Equipment operation and maintenance-oriented Internet of Things data uplink system and method

Abstract

The invention discloses an equipment operation and maintenance-oriented Internet of Things data uplink system and method, and belongs to the technical field of Internet of Things and block chains, and the system comprises an edge calculation module, an intelligent data screening module, a block chain evidence storage module, an intelligent contract module, a service integration module, and a data verification and auditing module. According to the method, key abnormal data are intelligently screened to be linked through a two-stage fusion anomaly detection algorithm, and the load on the chain is greatly reduced; a mixed framework of storing original data under a chain and storing hash and metadata on the chain is adopted, so that data integrity and verifiability are ensured; and automatic linkage of fault detection, maintenance work order creation, personnel assignment and spare part scheduling is realized through an intelligent contract. According to the method, the technical problems of high on-chain load, low data management efficiency and disjunction with a business process of an existing Internet of Things block chain scheme are solved, and high efficiency, automation and credibility of an equipment operation and maintenance process are realized.

Description

Technical Field

[0001] This invention relates to the fields of Internet of Things (IoT) and blockchain technology, and in particular to an IoT data on-chain system and method for equipment operation and maintenance. Background Technology

[0002] Blockchain technology, as a distributed ledger technology, integrates core technologies such as distributed data storage, peer-to-peer transmission, consensus mechanisms, encryption algorithms, and smart contracts to build a decentralized, immutable, and traceable foundation of trust. In the field of Industrial Internet of Things (IIoT), blockchain automatically executes equipment operation and maintenance through smart contracts, and stores data such as equipment identification, operating parameters, and maintenance records on the blockchain to enhance data security and transparency.

[0003] The development of blockchain technology has roughly gone through three stages: the first stage focused on digital currencies to achieve decentralized payments; the second stage expanded to the financial field, relying on smart contracts to achieve automated transactions; and the third stage has gradually penetrated into diversified scenarios such as public services, social governance, and industrial operations and maintenance. In the field of industrial operations and maintenance, early solutions mostly adopted public or private blockchains, but in recent years, consortium blockchains have become the mainstream form due to their balance between efficiency and controllability.

[0004] Despite the progress made by blockchain in industrial operations and maintenance, existing solutions still face many problems. For example, the invention patent with publication number CN114726599A provides a method for information processing of distributed energy stations that integrates blockchain and network. It achieves trusted operations and maintenance through data storage and consensus verification via blockchain nodes. However, this method tends to store all IoT device parameters and other information on the blockchain, relying on accounting nodes to compete for accounting rights. Its consensus mechanism has a slow processing speed, resulting in high on-chain storage and computing loads, making it difficult to meet the real-time processing needs of high-frequency data in industrial sites. Another example is the invention patent with publication number CN115086011A, which provides a blockchain-based network security operations and maintenance and defense method. This method divides network operations and maintenance data into critical and non-critical data and employs different strategies for data storage and defense. While blockchain-based data storage is used, this method is inefficient in managing massive amounts of raw data. Its data classification relies on static importance analysis and lacks a dynamic screening mechanism for data value. Furthermore, there is a lack of effective collaboration and integrity verification between on-chain and off-chain data, failing to achieve refined data lifecycle management. For example, the invention patent with publication number CN115510733A provides an industrial control system operation and maintenance method based on blockchain and artificial intelligence technologies. It integrates AI for fault prediction and uses smart contracts to automatically execute some operation and maintenance operations. However, the blockchain function in this method is mostly limited to data storage and process recording, failing to be deeply integrated with core operation and maintenance business processes such as fault diagnosis, work order creation, and spare parts scheduling. The degree of automation is limited, and there is a problem of disconnection from actual business scenarios. Summary of the Invention

[0005] To overcome the shortcomings of existing technologies, this invention provides an IoT data uplink system and method for equipment operation and maintenance.

[0006] To achieve the above objectives, the present invention adopts the following technical solution: In a first aspect, the present invention provides an IoT data uploading method for equipment operation and maintenance, comprising the following steps: S101: In the data acquisition and preprocessing stage, equipment operation data is acquired through a sensor network deployed at key monitoring points of the equipment, and edge computing is performed on the equipment operation data to generate initial data. The edge computing includes preliminary filtering and formatting preprocessing of the raw time-series data at the edge. S102: In the intelligent data filtering stage, a two-level fusion anomaly detection mechanism, including dynamic threshold screening and deep machine learning analysis, is used to filter out abnormal data from the generated initial data. If no anomalies are detected, the data is stored off-chain and not further processed on-chain.

[0007] S103: On-chain-off-chain collaborative storage phase. Initial data containing anomalous data is completely stored off-chain. Specifically, based on access frequency and value density, the initial data containing anomalous data is completely stored in different off-chain storage media. Simultaneously, the hash value of the anomalous data is calculated, generating a notarization data package, which is then sent to the blockchain network for on-chain storage. Specifically, the cryptographic hash value of the anomalous data is calculated, and this hash value is packaged with key metadata such as timestamps, device identifiers, and anomaly type tags to form a notarization data package containing key digest information, which is then sent to the blockchain network for on-chain storage. Consensus nodes verify and store this digest information, completing the trusted notarization of the data. S104: Smart Contract and Business Integration Phase. In response to the event of the evidence storage data packet being uploaded to the blockchain, a predefined smart contract is automatically triggered. The smart contract executes its internal operating logic and automatically performs operations such as maintenance work order creation, technician assignment, and spare parts scheduling through an oracle integrated with the off-chain business system. S105: Data Verification and Auditing Phase. A multi-level verification mechanism is employed to verify the data uploaded to the blockchain and to audit and trace the entire chain of operations involving the uploaded data. This step provides complete data credibility verification and audit tracing functions, ensuring transparency and trustworthiness throughout the entire chain from data collection to business execution.

