A renewable resource recycling data management system based on the Internet of Things

Abstract

This invention discloses an IoT-based data management system for recycled resources, relating to the fields of IoT data management and intelligent supervision of recycled resources. The system comprises: a terminal acquisition layer, an edge computing layer, a cloud service layer, and an application layer. The terminal acquisition layer covers core nodes across the entire recycling chain, collecting native business data throughout the process and generating a globally unique batch identifier for each batch of recycled resources. The edge computing layer performs local cleaning and structuring of the data, generating an immutable data fingerprint bound to the batch identifier. The cloud service layer incorporates cross-verification, data storage, and traceability early warning modules. The application layer adapts to multiple entities, enabling data management, compliance verification, and early warning handling. This invention, through multi-dimensional cross-verification across the entire chain and RFID physical binding design, combined with a trusted edge environment and consortium blockchain evidence storage architecture, fundamentally solves the problem of data falsification in recycling, achieving trusted full-process data control.

Description

Technical Field

[0001] This invention relates to the field of Internet of Things (IoT) data management and intelligent supervision of renewable resources, specifically to an IoT-based data management system for renewable resource recycling. Background Technology

[0002] After years of development, the recycling industry has formed a complete chain, encompassing individual waste disposal, collection at recycling stations, transportation and transshipment, sorting and processing, and reuse by waste utilization companies. Currently, most large-scale recycling sites and sorting centers are equipped with intelligent weighing equipment, video surveillance, and basic data recording tools, gradually replacing the traditional manual ledger system. As management requirements for the recycling industry continue to tighten across regions, the verifiability and reliability of recycling data have become core prerequisites for compliant operation by recycling entities and for routine inspections by regulatory authorities. There is a clear and urgent practical need within the industry for a management system that can truly control the authenticity of data.

[0003] Most existing recycling data management systems on the market can only identify basic weight anomalies and store data at a single site, making it impossible to prevent collusion among multiple entities across the entire recycling chain to falsify data and create empty ledgers. The verification logic of these systems is limited to data entered internally, without cross-validation with tamper-proof third-party data, making them vulnerable to exploitation. Furthermore, existing systems generally lack reliable control over data sources; data is often uploaded for documentation only after entry, making tampering during the entry process undetectable. Data gaps during transportation also prevent closed-loop verification across the entire chain, leaving the problems of fraudulent recycling and false ledgers unresolved. Summary of the Invention

[0004] The purpose of this invention is to overcome the shortcomings of existing technologies and provide an Internet of Things-based data management system for recycled resources. This system employs a multi-dimensional, end-to-end cross-verification design, using a unique batch identifier to link the entire process of recycled resources from deployment to reuse. First, it verifies the consistency of data entering and leaving the warehouse in adjacent stages according to the physical flow sequence. Then, it cross-verifies with third-party, tamper-proof data such as tax and electricity consumption data, thus preventing collusion and fraud at the source. Simultaneously, it utilizes a trusted edge execution environment and disposable anti-counterfeiting RFID ties to avoid tampering risks from the data collection source. Furthermore, through a hierarchical permission-based consortium blockchain for evidence storage and hierarchical early warning closed-loop rules, it protects the business information of operating entities and achieves trusted control over the entire recycling data process, solving the long-standing industry problems of false ledgers and lack of verification documentation.

[0005] To solve the above-mentioned technical problems, the present invention provides the following technical solution: an Internet of Things-based data management system for recycling renewable resources, the system comprising a terminal acquisition layer, an edge computing layer, a cloud service layer, and an application layer; The terminal acquisition layer covers the front-end placement, recycling stations, transportation, sorting, and waste utilization enterprise nodes of the entire recycling chain. Each node is configured with a corresponding acquisition terminal to collect native business data of the entire recycling process and generate a globally unique batch identifier code for each batch of recycled resources. The edge computing layer is communicatively connected to each acquisition terminal, including edge gateways deployed on each node, used to perform local cleaning and structuring processing on native business data, and generate an immutable data fingerprint that is uniquely bound to the batch identifier code; The cloud service layer communicates with the edge computing layer and has a built-in end-to-end multi-dimensional cross-verification module, data storage module, and source tracing and early warning module. The application layer communicates with the cloud service layer and includes functional ports for corresponding recycling operators, waste utilization enterprises, and regulatory departments, used to complete data management, compliance verification, and early warning and handling.

[0006] Furthermore, the full-link multi-dimensional cross-verification module uses the batch identifier code as an index to first perform full-link upstream and downstream chain logic verification, then connects to a third-party tamper-proof data source to perform external cross-verification, and generates the authenticity score of the batch data through the full-link data authenticity scoring model. The data storage module is used to store fingerprints, verification process data, and authenticity scoring results of batch data throughout their entire lifecycle. The source tracing and early warning module is used to perform full-chain source tracing for batches whose authenticity score is lower than a preset threshold, and push early warning information according to the level of abnormality.

[0007] Furthermore, the edge gateway has a built-in Trusted Execution Environment (TEE) that is physically isolated from the device operating system. The entire process of cleaning, structuring, and generating data fingerprints of native business data is completed within the TEE. The edge gateway has a built-in lightweight anomaly detection model, which is used to perform real-time anomaly detection on the collected weight, category, and location data locally, and to perform local interception and early warning for abnormal data that exceeds a preset threshold. The edge gateway supports offline operation when the network is down. It completes local data processing and encrypted storage when the network is down, and automatically synchronizes the data to the cloud service layer after the network is restored.

