entropy degradation optimization matrix to obtain the sparse optimized garbage disposal progress generation feature vector.

Images