[0008] As a preferred embodiment of the present invention, the two-level fusion anomaly detection logic executed in the intelligent data filtering stage specifically includes: Level 1: Dynamic Threshold Screening. Based on historical equipment operating data, dynamic threshold models are established for various monitoring indicators under different operating conditions. These thresholds can be dynamically adjusted using a linear regression model based on factors such as equipment load and ambient temperature. If multiple real-time data points exceed the current dynamic threshold range, they are marked as "suspicious."

[0009] Level 2: Deep Machine Learning Analysis. For "suspicious" data, feature vectors containing time-domain and frequency-domain features are extracted and input into a pre-trained lightweight isolated forest model for analysis. This model calculates anomaly scores based on the path length of data points in the isolated trees. When the score exceeds a preset confidence level, the data is ultimately determined to be "abnormal" and triggered to be uploaded to the blockchain.

[0010] As a preferred embodiment of the present invention, the internal execution logic of the smart contract in the business integration phase is specified as follows: In response to an event where a data packet containing multidimensional physical event data and event results is uploaded to the blockchain, predefined modular smart contracts on the blockchain are automatically triggered. These smart contracts include work order management contracts, resource optimization contracts, and supply chain collaboration contracts.

[0011] The work order management contract calls its declared event result judgment logic to perform conflict judgment on the event data in the evidence data packet. If there is a conflict, the final result of the physical event is determined based on the preset consensus protocol. Based on this, the operation and maintenance work order is generated or updated and its status is tracked. The work order status is managed using a state machine model, including the creation, allocation, execution, waiting for spare parts, completion and cancellation status. The resource optimization contract uses the operation and maintenance work order information generated by the work order management contract to execute a multi-factor evaluation algorithm to select the best engineer from the maintenance technical service personnel database for work order assignment. The evaluation algorithm comprehensively considers skill matching, workload balancing, geographical proximity and engineer rating factors. The supply chain collaboration contract automatically generates spare parts demand orders based on the abnormal events or fault types identified by the work order management contract, and transmits the demand to the external inventory management system through the oracle service, while tracking the outbound and logistics status of spare parts. Oracle services act as a bridge between the blockchain and external business systems, converting blockchain events into API calls that can be recognized by the business systems. During the data conversion process, they perform format validation and business rule checks, and adopt the Saga pattern to manage distributed transactions across multiple business systems to ensure eventual consistency of the system.

[0012] As a preferred embodiment of the present invention, the multi-level verification mechanism executed during the data verification and auditing phase specifically includes: Digital signature verification involves using the sender's public key to decrypt the digital signature attached to the received data to obtain a first hash value (H1), independently calculating a second hash value (H2) of the received data, and comparing H1 and H2 to verify the data source and integrity. Merkel proof involves recursively calculating the root hash value based on the data to be verified, the Merkel path corresponding to the data to be verified, and the trusted Merkel root stored on the chain. The calculated root hash value is then matched with the trusted Merkel root to verify that the data to be verified belongs to a specific data set. Hash value comparison involves recalculating the hash value of the abnormal data contained in the initial data corresponding to the abnormal data on the chain, and comparing the recalculated hash value with the initial hash value stored on the chain. If the hash values ​​match, it is determined that the data has not been tampered with.

[0013] Secondly, this invention provides an IoT data uplink system for device operation and maintenance that implements the above-mentioned method. The system adopts a microservice-based distributed architecture and specifically includes: Edge computing module: As the system data entry point, it includes industrial-grade edge gateway hardware and a layered software architecture. The software architecture includes a device driver layer that supports multiple industrial protocols, a data processing layer that integrates a stream processing engine, and an application service layer that provides data interfaces, which is responsible for real-time data acquisition, preprocessing, and caching.

[0014] Intelligent data filtering module: Through a two-level fusion anomaly detection algorithm that combines dynamic threshold screening with machine learning deep analysis, the pre-processed device data is intelligently filtered to ensure that only exact abnormal data is submitted to the blockchain for notarization, thereby effectively improving data value density and reducing network load.

[0015] Blockchain Evidence Storage Module: Built on a permissioned consortium blockchain network, it adopts a practical Byzantine fault-tolerant consensus mechanism, provides a distributed ledger, state database and integration with a distributed file system, and builds a security mechanism based on public key infrastructure and attribute-based access control model.

[0016] Smart Contract Module: Provides a complete contract development toolchain, security auditing tools, and a high-performance execution environment based on WebAssembly, supporting the deployment, upgrade, and full lifecycle management of contract modules such as work order management, resource optimization, and supply chain collaboration.

[0017] Business integration module: It realizes protocol conversion and security authentication through API gateway, realizes data format conversion and business rule processing through data conversion engine and rule engine, and ensures eventual consistency of cross-system distributed transactions based on Saga pattern transaction manager.

[0018] Data Verification and Auditing Module: Provides data credibility verification and end-to-end audit trail capabilities. Its integrity verification system uses a three-level verification mechanism consisting of digital signatures, Merkel proofs, and hash comparisons to ensure that the data has not been tampered with since its generation. Its audit trail system records the entire operation log from data collection to contract execution, supporting multi-dimensional query analysis and automated compliance report generation.

[0019] Compared with the prior art, the beneficial effects of the present invention are as follows: This invention achieves a fundamental shift from "all on the chain" to "abnormal on the chain" through an innovative two-level fusion detection model for intelligent screening. It filters out more than 95% of normal data from the source, greatly reducing the storage and computing pressure on the blockchain and significantly reducing operating costs. A hybrid storage architecture with clearly defined responsibilities was constructed. The off-chain system is responsible for storing and processing massive amounts of raw data with high performance, while the blockchain acts as a "trust anchor," ensuring that any tampering with off-chain data can be detected quickly and irrefutably through the cryptographic hash of the stored data, thus perfectly balancing performance and trustworthiness. By designing smart contracts with concrete, executable logic, trusted data flows are directly transformed into trusted business flows. The system can automatically initiate and record the entire maintenance response process the moment a fault is detected, without human intervention, forming a closed-loop automation from "perception" to "execution," significantly improving operational efficiency and process transparency. Attached Figure Description

[0020] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with the invention and, together with the description, serve to explain the principles of the invention.