[0008] Furthermore, the recycling station node of the terminal collection layer is equipped with an RFID tag writing device and a disposable anti-counterfeiting RFID cable tie. The unique batch identification code is written into the passive RFID tag built into the cable tie. The cable tie is used to package the corresponding batch of recycled resources. After the cable tie is cut, the built-in RFID tag is physically damaged, cannot be read and cannot be reused. The transportation nodes, sorting nodes, and waste utilization enterprise nodes are all equipped with RFID reading and writing devices. Each node must scan the corresponding batch cable ties that have not expired through the RFID reading and writing device in order to complete the node confirmation and entry of the batch data.

[0009] Furthermore, when the full-link multi-dimensional cross-verification module performs upstream and downstream chain-like logical verification, it uses the batch identifier code as an index and performs relative deviation verification on the same batch and same type of recycled resource data of adjacent nodes according to the physical flow sequence. The verification object is the outbound weight data of the previous node and the inbound weight data of the next node. Node data with a relative deviation exceeding the preset threshold is marked as abnormal data. The preset threshold is set in a fixed manner according to the node type. The threshold for the front-end delivery node and the recycling station node is 3%, and the threshold for the other adjacent nodes is 2%.

[0010] Furthermore, the third-party tamper-proof data sources connected to the full-link multi-dimensional cross-verification module include tax invoice systems, power consumption monitoring systems, and commodity price information systems; When performing external cross-validation, at least three validation rules must be included: First, the matching and verification of the batch production input data of waste utilization enterprises with the power consumption data of production equipment in the same production cycle; Secondly, verify the consistency of category, quantity, and amount between the corresponding batch of recycled transaction data and the corresponding tax invoice data; Third, verify the reasonableness of the corresponding batch recycling price and the national average market price of the same type of bulk commodity in the natural week in which the transaction occurred.

[0011] Furthermore, the full-link data authenticity scoring model takes the consistency results of the full-link chain logic verification, the matching degree results of external cross-verification, the rationality of historical data fluctuations in the same dimension, and the compliance of the operation behavior of all process nodes as core feature dimensions, and outputs an authenticity score ranging from 0 to 100. The cloud service layer has a built-in sub-module for dynamically adaptively adjusting the scoring threshold. It uses the historical compliance record of the recycling entity, the risk level of counterfeiting of the recycling category, regional regulatory requirements, and market price fluctuations as input features to dynamically adjust the preset threshold for the authenticity score of the corresponding batch.

[0012] Furthermore, the data storage module adopts a consortium blockchain architecture, where recycling operators, waste utilization companies, and regulatory departments are all peer nodes in the consortium blockchain, and each node is configured with independent and fixed hierarchical access permissions: Recycling operators can only access the on-chain data they generate; waste utilization enterprises can only access the on-chain data corresponding to their own batches; and regulatory authorities can access all on-chain data within their jurisdiction. All on-chain data can only be written using the private key of the corresponding consortium blockchain node generated from the data. Once written, the data cannot be tampered with, cannot be deleted, and is traceable throughout the entire process.

[0013] Furthermore, the source tracing and early warning module is divided into three non-overlapping risk levels based on the authenticity score: >40 points and ≤60 points are general risk, >20 points and ≤40 points are medium risk, and ≤20 points are major risk. The tiered early warning push rules are fixed as follows: general risk early warning information is only pushed to the corresponding recycling business entity; medium risk early warning information is simultaneously pushed to the corresponding recycling business entity, waste utilization enterprise and local regulatory department; and major risk early warning information is immediately pushed to all relevant entities and regulatory departments.

[0014] Furthermore, as stated above.

[0015] Compared with existing technologies, this Internet of Things-based data management system for recycling renewable resources has the following advantages: I. This invention employs a core design of full-chain, multi-dimensional cross-verification, using a unique batch identifier to link data across the entire process from placement, collection, transportation, sorting to reuse. First, it verifies the consistency of inbound and outbound data in adjacent stages according to the physical flow sequence. Then, it cross-verifies with three types of tamper-proof third-party data: tax invoices, electricity consumption, and market information, fundamentally blocking loopholes for collusion and fraud at multiple stages. Simultaneously, it binds the batch identifier to disposable anti-counterfeiting RFID straps, ensuring seamless data-physical correspondence. Furthermore, an authenticity scoring model quantifies the data's credibility, allowing the recycling entity's daily operational records and regulatory department's verification work to directly access the full-chain traceable data for each batch, solving the previous problems of lacking clear verification methods and data documentation.

[0016] Second, this invention utilizes an edge-end trusted execution environment combined with a consortium blockchain for evidence storage. The entire process of data cleaning, format processing, and unique fingerprint generation is completed in a trusted environment physically isolated from the main device system. This avoids the risk of data tampering from the data collection source. The edge gateway can also operate normally even when offline, adapting to the usage needs of different scenarios such as urban sites and remote rural recycling points. Simultaneously, by storing the entire process data through a consortium blockchain with tiered permissions, coupled with tiered early warning and rectification closed-loop rules, it protects the business information of operating entities and achieves trusted control over the entire data process.