[0021] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, for those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0022] Figure 1 This is a flowchart of an IoT data uplink method for device operation and maintenance provided in an embodiment of the present invention.

[0023] Figure 2 This is a flowchart of the data acquisition and preprocessing steps provided in the embodiments of the present invention.

[0024] Figure 3 This is a flowchart of the intelligent data filtering steps provided in an embodiment of the present invention.

[0025] Figure 4This is a flowchart of the on-chain-off-chain collaborative storage steps provided in an embodiment of the present invention.

[0026] Figure 5 This is a flowchart of the smart contract and business integration steps provided in the embodiments of the present invention.

[0027] Figure 6 This is a flowchart of the data verification and auditing steps provided in the embodiments of the present invention.

[0028] Figure 7 This is a structural block diagram of an IoT data uplink system for device operation and maintenance provided in an embodiment of the present invention. Detailed Implementation

[0029] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of them. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present invention.

[0030] It should be noted that all directional indications (such as up, down, left, right, front, back, etc.) in the embodiments of the present invention are only used to explain the relative positional relationship and movement of each component in a certain specific posture (as shown in the figure). If the specific posture changes, the directional indication will also change accordingly.

[0031] Furthermore, the use of terms such as "first" and "second" in this invention is for descriptive purposes only and should not be construed as indicating or implying their relative importance or implicitly specifying the number of technical features indicated. Therefore, features defined with "first" and "second" may explicitly or implicitly include at least one of those features. Additionally, the technical solutions of the various embodiments can be combined with each other, but only on the basis of being achievable by those skilled in the art. When the combination of technical solutions is contradictory or impossible to implement, such a combination of technical solutions should be considered non-existent and not within the scope of protection claimed by this invention.

[0032] Example 1 like Figure 1 As shown, the present invention provides an IoT data uploading method for device operation and maintenance, comprising the following steps: S101: In the data acquisition and preprocessing stage, equipment operation data is acquired through a sensor network deployed at key monitoring locations of the equipment, and edge computing is performed on the equipment operation data to generate initial data. The edge computing includes preliminary filtering and formatting preprocessing of the raw time-series data at the edge.

[0033] This phase employs a systematic data acquisition and preprocessing process to ensure the acquisition of high-quality equipment operation data, providing a reliable foundation for subsequent analysis. For example... Figure 2 As shown, the data acquisition and preprocessing stage specifically includes the following processes: 1) Data collection: Data acquisition employs a sensor network architecture to obtain equipment operation data. At the physical deployment level, the system installs various types of sensors on key monitoring components of the equipment, including vibration acceleration sensors, temperature sensors, and electrical parameter sensors. Vibration sensors are primarily installed on critical mechanical components such as bearing housings and gearbox housings to monitor the mechanical vibration state of the equipment; temperature sensors are installed on heat-prone areas such as motor windings and bearing outer rings; and electrical parameter sensors monitor the electrical operating status of the equipment through current and voltage transformers.

[0034] During data acquisition, the system employs synchronous acquisition technology to ensure data time consistency. A unified time reference is provided for all acquisition nodes via a GPS / BeiDou clock source, achieving precise synchronization of data from each sensor. For vibration data, the system acquires data at a sampling frequency of 5kHz-10kHz to ensure the capture of high-frequency vibration characteristics of the equipment; for temperature data, a sampling frequency of 1Hz is used; and for electrical parameters, a sampling frequency of 1kHz is used.

[0035] The system employs an anti-aliasing filter before analog-to-digital conversion to prevent high-frequency noise from aliasing into low-frequency data. The anti-aliasing filter is a crucial low-pass filter whose core function is to prevent aliasing by limiting the bandwidth of the analog data before digitization. This filter attenuates or filters out high-frequency components above the Nyquist frequency (i.e., half the sampling frequency) in the data, ensuring that the data spectrum processed by the analog-to-digital converter does not produce distorted spurious frequencies due to high-frequency components folding into the baseband, thus guaranteeing the authenticity and accuracy of the data. The transfer function of the anti-aliasing filter is: Where H(f) is the gain at frequency f, f is the frequency of the current data, and f c Here, n is the cutoff frequency, and n is the filter order. Simultaneously, based on the amplitude range of the sensor output data, the system automatically adjusts the amplification factor of the programmable gain amplifier to ensure full utilization of the data's dynamic range.

[0036] 2) Edge preprocessing: Edge preprocessing employs a multi-stage pipeline architecture for data preprocessing. First, data quality control is performed. The system uses statistical principles to detect outliers in the collected data, marking data points that exceed the normal range. When data is missing, the system employs different imputation strategies depending on the type of missing data: For occasional missing data, cubic spline interpolation is used. This method first divides the interpolation interval based on known data points around the missing location, and independently constructs a cubic polynomial function for each sub-interval. Then, by forcing all polynomials to satisfy the conditions of continuity of function values, first derivatives, and second derivatives at the nodes within the interval, and combining this with specific boundary constraints to solve a system of linear equations, all interpolation function parameters are determined. Finally, using this continuous and smooth curve generated through global data constraints, high-precision estimation of missing values ​​within data gaps can be achieved. Its advantage lies in effectively preserving the local variation trend and overall morphological characteristics of the original data, making it particularly suitable for completing sporadic data in continuously changing scenarios such as sensor monitoring and physical quantity measurement. For continuous missing data, a data quality alarm mechanism is activated. Outlier detection uses the 3σ criterion. in, The mean of the data. The standard deviation is denoted as .