[0017] Other advantages, objectives and features of the invention will be set forth in part in the description which follows, and in part will be apparent to those skilled in the art from the following examination or study, or may be learned from the practice of the invention. Attached Figure Description

[0018] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the accompanying drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are merely some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without any creative effort.

[0019] Figure 1 This is a schematic diagram of the four-layer architecture of the system of the present invention; Figure 2 This is a schematic diagram of the entire data flow of renewable resources according to the present invention; Figure 3 This is a schematic diagram of the closed-loop process for authenticity scoring, grading, early warning, and handling in this invention. Detailed Implementation

[0020] To further illustrate the technical means and effects of the present invention in achieving its intended purpose, the following detailed description of the specific implementation methods, structures, features, and effects of the present invention, in conjunction with the accompanying drawings and preferred embodiments, is provided below.

[0021] This embodiment details the specific implementation of the IoT-based data management system for recycled resources disclosed in this invention. Its core objective is to address the fundamental problems of existing recycling data management systems, such as easily tampered data sources, inability to detect collusion and fraud across multiple stakeholders, and the lack of closed-loop verification logic. This embodiment fully elucidates the hardware composition, working principles, and data flow logic of each module at each level of the system. It details the application methods and parameter meanings of core algorithms and mathematical formulas, and completely reconstructs the entire system operation mechanism, ensuring that those skilled in the art can fully understand and reproduce all the technical solutions of this system.

[0022] This system adopts a four-layer loosely coupled and scalable architecture, consisting of the application layer, cloud service layer, edge computing layer, and terminal acquisition layer from top to bottom. For example... Figure 1 As shown, the four-layer architecture completes data interaction through communication links encrypted with national cryptographic algorithms, uses the national cryptographic SM2 algorithm for identity authentication and signature, and the SM4 algorithm for symmetric encryption of communication data, forming a complete technical closed loop of data collection, trusted processing, cross-verification, evidence storage and traceability, and closed-loop processing.

[0023] The terminal data acquisition layer covers the entire recycling chain, including front-end delivery nodes, recycling station nodes, transportation nodes, sorting center nodes, and waste utilization enterprise nodes. Each node is equipped with a corresponding data acquisition terminal to collect native business data from the entire process of recycling resources, from delivery, collection, transportation, sorting to resource utilization. At the same time, a globally unique batch identifier code is generated for each batch of recycling resources at the recycling station node.

[0024] The batch identifier uses an 18-bit fixed-length code, which includes three fields: recycling station node number, generation timestamp, and random check code, to ensure global uniqueness and non-repeating.

[0025] The front-end delivery node is equipped with a smart recycling bin, featuring a built-in weighing sensor with a weighing accuracy of ±50g, a high-definition camera with a resolution of 2 megapixels or higher, a QR code scanning module, and a touch screen. It collects data on the type, weight, time, location, delivery person, and video data of the delivery process. Once collected, the data is synchronized to the corresponding edge gateway in real time.

[0026] The recycling station node is equipped with a desktop weighing terminal, an industrial handheld terminal, a monitoring camera, an RFID tag writing device, and disposable anti-counterfeiting RFID cable ties. These are used to collect inbound and outbound data during the collection process, and simultaneously write the generated globally unique batch identification code into the passive RFID tag built into the cable tie. The RFID tag's antenna is embedded inside the cable tie itself. The cable tie is used to package the corresponding batch of recyclable resources. When the cable tie is cut, the built-in antenna breaks simultaneously, physically damaging the RFID tag, rendering it unreadable and unusable, thus physically preventing the tag from being reused for counterfeiting.

[0027] The transportation node is equipped with an intelligent vehicle-mounted terminal, integrating a dual-frequency GPS positioning module, a beam-type load-bearing sensor module with a weighing accuracy of ±10kg, an electronic door lock sensor module for the cargo compartment, four infrared night vision cameras, and a vehicle-mounted edge gateway. It is used to collect location, trajectory, load, cargo compartment opening / closing records, and video data inside the cargo compartment throughout the entire transportation process. The data collection frequency can be dynamically adjusted according to the vehicle's driving status; during vehicle movement, the data collection frequency is no less than once per minute, and when the vehicle is stationary, the collection frequency can be reduced to once every 10 minutes.

[0028] The sorting center nodes are equipped with weighbridge terminals, hyperspectral imaging recognition equipment, sorting line operation sensors, and RFID readers / writers. These are used to collect data from the entire sorting process, including inbound, sorting, and outbound operations. Simultaneously, the hyperspectral imaging equipment accurately identifies the type, purity, and impurity percentage of recycled resources. The identification results are then incorporated into the structured dataset for that batch, serving as a supplementary dimension for subsequent chain-based logical verification and authenticity scoring.

[0029] Waste utilization enterprises are equipped with weighing terminals, production equipment operation status acquisition modules, power consumption acquisition modules, and RFID reading and writing devices. These are used to collect data on raw material warehousing, production input, energy consumption, and finished product output in the resource utilization process. The energy consumption data is collected with precision down to a single production cycle of a single production line. The production cycle is the complete production process from raw material input to finished product output for a single batch of raw materials.