[0037] In the data processing stage, the system performs combined time-domain and frequency-domain processing. Time-domain processing includes trend removal and outlier removal. Outlier removal uses a median filter to remove impulse interference. The median filter is a nonlinear digital data processing technique based on ordination statistics theory. Its core principle is to replace the value of a point in the data with the median of all data points in its neighborhood window. This method uses a sliding template of odd size (such as 3×3 or 5×5) to traverse the data, sorts the values ​​in each window, and takes the median as the output, thus effectively filtering out impulse noise while significantly preserving data edge features and avoiding the edge blurring problem common in linear filtering methods. Trend removal uses least squares fitting to remove linear or polynomial trend terms from the data. Least squares fitting is a mathematical optimization technique whose core is to find the best function match for a set of data points by minimizing the sum of squared errors to achieve parameter estimation or curve fitting. This method is based on a selected combination of linearly independent functions, using the criterion of minimizing the sum of squared differences between the observed values ​​of all data points and the estimated values ​​of the fitted function to solve an unconstrained optimization problem. By establishing a system of normal equations (canonical equations) and solving for the undetermined parameters, the objective function is to minimize the overall deviation of the fitted curve from the measured data points within a given function class. Where yi is the actual observed value of the i-th data point, a is the translation parameter, b is the scale parameter, and xi is the feature value of the i-th data point.

[0038] Frequency domain processing mainly includes order analysis and spectrum analysis. Order analysis is based on equal-angle resampling of rotational speed data to eliminate the influence of rotational speed fluctuations on spectrum analysis. Spectrum analysis obtains frequency characteristics by performing fast Fourier transform on vibration data.

[0039] To extract richer data features, the system also performs time-frequency analysis, employing methods such as Hilbert-Huang transform and wavelet transform. The Hilbert-Huang transform obtains the intrinsic mode functions of the data through empirical mode decomposition; the wavelet transform uses the Db8 wavelet basis for multi-scale decomposition to extract data features from different frequency bands, better reflecting the time-varying characteristics of the data. The mathematical expression for the wavelet transform is: in, wavelet coefficients Let be the wavelet basis function, a be the scaling parameter, and b be the translation parameter.

[0040] Data normalization and compression are the final steps in preprocessing. The system standardizes different types of monitoring data to eliminate the influence of dimensions and make data on different physical quantities comparable. For data storage, the system adopts a hierarchical compression strategy: lossy compression based on wavelet transform is used for large data volumes such as vibration data to achieve a high compression ratio while ensuring the quality of reconstructed data; lossless compression is used for key feature parameters to ensure data accuracy.

[0041] S102: In the intelligent data filtering phase, a two-level fusion anomaly detection mechanism, including dynamic threshold screening and deep machine learning analysis, is used to filter out anomalous data from the generated initial data. If no anomalies are detected, the data is stored off-chain and not subsequently added to the blockchain. Figure 3 As shown, the intelligent data filtering stage is specifically as follows: Level 1 Detection: Dynamic Threshold Screening The dynamic threshold detection unit employs an adaptive threshold algorithm based on equipment operating status. This algorithm does not use a fixed threshold range, but rather dynamically adjusts the anomaly judgment criteria based on the real-time operating conditions of the equipment using a linear regression model. The system establishes a dynamic threshold model for each monitored parameter, which comprehensively considers multiple factors such as equipment load rate, environmental parameters, and cumulative operating time.

[0042] During threshold calculation, the system first establishes a baseline threshold based on historical equipment operating data, and then dynamically adjusts it according to the current operating conditions. For example, for temperature parameters, the threshold is related to ambient temperature and cumulative operating time; in high-temperature environments or under equipment aging conditions, the system will correspondingly lower the temperature threshold. For vibration intensity parameters, when the equipment operates under high load conditions, the system will appropriately widen the vibration intensity threshold range; its dynamic threshold is a function of the load rate: in, For dynamic threshold is the baseline threshold, k is the adjustment coefficient, and L is the load rate.

[0043] Anomaly detection employs a multi-feature joint evaluation mechanism. The system not only considers the exceeding of limits for a single parameter but also comprehensively analyzes the changing trends of multiple related parameters. When a parameter exceeds the dynamic threshold range, the system further evaluates the status of other related parameters. Only when multiple parameters exhibit abnormal characteristics is the data marked as a suspicious anomaly and proceed to the second level of detection.

[0044] Level 2 Detection: Deep Machine Learning Analysis The deep analysis unit employs a combination of feature engineering and machine learning for anomaly detection. Feature extraction is the foundation of deep analysis; the system extracts rich feature parameters from three dimensions: time domain, frequency domain, and time-frequency domain.

[0045] Temporal feature extraction includes two categories: dimensional indicators and dimensionless indicators. Dimensional indicators mainly reflect the absolute intensity of data, including root mean square value and peak value; dimensionless indicators reflect the shape characteristics of data, including kurtosis, skewness, waveform indicators, and impulse indicators. These indicators describe the statistical characteristics of data from different perspectives and have different sensitivities to different types of faults.

[0046] Frequency domain feature extraction is achieved through spectral analysis of the data. The system first performs a Fast Fourier Transform on the data to obtain the spectrum, and then calculates the energy distribution characteristics of each frequency band. In particular, the system calculates the theoretical fault characteristic frequencies based on the structural parameters of the equipment and extracts the energy amplitudes near these frequencies as an important basis for fault diagnosis.

[0047] Time-frequency domain feature extraction employs wavelet packet decomposition. The system performs multi-scale wavelet packet decomposition on the data, calculates the energy proportion of each decomposition node, and further calculates feature parameters such as wavelet energy entropy. These features can simultaneously reflect the time-domain and frequency-domain characteristics of the data, resulting in better anomaly detection for non-stationary data.

[0048] Based on feature extraction, the system employs Isolation Forest for anomaly detection. Isolation Forest is a tree-based unsupervised anomaly detection algorithm. Its core idea is to quickly identify outliers that are significantly different from other data by randomly partitioning the feature space. This algorithm is based on the assumption that "abnormal data is scarce and variable." It constructs multiple isolated trees by recursively and randomly selecting features and split points, forming a forest structure. In this process, outliers, due to their sparse distribution and distance from dense data areas, are typically isolated with shorter average path lengths. Each base detector evaluates the degree of anomaly from different perspectives. The system obtains the final anomaly score by weighted fusion of the outputs of each base detector. When the score is higher than a preset confidence level, the data is considered anomalous. The anomaly scoring function of the Isolation Forest algorithm is: in, Let E(h(x)) be the anomaly score for sample x, E(h(x)) be the path length of sample x in the isolated tree, and c(n) be the average path length.