[0030] RFID readers at transportation nodes, sorting centers, and waste utilization enterprises must scan the corresponding unexpired cable ties tags for each batch in order to complete the node confirmation and entry of the batch data. If a tag scan fails or is identified as an expired tag, the system will prevent the batch data from entering the next process and trigger an anomaly warning to ensure that the data and physical goods are bound together throughout the entire process.

[0031] Each data acquisition terminal at each node in the terminal acquisition layer adapts to the corresponding communication method depending on the deployment scenario. In urban areas with good network coverage, 4G or 5G communication is used, while in remote towns and villages with weak network coverage, LoRa or NB-IoT low-power wide-area network communication is used to ensure that the data collected by all nodes can be stably transmitted to the corresponding edge gateway.

[0032] The edge computing layer communicates with each acquisition terminal in the terminal acquisition layer, with the core being the edge gateway deployed on each node. The edge gateway uses a trusted computing chip that supports national cryptographic standards and has a built-in trusted execution environment (TEE) that is physically isolated from the device's operating system at the hardware level. The entire process of cleaning, structuring, and generating data fingerprints of native business data is completed within the TEE, which cannot be accessed or tampered with by external programs within the device's main system, thus avoiding the risk of tampering from the source of data processing.

[0033] After receiving the native business data uploaded by the collection terminal, the edge gateway first performs data cleaning within the Trusted Execution Environment (TEE) to remove duplicate data, invalid data with missing key fields, and abnormal data that exceeds the reasonable weight range for the product category.

[0034] After the data cleaning is completed, the edge gateway performs structured processing on the cleaned data within the Trusted Execution Environment (TEE). According to the preset unified data format, the unstructured data from different terminals and nodes is converted into a standardized structured dataset.

[0035] After the structuring process is completed, the edge gateway generates an immutable data fingerprint that is uniquely bound to the corresponding batch identifier code through a secure hash algorithm within the Trusted Execution Environment (TEE).

[0036] The calculation formula used for data fingerprint generation is as follows: In the formula, H is the unique data fingerprint of a single batch of recycled resource data, which is a fixed 256-bit hash value; SHA-256 is the secure hash algorithm used in this embodiment, which is used to perform irreversible encryption calculation on the concatenated original data to ensure that any slight change in the original data will result in a complete change in the generated hash value, thus ensuring the uniqueness and immutability of the data fingerprint from the algorithm level. This is a standardized dataset of a single batch of recycled resources after structured processing, containing all core fields of the batch, including category, weight, time, location, and operator. The double vertical lines in the formula are binary byte stream concatenation operators, used to seamlessly concatenate the binary byte streams of preceding and following fields in sequence, ensuring that all related fields participate in fingerprint generation, and that tampering with a single field can trigger a complete change in the fingerprint. The data preprocessing is completed with millisecond-level timestamps, accurate to one-thousandth of a second, to ensure that fingerprints generated at different times in the same batch are unique and to prevent historical compliant data from being repeatedly used for fraud. It is a globally unique batch identifier for a single batch of recycled resources, a unique index for data association across the entire chain, and is bound to the corresponding physical batch throughout the entire process; This is a unique hardware serial number for the trusted execution environment of the edge gateway that performs this calculation, ensuring that the generated data fingerprint can be traced back to the corresponding acquisition terminal and edge device, and cannot be counterfeited across devices.

[0037] The edge gateway has a built-in lightweight anomaly detection model. After the model is fully trained in the cloud, it is trimmed, quantized and compressed, and then sent to the edge gateway for local operation, so that real-time inference can be completed without relying on cloud computing power.

[0038] The model takes weight, category images, and location data uploaded by the acquisition terminal as input and outputs the probability of data anomalies. The preset threshold can be customized according to the needs of the scenario. The default threshold is 80%. When the probability of anomalies exceeds the threshold, the edge gateway directly triggers data interception and audible and visual warnings locally. At the same time, the abnormal event record is synchronized to the cloud service layer. There is no need to wait for cloud processing instructions, which reduces the computing pressure on the cloud and enables real-time handling of abnormal data.

[0039] Edge gateways support offline operation even when the network is down. In the offline state, both the raw data collected and the processed structured data are stored in an encrypted and secure storage area bound to the Trusted Execution Environment (TEE), which cannot be tampered with locally. After the network is restored, the edge gateway signs the offline data with its own node private key and automatically synchronizes it to the cloud service layer, ensuring uninterrupted business operations, no data loss, and full traceability.

[0040] The cloud service layer and the edge computing layer are connected via encrypted communication over a wide area network. It has three core functional units built in: a full-link multi-dimensional cross-verification module, a data storage module, and a source tracing and early warning module. It is the core computing and control hub of this system.

[0041] The end-to-end multi-dimensional cross-verification module is the core functional unit of this system. Using a unique batch identifier as an index, it completes verification in two phases. The first phase is real-time node verification: after data entry is completed at each flow node, the module triggers a chain-like logical verification between that node and the previous node in real time, identifying single-node anomalies in advance. The second phase is end-to-end closed-loop verification: after the batch completes the entire production and feeding process at the waste utilization enterprise, the module performs a complete two-dimensional cross-verification, ultimately generating a final authenticity score for the batch data through the end-to-end data authenticity scoring model.

[0042] When performing upstream and downstream chain logic verification, the module uses a unique batch identifier as an index and performs relative deviation verification on the same batch and type of recycled resource data of adjacent nodes according to the physical flow sequence. The verification object is the outbound weight data of the previous node and the inbound weight data of the next node. Node data with relative deviation exceeding the preset threshold is marked as abnormal data.

[0043] The calculation formula used for verifying the relative deviation of data between adjacent nodes is as follows: In the formula, For the i-th node and the i-th node in the entire link The relative deviation of recycling data between adjacent nodes is as follows: the larger the value, the greater the difference between the two nodes and the higher the risk of anomaly; i is the node number of the entire recycling chain, which is in the order of physical flow: front-end placement node, recycling station node, transportation node, sorting center node, and waste utilization enterprise node. Each node corresponds to a fixed number to ensure that the verification order is completely consistent with the physical flow order. The cumulative outbound weight of the product category for the i-th node is the valid measurement data of that node after verification by the trusted execution environment at the edge. It is bound to a unique batch identifier and cannot be tampered with. For the first The cumulative inbound weight of the same category for each node corresponds strictly to the outbound data of the previous node for the same batch of recycled resources, ensuring the consistency of the verification objects.

[0044] The preset thresholds are fixed according to the node type. The threshold for front-end delivery nodes and recycling bin nodes is 3%, and the threshold for other adjacent nodes is 2%. Node data with relative deviations exceeding the corresponding thresholds are directly marked as abnormal data and included in the calculation dimension of the final authenticity score.

[0045] After completing the upstream and downstream chain logic verification, the module connects to third-party tamper-proof data sources to perform external cross-verification. The data sources connected include tax invoice systems, power consumption monitoring systems, and commodity price information systems. All data sources are connected through official compliant interfaces, and the data is tamper-proof and tamper-proof.

[0046] The verification rules to be executed must include at least three items: The first item is to match and verify the batch production input data of waste utilization enterprises with the power consumption data of production equipment in the same production cycle. The verification logic is that the raw material processing volume corresponding to the unit energy consumption of a single production line is fixed within a reasonable range. If the input data and energy consumption data exceed the reasonable matching range, they are marked as abnormal data to avoid falsification of input data for idle running without actual production.

[0047] The second item is to verify the consistency of the category, quantity, and amount of the corresponding batch of recycled transaction data with the corresponding tax invoice data. It verifies whether the transaction data and the invoice data filed with the tax authority are completely matched, and any mismatches are marked as abnormal data.

[0048] The third item is to verify the reasonableness of the corresponding batch recycling price and the national average market price of the same type of bulk commodities in the natural week of the transaction date. The national average market price of the same type of bulk commodities in the natural week of the transaction date is used as the benchmark. If the price exceeds the benchmark by 20% above or below, it is marked as abnormal data to identify fraudulent transactions with abnormal pricing.

[0049] After completing the two-dimensional cross-validation, the module generates the final authenticity score for this batch of data using a full-link data authenticity scoring model. The calculation formula used in the scoring model is as follows: In the formula, S is the final authenticity score of a single batch of recycled resource data, ranging from 0 to 100. The higher the score, the higher the authenticity of the data and the lower the risk of falsification. , , , These are the weight coefficients for the four core feature dimensions, with the sum of the four weight coefficients being 1. Each coefficient ranges from 0 to 1, and all weight coefficients are non-negative. The initial values ​​of all weight coefficients are obtained by fitting a pre-trained graph convolutional neural network model. The training set used for fitting consists of historical data from the entire recycling chain over the past three years and data from confirmed fraudulent recycling events. The weight coefficients can be dynamically adjusted based on batch type, the entity's historical compliance records, and risk level. For example, for high-value, high-risk counterfeiting categories such as scrap metal, the corresponding weight coefficients are... The weighting coefficient of external cross-validation matching degree will be automatically increased to strengthen the validation weight of external data; for entities that have long been compliant with regulations, the corresponding... The weighting coefficient for the reasonableness of data fluctuations will be appropriately lowered to reduce the probability of misjudgment caused by normal business fluctuations. C represents the consistency score of the end-to-end chain logic verification, ranging from 0 to 100. The score is negatively correlated with the number of abnormal nodes and the magnitude of deviations. Batches with no abnormal nodes receive a full score of 100 for this item; each abnormal node deducts a corresponding score based on the deviation ratio. M represents the matching degree score of third-party external data cross-verification, ranging from 0 to 100. The score is negatively correlated with the number of external verification anomalies; batches with no external verification anomalies receive a full score of 100 for this item; each anomaly deducts a corresponding score based on the matching degree deviation. R represents the reasonableness score of data fluctuations, ranging from 0 to 100. This corresponds to the degree of matching between the batch data and historical compliant data of the same subject, category, and time period. Data fluctuations within the historical normal range receive a full score of 100; fluctuations exceeding the normal range deduct a corresponding score based on the magnitude. B represents the compliance score for operational behavior, ranging from 0 to 100. It corresponds to the compliance results of the entire batch of operational nodes. A score of 100 is awarded for no abnormal operations, no out-of-bounds permissions, and no offline tampering records. For non-compliant operations, the corresponding score is deducted according to the severity.