[0049] Model updates and maintenance are crucial for ensuring consistently stable detection results. The system employs an incremental learning mechanism, enabling the model to adaptively update as device status changes. Simultaneously, the system also establishes a concept drift detection mechanism to monitor model performance trends and promptly trigger model retraining when a significant change in data distribution is detected.

[0050] S103: On-chain-off-chain collaborative storage phase. Initial data containing anomalous data is completely stored off-chain. Specifically, based on its access frequency and value density, the initial data containing anomalous data is completely stored in different off-chain storage media. Simultaneously, the hash value of the anomalous data is calculated, a notarization data packet is generated, and the notarization data packet is sent to the blockchain network for on-chain storage. Consensus nodes verify and store the digest information, completing the trusted notarization of the data. Figure 4 As shown, the on-chain-off-chain collaborative storage stage is specifically as follows: 1) Off-chain storage management system: Off-chain storage employs a distributed architecture to ensure high data availability and scalability. The system uses a consistent hashing algorithm to distribute data across multiple storage nodes, with each data block having multiple replicas, achieving data redundancy backup. The system also establishes a dynamic load balancing mechanism to automatically adjust data distribution based on the load of each node.

[0051] Data lifecycle management is a crucial function of off-chain storage. The system categorizes data into three levels—hot data, warm data, and cold data—based on access frequency and value density. Hot data refers to recently generated data that requires frequent access and is stored on high-performance solid-state drives (SSDs). Warm data refers to historical data that still requires periodic access and is stored on hard disk drives (HDDs). Cold data refers to rarely accessed archived data and is migrated to lower-cost storage media.

[0052] 2) On-chain evidence storage mechanism: The core idea of ​​on-chain evidence storage is to establish trusted proof of off-chain data through cryptographic methods. The system uses the SHA-256 hash function to calculate the digital fingerprint of off-chain data and stores the hash value along with key metadata on the blockchain. The system constructs a structured evidence storage data package, including version information, device identifier, timestamp, anomaly analysis results, data reference information, etc.

[0053] To improve the efficiency of evidence storage and the flexibility of data organization, the system introduces a Merkle tree structure. A Merkle tree is a binary tree data structure built using cryptographic hash functions. Its leaf nodes are the hash values ​​of the underlying data blocks, and each non-leaf node is composed of the combined hash of its child node hash values. This process is recursively performed until a unique root hash value, called the Merkle root, is generated. Merkle proofs are a mechanism based on Merkle trees used to verify the existence and integrity of a specific data block. This proof provides the hash values ​​of all sibling nodes required on the path from the leaf node to the root hash of the target data block. Verification only requires recalculating the path hash and checking if the final result matches the known Merkle root to efficiently confirm data ownership without exposing or verifying the entire dataset. Merkle trees aggregate the hash values ​​of multiple data blocks into a single root hash through hierarchical hash calculations. This structure has two important characteristics: first, any modification to the underlying data will lead to a change in the root hash; second, Merkle proofs can quickly verify whether a data block belongs to the tree.

[0054] The blockchain transaction process includes three main stages: transaction construction, signature verification, and consensus confirmation. In the transaction construction stage, the system serializes the evidence data and calculates the transaction hash. In the signature verification stage, the device's private key is used to digitally sign the transaction, ensuring the trustworthiness of the transaction's origin. In the consensus confirmation stage, the transaction is broadcast through a P2P network, verified by a consensus algorithm, and then packaged into a block for permanent evidence storage.

[0055] S104: Smart Contract and Business Integration Phase. In response to the event of the evidence storage data packet being uploaded to the blockchain, a predefined smart contract is automatically triggered. The smart contract executes its internal operating logic, automatically performing operations such as maintenance work order creation, technician assignment, and spare parts scheduling through an oracle integrated with the off-chain business system. Figure 5As shown, the smart contract and business integration phase is as follows: 1) Smart contract architecture design: The smart contract adopts a modular design, mainly including three core modules: work order management contract, resource optimization contract, and supply chain collaboration contract.

[0056] The work order management contract is responsible for creating and tracking the status of abnormal work orders. When new abnormal data is recorded on the blockchain, the work order management contract is automatically triggered, calculating the work order priority based on the abnormality type and confidence level, and generating structured work order information. The work order status is managed using a state machine model, including states such as creation, allocation, execution, waiting for spare parts, completion, and cancellation, ensuring complete tracking of the work order lifecycle.

[0057] Resource optimization contracts enable intelligent allocation of maintenance resources. The contract maintains a database of technical service personnel, recording each engineer's skills, expertise, geographical location, workload, and other information. When a work order needs to be assigned, the contract executes a multi-factor evaluation algorithm, comprehensively considering factors such as skill matching, workload balancing, geographical proximity, and engineer ratings to select the optimal engineer for assignment. The resource allocation algorithm employs multi-objective optimization: in, These are the weighting coefficients. For the first An optimization objective function.

[0058] Supply chain collaboration contracts are responsible for the automatic triggering and tracking of spare parts requirements. The contracts have a built-in mapping relationship between fault types and spare parts. When a specific type of fault is detected, a spare parts request form is automatically generated and transmitted to the inventory management system via an oracle service. The contracts also track the outbound and logistical status of spare parts to ensure timely availability of repair resources.

[0059] 2) Oracle service architecture: Oracle services act as a bridge between blockchain and the real world, enabling seamless integration of smart contracts with existing business systems. Oracles employ an adapter architecture, developing dedicated connectors for different types of business systems, including connectors for work order management systems, enterprise resource planning systems, and warehouse management systems.

[0060] Data transformation is the core function of oracles. Oracles are responsible for converting blockchain events into API calls that business systems can understand, and simultaneously converting the responses from business systems into blockchain transactions. During the data transformation process, oracles perform operations such as data format verification and business rule checks to ensure the accuracy and consistency of the data.