[0050] The cloud service layer has a built-in sub-module for dynamically adaptively adjusting the scoring threshold. It takes the historical compliance record of the recycling entity, the risk level of counterfeit recycling categories, regional regulatory requirements, and market price fluctuations as input features. It outputs recommended thresholds through a lightweight gradient boosting tree model. The system automatically adjusts the preset threshold for the authenticity score of the corresponding batch based on the recommended threshold, while retaining the manual adjustment authority of regulatory departments to ensure that the thresholds are adapted to the regulatory needs of different scenarios.

[0051] The data storage module is built using a consortium blockchain architecture, with recycling operators, waste utilization companies, and regulatory departments all serving as peer nodes in the consortium blockchain. Each node is configured with independent and fixed hierarchical access permissions.

[0052] Recycling operators can only access the on-chain data they generate themselves, waste utilization enterprises can only access the on-chain data corresponding to their own batches entering the warehouse, and regulatory authorities can access all on-chain data within their jurisdiction. At the same time, they can access the on-chain data of corresponding batches across regions according to their regulatory authority, ensuring full-process supervision of batches transferred across regions.

[0053] The consortium blockchain adopts a practical Byzantine fault-tolerant consensus mechanism. Node data synchronization requires three-stage consensus confirmation to ensure the consistency and immutability of the data uploaded to the chain. At the same time, it adapts to the business needs of high-frequency data recycling and uploading to the chain, ensuring the efficiency of data uploading.

[0054] All data uploaded to the blockchain can only be written using the private key of the corresponding consortium blockchain node generated from the data. Once written, it cannot be tampered with, deleted, or altered, and the entire process is traceable. The data uploaded to the blockchain includes fingerprints of single batches of data, data from the entire verification process, authenticity scoring results, and early warning and handling records. The data is synchronized to the blockchain in real time after generation, ensuring full traceability.

[0055] The source tracing and early warning module communicates with the end-to-end multi-dimensional cross-verification module and the data storage module, respectively. It is used to perform end-to-end anomaly tracing and location for batch data with authenticity scores below a preset threshold, and push early warning information according to the anomaly level.

[0056] like Figure 3 As shown, the module is divided into three non-overlapping risk levels based on the authenticity score: >40 and ≤60 is general risk, >20 and ≤40 is medium risk, and ≤20 is major risk.

[0057] The tiered early warning push rules are fixed as follows: general risk warnings are only pushed to the application ports of the corresponding recycling operators; medium risk warnings are simultaneously pushed to the application ports of the corresponding recycling operators, waste utilization enterprises, and local regulatory departments; and major risk warnings are immediately pushed to the application ports of all relevant entities and regulatory departments. Push methods include in-app messages, SMS, and system pop-ups. Major risk warnings are also simultaneously pushed to the dedicated government affairs interface of regulatory departments.

[0058] When the module performs end-to-end anomaly tracing, it uses the unique batch identifier as an index to retrieve the original data, on-chain evidence records, and node operation records of the entire chain for that batch. It uses the data association graph to reverse locate the specific node, anomaly type, and deviation value of the abnormal data, and generates a tracing report containing abnormal node information, deviation details, and original data vouchers, which is then pushed to the corresponding entity and regulatory department.

[0059] The module tracks the entire process of handling all early warning events: For recycling operations corresponding to general risks, the responsible entity must upload rectification certificates and supporting materials within 3 working days from the date the early warning information is sent; for all relevant entities corresponding to medium and major risks, the responsible entity must provide feedback on the handling progress within 1 working day from the date the early warning information is sent. All working days mentioned are national statutory working days, excluding holidays.

[0060] For batches that have completed rectification, the module re-executes end-to-end chain-based logical verification and external cross-verification. Once verification is passed, the warning is lifted. If two consecutive verifications fail, the system automatically marks the corresponding entity as a high-risk entity. The verification threshold for all subsequent batches of that entity is automatically increased, with the increase matching the risk level, and the weighting coefficient is adjusted synchronously to strengthen fraud risk prevention and control. Entities that fail to complete rectification on time are automatically pushed to regulatory authorities for supervision, forming a complete closed-loop handling process.

[0061] The application layer communicates with the cloud service layer, including functional ports corresponding to recycling operators, waste utilization enterprises, and regulatory departments, which are used by the corresponding entities to complete data management, compliance verification, and early warning and disposal.

[0062] The recycling business entity's portal can complete equipment status management, full-link data viewing, automatic generation of business ledgers, viewing and handling of early warning information, and historical data tracing. The system can automatically generate business ledgers that meet regulatory requirements based on full-link data, eliminating the need for manual recording.

[0063] The waste utilization enterprise portal can complete the verification of incoming data, upload production data, generate compliance ledgers, and trace the entire chain of supplier recycling data. It can perform full-chain verification of batch data of incoming recycled resources and avoid compliance risks caused by receiving false recycled raw materials.

[0064] The regulatory authority's portal can provide a global visual display of recycling data within its jurisdiction, verify all batches of data, issue early warnings and supervise anomalies, conduct full-chain traceability and verification, and automatically generate regulatory reports. It can view the operational status and data authenticity of all recycling entities within its jurisdiction in real time, and automatically generate monthly and annual regulatory reports. All regulatory data comes from immutable data stored on the consortium blockchain and can be directly used as the basis for compliance verification.