[0061] Transaction consistency management employs the Saga pattern to handle distributed transactions. When business operations need to be executed across multiple systems, the oracle breaks down complex business processes into a series of compensable atomic operations. Each atomic operation is logged upon successful execution; if an operation fails, compensating operations are executed in reverse order to ensure eventual consistency of the system.

[0062] S105: Data Verification and Auditing Phase. A multi-level verification mechanism is employed to verify the data uploaded to the blockchain and to audit and trace the entire chain of operations involving the uploaded data. This step provides complete data credibility verification and audit trail functions, ensuring transparency and trustworthiness throughout the entire chain from data collection to business execution. For example... Figure 6 As shown, the data verification and auditing phase is as follows: 1) Integrity verification system: The integrity verification system employs a multi-level verification mechanism to verify the integrity and authenticity of data from different perspectives. Specifically, this includes: The first layer is digital signature verification: Verifying the digital signature confirms the trustworthiness of the data source. The sender uses their private key to encrypt the hash value of the data (i.e., the digital fingerprint of the data), generating a unique digital signature. During verification, the receiver uses the sender's public key to decrypt the signature, restoring the first hash value (H1), and simultaneously independently calculates the second hash value (H2) of the received data. Finally, H1 and H2 are compared to verify the data source and integrity: if they match perfectly, it proves that the data was indeed issued by the claimed sender and has not been tampered with since signing; if they do not match, the verification fails.

[0063] The second layer involves Merkel proof verification: Verifying a data block's Merkel proof confirms its membership in the existing data set. The verifier holds the data to be verified, several key hash values ​​(i.e., the Merkel path) along a root hash (Merkel root) path from the data's hash value to the entire data set, and a publicly available trusted Merkel root on the blockchain. Starting with the data to be verified, the verifier recursively calculates layer by layer using the hash values ​​in the provided Merkel path, ultimately deriving a root hash value. If this calculated root hash perfectly matches the trusted Merkel root stored on the blockchain, it proves that the data is undoubtedly a formal member of the set.

[0064] The third layer involves hash value comparison. The hash value of the abnormal data contained in the initial data corresponding to the abnormal data on the chain is recalculated, and the recalculated hash value is compared with the hash value stored on the chain. If the hash values ​​match, it indicates that the data has not been tampered with.

[0065] To support large-scale verification needs, the system implements batch verification optimization. Batch verification improves verification efficiency by combining multiple verification tasks into batches and utilizing parallel computing techniques. The system also establishes a verification result caching mechanism to skip duplicate verifications of already verified data, further optimizing performance.

[0066] 2) Audit trail system: The audit trail system records operation logs throughout the entire process, from data collection to business execution. Each log entry includes detailed information such as timestamp, operation type, executor, input parameters, output results, and associated transaction hashes. Log data is stored in a time-series database, supporting queries by multiple dimensions such as time range, device identifier, and operation type.

[0067] The system offers a wealth of audit and analysis functions, including operation statistics, anomaly detection, and compliance checks. The operation statistics function summarizes system usage from different dimensions, helping administrators understand the system's operational status; the anomaly detection function identifies suspicious behavior by analyzing operation patterns; and the compliance check function verifies whether various operations comply with predefined security policies and business specifications.

[0068] Audit report generation is a key output of the audit system. The system periodically generates compliance reports, summarizing information such as system operation status, anomalies, and compliance status within a specified time period. The reports use a structured format, facilitating quick understanding of system performance and decision-making by administrators.

[0069] Example 2 This embodiment provides a system for implementing the above method.

[0070] System overall architecture design This system adopts a microservice-based distributed architecture, decomposing system functions into multiple loosely coupled service units. The overall architecture consists of six layers: edge computing layer, data filtering layer, blockchain layer, smart contract layer, business integration layer, and data verification and auditing layer. Each layer communicates with other layers through well-defined interfaces, ensuring the system's flexibility and scalability. The system uses a communication method combining RESTful APIs and message queues, supporting both synchronous request-response and asynchronous event-driven modes.

[0071] The edge computing layer is responsible for collecting and initially processing device data, and is deployed on the edge gateway device at the device site; the data filtering layer is responsible for filtering out abnormal data, and is deployed on the edge computing layer; the blockchain layer provides distributed and trusted evidence storage capabilities, and is built on a consortium blockchain network; the smart contract layer realizes the automated execution of business logic, and is deployed in the blockchain network; the business integration layer is responsible for integration with existing enterprise information systems, and realizes data exchange through oracle services; the data verification and auditing layer is responsible for data credibility verification and full-link audit tracing.

[0072] like Figure 7 As shown, the system specifically includes interconnected systems: Edge computing module The edge computing module is the system's data entry point, responsible for the real-time acquisition and processing of device operation data.

[0073] In terms of hardware, the system adopts an industrial-grade edge computing gateway, equipped with a multi-core processor, large-capacity memory, and abundant I / O interfaces. The gateway device supports a wide operating temperature range, and its mean time between failures (MTBF) meets industrial-grade standards. The sensor network includes various types of industrial sensors such as vibration, temperature, pressure, and current sensors, with accuracy levels meeting relevant industry standards.

[0074] The software architecture adopts a layered design. The bottom layer is the device driver layer, which provides a unified device access framework and supports mainstream industrial communication protocols such as Modbus, OPC UA, and PROFIBUS. The middle layer is the data processing layer, which integrates a stream processing engine and a complex event processing framework to achieve real-time data processing and analysis. The upper layer is the application service layer, which provides functions such as data interfaces, configuration management, and status monitoring.

[0075] The system implements several key technical features. In terms of data acquisition, it supports multiple sampling modes and triggering conditions to adapt to different monitoring needs; in terms of data processing, it provides a configurable processing pipeline, allowing users to combine different processing algorithms as needed; and in terms of communication, it supports multiple network connection methods, including wired and wireless, and features reconnection after disconnection and data caching mechanisms.

[0076] Intelligent data filtering module The intelligent data filtering module is the core of the system's data preprocessing. It is responsible for filtering massive amounts of device data through intelligent algorithms to ensure that only exact abnormal data is submitted to the blockchain for evidence storage.

[0077] The intelligent data filtering module adopts a two-level cascaded fusion detection architecture. The first level is a dynamic threshold screening unit, which implements an adaptive threshold algorithm based on equipment operating status. This unit establishes a dynamic threshold model for each monitoring parameter. The model input includes multi-dimensional factors such as equipment load rate, environmental parameters, and cumulative operating time, allowing the anomaly judgment criteria to be dynamically adjusted according to operating conditions. Anomaly judgment adopts a multi-feature joint evaluation mechanism to avoid false alarms caused by fluctuations in a single parameter. Data is only passed to the second level when multiple related parameters are abnormal. The second level is a machine learning deep analysis unit, which uses a combination of feature engineering and isolated forest methods. Feature extraction covers the time domain (such as kurtosis, impulse index), frequency domain (such as feature frequency energy), and time-frequency domain (such as wavelet energy entropy), constructing a comprehensive feature vector.

[0078] The intelligent data filtering module achieves key technical features. Regarding algorithm adaptability, an incremental learning mechanism enables the model to continuously self-optimize, and concept drift detection is integrated, automatically triggering model retraining when data distribution changes, ensuring the long-term effectiveness of detection capabilities. In terms of processing performance, the streaming engine guarantees real-time data processing, meeting the low-latency requirements of industrial scenarios.

[0079] The intelligent data filtering module management function supports flexible strategy configuration. The threshold management interface allows users to adjust the baseline value and sensitivity of dynamic thresholds; the feature and model management interface supports the selection and configuration of feature combinations and detection algorithms; the monitoring interface displays key indicators such as data filtering rate and anomaly confirmation rate in real time and is integrated with the system monitoring platform.

[0080] Blockchain Evidence Storage Module The blockchain evidence storage module provides the system with a trusted data evidence storage infrastructure.

[0081] The network architecture adopts a permissioned consortium blockchain model, jointly maintained by multiple stakeholders including equipment manufacturers, maintenance service providers, and equipment users. Each organization runs at least one blockchain node, and nodes are connected via a P2P network. The consensus mechanism employs a practical Byzantine fault-tolerant algorithm, providing good performance while ensuring consistency.

[0082] The storage architecture design fully considers business needs. The distributed ledger adopts a multi-chain structure, storing different types of data on different chains to improve the system's parallel processing capabilities. The state database uses a key-value database that supports rich query functions to store the system's current state. For large volumes of unstructured data, the system integrates a distributed file system, which is associated with the blockchain through hash references.

[0083] Security mechanisms are a crucial component of the blockchain module. The system establishes an identity management system based on public key infrastructure, where each device and user possesses a unique digital identity. Access control employs an attribute-based access control model, supporting fine-grained permission management. Data transmission is encrypted throughout the entire process using transport layer security protocols to prevent data theft or tampering during transmission.

[0084] Smart contract module The smart contract module implements the coding and automated execution of business logic.

[0085] Contract development offers complete toolchain support, including contract programming, testing, deployment, and upgrade capabilities. The system provides a rich contract template library covering common operational scenarios, reducing the difficulty of contract development. Security auditing tools automatically detect security vulnerabilities and logical errors in contracts, ensuring contract quality and security.

[0086] The contract execution environment has been specifically optimized. The virtual machine uses WebAssembly technology to provide high-performance contract execution capabilities. The gas mechanism implements reasonable resource billing to prevent malicious contracts from consuming excessive system resources. The event mechanism provides comprehensive event publish-subscribe functionality, supporting decoupled communication between system components.

[0087] The contract management feature supports full lifecycle management of contracts. The version control system records the contract's change history and supports contract rollback and canary releases. The access control system controls contract access permissions, ensuring that only authorized users can invoke sensitive operations. The monitoring system tracks the contract's execution status in real time, promptly detecting and handling anomalies.

[0088] Business integration module The business integration module enables seamless integration between the blockchain network and existing enterprise information systems.

[0089] The API gateway serves as the unified entry point for business integration modules, providing functions such as protocol conversion, traffic control, and security authentication. The gateway supports multiple API protocols, including RESTful, GraphQL, and gRPC, adapting to the integration needs of different systems. The traffic control mechanism prevents sudden traffic surges from impacting the backend system, ensuring system stability. The security authentication module verifies the legitimacy of each API request, preventing unauthorized access.

[0090] The data transformation engine is responsible for converting data formats between different systems. The engine supports multiple data formats such as XML, JSON, and Protocol Buffers, and provides a visual mapping configuration interface. The rules engine allows users to define business rules, enabling flexible data processing and validation logic. The data validation module rigorously checks input and output data to ensure data integrity and accuracy.

[0091] The transaction manager handles distributed transactions across systems. Based on the Saga pattern, it implements eventual consistency transaction management, decomposing complex business operations into a series of compensable atomic operations. The compensation mechanism ensures correct rollback in case of operation failure, maintaining system consistency. The retry strategy implements an intelligent retry mechanism, improving the system's fault tolerance to temporary failures.

[0092] Data Validation and Auditing Module The data verification and auditing module provides the system with end-to-end data credibility assurance and transparent operation auditing capabilities, and is a key infrastructure for building a trusted operation and maintenance ecosystem.

[0093] The data verification and auditing module consists of two core subsystems. The integrity verification system implements a multi-layered trusted verification mechanism. First, the system verifies the digital signature of transactions to ensure the authenticity and non-repudiation of the data source. Second, based on Merkel proofs, it verifies that a specific data block belongs to a consensus-based block, confirming its existence. Finally, by comparing the consistency between the original off-chain data and the hash value stored on-chain, it completes the ultimate verification of data integrity. To cope with high-concurrency verification requirements, the system integrates batch verification optimization and result caching mechanisms, significantly improving verification efficiency. The audit trail system is responsible for recording the entire operation log. The system captures every key event from data acquisition, edge processing, blockchain notarization to smart contract calls. The log records include precise timestamps, operating entities, inputs and outputs, related transaction hashes, and other detailed information, and is stored in a high-performance time-series database.