[0065] The complete system operation steps for a single batch of recyclable resource recycling are as follows: Figure 2 As shown: The first step involves residents depositing recyclable resources into smart recycling bins at the front-end delivery nodes. The terminal collects the deposit data and synchronizes it to the edge computing module to complete local preprocessing and data fingerprint generation. The data is then synchronized in real time to the cloud service layer and the data storage module for on-chain storage.

[0066] The second step involves the recycling station collecting the corresponding recyclable resources, weighing and collecting data for warehousing, packaging them, generating a globally unique batch identification code, and writing it into a one-time anti-counterfeiting RFID cable tie to bind the physical item with the data. Outbound data is generated simultaneously, and all data is processed by the edge gateway and then uploaded to the blockchain for verification. After node data entry is complete, the cloud triggers a real-time chain-like logical verification between that node and the previous node, providing immediate alerts for any anomalies.

[0067] The third step involves the transport vehicle scanning the batch identification code to confirm its departure from the warehouse. During transportation, the intelligent vehicle terminal collects real-time data on trajectory, load, and cargo box status. This data is then processed by the vehicle edge gateway and uploaded to the cloud in real time, simultaneously stored on the blockchain for evidence. Once the node data is entered, the cloud triggers a chain-like logical verification between that node and the previous node in real time, identifying anomalies and issuing immediate warnings.

[0068] The fourth step involves the sorting center scanning the batch identification code to complete the warehousing confirmation, collecting data from the entire sorting process, processing it through the edge gateway, uploading it to the cloud, and simultaneously storing it on the blockchain. Once sorting is complete, outbound data is generated. After node data entry is complete, the cloud triggers a real-time chain-like logical verification between that node and the previous node, identifying anomalies and issuing immediate alerts.

[0069] Fifth, the waste utilization company scans the batch identification code to complete the warehousing confirmation, collects data on raw material warehousing, production input, and power consumption, processes the data through the edge gateway, and uploads it to the cloud for simultaneous on-chain storage. After the node data is entered, the cloud triggers a chain-like logical verification between the node and the previous node in real time, and issues immediate warnings for any anomalies.

[0070] The sixth step involves a cloud-based, multi-dimensional cross-verification module that uses batch identifiers as indexes to complete closed-loop chain-like logical verification and third-party external cross-verification. A final score is generated through an authenticity scoring model. Batches with scores above a preset threshold are marked as compliant, and a compliance verification report is generated and stored. Batches with scores below the preset threshold trigger a traceability and tiered early warning process, completing the closed-loop handling.

[0071] For different application scenarios, the system can flexibly adjust the configuration type of terminal collection devices, the rule threshold of cross-validation, and the allocation ratio of weight coefficients. Without reconstructing the system architecture, it can adapt to the business needs of different scenarios such as urban residential community recycling of recyclable resources, township and village-level recycling networks, industrial park recycling of industrial recyclable resources, large supermarket recyclable material collection, full-chain management of recycling industrial parks, and bulky waste recycling and disposal.

[0072] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention in any way. Although the present invention has been disclosed above with reference to preferred embodiments, it is not intended to limit the present invention. Any person skilled in the art can make some modifications or alterations to the above-disclosed technical content to create equivalent embodiments without departing from the scope of the present invention. Any simple modifications, equivalent changes and alterations made to the above embodiments based on the technical essence of the present invention without departing from the scope of the present invention shall still fall within the scope of the present invention.

Claims

1. A data management system for recycling renewable resources based on the Internet of Things, characterized in that, The system comprises a terminal acquisition layer, an edge computing layer, a cloud service layer, and an application layer; The terminal acquisition layer covers the front-end placement, recycling stations, transportation, sorting, and waste utilization enterprise nodes of the entire recycling chain. Each node is configured with a corresponding acquisition terminal to collect native business data of the entire recycling process and generate a globally unique batch identifier code for each batch of recycled resources. The edge computing layer is communicatively connected to each acquisition terminal, including edge gateways deployed on each node, used to perform local cleaning and structuring processing on native business data, and generate an immutable data fingerprint that is uniquely bound to the batch identifier code; The cloud service layer communicates with the edge computing layer and has a built-in end-to-end multi-dimensional cross-verification module, data storage module, and source tracing and early warning module. The application layer communicates with the cloud service layer and includes functional ports for corresponding recycling operators, waste utilization enterprises, and regulatory departments, used to complete data management, compliance verification, and early warning and handling.

2. The Internet of Things-based data management system for recycling renewable resources according to claim 1, characterized in that, The full-link multi-dimensional cross-verification module uses the batch identifier code as an index to first perform chain-like logical verification of the upstream and downstream of the entire link, and then connects to a third-party tamper-proof data source to perform external cross-verification. It generates the authenticity score of the batch data through the full-link data authenticity scoring model. The data storage module is used to store fingerprints, verification process data, and authenticity scoring results of batch data throughout their entire lifecycle. The source tracing and early warning module is used to perform full-chain source tracing for batches whose authenticity score is lower than a preset threshold, and push early warning information according to the level of abnormality.

3. The Internet of Things-based data management system for recycling renewable resources according to claim 1, characterized in that, The edge gateway has a built-in Trusted Execution Environment (TEE) that is physically isolated from the device operating system. The entire process of cleaning, structuring, and generating data fingerprints of native business data is completed within the TEE. The edge gateway has a built-in lightweight anomaly detection model, which is used to perform real-time anomaly detection on the collected weight, category, and location data locally, and to perform local interception and early warning for abnormal data that exceeds a preset threshold. The edge gateway supports offline operation when the network is down. It completes local data processing and encrypted storage when the network is down, and automatically synchronizes the data to the cloud service layer after the network is restored.

4. The Internet of Things-based data management system for recycling renewable resources according to claim 1, characterized in that, The recycling station node of the terminal collection layer is equipped with an RFID tag writing device and a disposable anti-counterfeiting RFID cable tie. The unique batch identification code is written into the passive RFID tag built into the cable tie. The cable tie is used to package the corresponding batch of recycled resources. After the cable tie is cut, the built-in RFID tag is physically damaged, cannot be read and cannot be reused. The transportation nodes, sorting nodes, and waste utilization enterprise nodes are all equipped with RFID reading and writing devices. Each node must scan the corresponding batch cable ties that have not expired through the RFID reading and writing device in order to complete the node confirmation and entry of the batch data.

5. The Internet of Things-based data management system for recycling renewable resources according to claim 1, characterized in that, When the full-link multi-dimensional cross-verification module performs upstream and downstream chain logic verification, it uses the batch identifier code as an index and performs relative deviation verification on the same batch and same type of recycled resource data of adjacent nodes according to the physical flow sequence. The verification object is the outbound weight data of the previous node and the inbound weight data of the next node. Node data with a relative deviation exceeding the preset threshold is marked as abnormal data. The preset threshold is set in a fixed manner according to the node type. The threshold for the front-end delivery node and the recycling station node is 3%, and the threshold for the other adjacent nodes is 2%.

6. The Internet of Things-based data management system for recycling renewable resources according to claim 1, characterized in that, The end-to-end multi-dimensional cross-verification module connects to third-party tamper-proof data sources, including tax invoice systems, power consumption monitoring systems, and commodity price information systems. When performing external cross-validation, at least three validation rules must be included: First, the matching and verification of the batch production input data of waste utilization enterprises with the power consumption data of production equipment in the same production cycle; Secondly, verify the consistency of category, quantity, and amount between the corresponding batch of recycled transaction data and the corresponding tax invoice data; Third, verify the reasonableness of the corresponding batch recycling price and the national average market price of the same type of bulk commodity in the natural week in which the transaction occurred.

7. A data management system for recycling renewable resources based on the Internet of Things according to claim 1, characterized in that, The full-link data authenticity scoring model takes the consistency results of the full-link chain logic verification, the matching degree results of external cross-verification, the rationality of historical data fluctuations in the same dimension, and the compliance of the operation behavior of all process nodes as core feature dimensions, and outputs an authenticity score with a value range of 0 to 100. The cloud service layer has a built-in sub-module for dynamically adaptively adjusting the scoring threshold. It uses the historical compliance record of the recycling entity, the risk level of counterfeiting of the recycling category, regional regulatory requirements, and market price fluctuations as input features to dynamically adjust the preset threshold for the authenticity score of the corresponding batch.

8. The Internet of Things-based data management system for recycling renewable resources according to claim 1, characterized in that, The data storage module adopts a consortium blockchain architecture, where recycling operators, waste utilization companies, and regulatory departments are all peer nodes in the consortium blockchain, and each node is configured with independent and fixed hierarchical access permissions: Recycling operators can only access the on-chain data they generate; waste utilization enterprises can only access the on-chain data corresponding to their own batches; and regulatory authorities can access all on-chain data within their jurisdiction. All on-chain data can only be written using the private key of the corresponding consortium blockchain node generated from the data. Once written, the data cannot be tampered with, cannot be deleted, and is traceable throughout the entire process.

9. A data management system for recycling renewable resources based on the Internet of Things according to claim 1, characterized in that, The source tracing and early warning module is divided into three non-overlapping risk levels based on the authenticity score: >40 and ≤60 is general risk, >20 and ≤40 is medium risk, and ≤20 is major risk. The tiered early warning push rules are fixed as follows: general risk early warning information is only pushed to the corresponding recycling business entity; medium risk early warning information is simultaneously pushed to the corresponding recycling business entity, waste utilization enterprise and local regulatory department; and major risk early warning information is immediately pushed to all relevant entities and regulatory departments.

10. A data management system for recycling renewable resources based on the Internet of Things according to claim 9, characterized in that, The source tracing and early warning module tracks the entire process of handling all early warning events: For recycling operators facing general risks, they must upload rectification certificates and supporting materials within 3 working days from the date the early warning information is sent. All relevant entities involved in medium-risk and major-risk situations must provide feedback on the progress of their actions within one working day from the date the early warning information is sent. The traceability and early warning module re-executes the full-chain upstream and downstream logical verification and external cross-verification for batches that have completed rectification. The early warning is lifted after the verification is passed. For those that have not completed rectification on time, the module is automatically pushed to the regulatory authorities for supervision.

Images