[0094] The data verification and auditing module incorporates several key technological innovations. In verification, it employs lightweight verification technology, allowing users to quickly complete verification without downloading the complete blockchain data, thus reducing client resource requirements. For auditing and analysis, it integrates a powerful rule engine and machine learning algorithms, supporting real-time anomaly detection and automated compliance checks, proactively identifying potential security risks or violations.

[0095] The data verification and auditing module provides comprehensive management functions. Audit logs support granular querying and visualization across multiple dimensions (such as time, device, and operation type) to meet various investigation needs. The system has flexible audit report generation capabilities, allowing for the customization of periodic compliance reports or specific event analysis reports as needed. All configuration items, such as verification rules, audit strategies, and report templates, are centrally managed through the configuration management center, ensuring the controllability and compliance of system operations.

[0096] System monitoring and management platform The system monitoring and management platform provides comprehensive system operation and maintenance support.

[0097] The monitoring metrics system covers the operational status of all levels of the system. Infrastructure monitoring tracks server resource usage, including metrics such as CPU, memory, disk, and network; application performance monitoring tracks the execution efficiency of business functions, including metrics such as response time, throughput, and error rate; and blockchain monitoring tracks the operational status of the blockchain network, including metrics such as block height, number of transactions, and network latency.

[0098] The log management system centrally collects, stores, and analyzes logs. The log collector gathers log data from various system components and stores it uniformly in a time-series database. The log analysis engine provides powerful query and analysis capabilities, supporting real-time log search and statistical analysis. The alarm system automatically detects anomalies based on predefined rules and notifies operations and maintenance personnel through multiple channels.

[0099] The configuration management center provides unified system configuration management functions. Configuration information is version-based, supporting configuration tracking and rollback. Configuration changes undergo a strict approval process to ensure compliance. A configuration distribution mechanism ensures that configuration information is synchronized to all system components in a timely manner.

[0100] System Deployment Architecture The system supports flexible deployment options to adapt to different application scenarios and resource conditions.

[0101] The cloud deployment solution deploys all system components on a cloud platform, making it suitable for scenarios with ample resources and high availability requirements. This solution leverages the elastic scaling capabilities of the cloud platform to dynamically adjust resource allocation based on business load, while also benefiting from the high availability and disaster recovery capabilities provided by the cloud platform.

[0102] Edge deployment solutions place critical components at edge nodes, making them suitable for scenarios with high data privacy requirements and poor network conditions. This approach reduces data transmission over the network, lowers network latency, and better meets compliance requirements for data localization.

[0103] The hybrid deployment solution combines the advantages of cloud and edge computing to achieve a reasonable distribution of load and data. The core blockchain network and business integration components are deployed in the cloud, while the edge computing components are deployed on-site, ensuring both centralized system management capabilities and meeting the real-time requirements of data processing.

[0104] Example 3 The present invention also provides an electronic device, including: a processor, a transmitting device, an input device, an output device, and a memory. The processor may be implemented using a general-purpose CPU (Central Processing Unit), a microprocessor, an application-specific integrated circuit, or one or more integrated circuits, and is used to execute relevant programs to implement the technical solutions provided in the embodiments of this application. The memory may be implemented using a read-only memory (ROM), a static storage device, a dynamic storage device, or a random access memory (RAM), and is used to store computer program code. The computer program code includes computer instructions. When the processor executes the computer instructions, the electronic device executes a method as described in any of the above possible implementation methods.

[0105] Example 4 The present invention also provides a computer-readable storage medium storing a computer program, the computer program including program instructions, which, when executed by a processor of an electronic device, cause the processor to perform a method as described in any of the above possible implementations.

[0106] In the description of this specification, the references to terms such as "an embodiment," "example," "specific example," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of the present invention. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples.

[0107] The above description is merely a specific embodiment of the present invention, enabling those skilled in the art to understand or implement the invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the invention. Therefore, the present invention is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features claimed herein.

Claims

1. A method for uploading IoT data to the blockchain for equipment operation and maintenance, characterized in that, Includes the following steps: S101, Data Acquisition and Preprocessing Stage: Data on equipment operation is acquired through a sensor network deployed in key parts of the equipment, and edge computing is performed on the data to generate initial data. S102, Intelligent data screening stage: A two-level fusion anomaly detection mechanism, including dynamic threshold screening and machine learning deep analysis, is used to screen out abnormal data from the initial data. S103, On-chain-off-chain collaborative storage stage: The initial data containing abnormal data is completely stored in the off-chain system, the hash value of the abnormal data is calculated, a data storage data package is generated, and the data storage data package is sent to the blockchain network for on-chain storage. S104, Smart Contract and Business Integration Stage: In response to the event of the evidence storage data packet being uploaded to the blockchain, a predefined smart contract is automatically triggered to automatically execute at least one operation in operation and maintenance work order management, resource allocation, and supply chain collaboration. S105, Data Verification and Auditing Phase: A multi-level verification mechanism is used to verify the on-chain data and to audit and track the entire chain operation of the on-chain data.

2. The method according to claim 1, characterized in that, A two-level fusion anomaly detection mechanism, incorporating dynamic threshold screening and deep machine learning analysis, is used to filter out anomalous data, including: Level 1: Dynamic threshold screening, specifically: Based on the equipment's historical operating data, establish dynamic threshold models for various monitoring indicators under different operating conditions, wherein the thresholds are dynamically adjusted according to the equipment load and ambient temperature; if multiple real-time data exceed the current dynamic threshold range, they are marked as suspicious data; The second level is machine learning deep analysis, which specifically involves extracting feature vectors containing time-domain and frequency-domain features from the suspicious data and inputting them into a pre-trained lightweight isolated forest model for analysis, outputting abnormal data.

3. The method according to claim 2, characterized in that, The input is analyzed in a pre-trained lightweight isolated forest model, and the output includes anomalous data, including: The lightweight isolated forest model determines the anomalous probability of the input feature vector by calculating the path length of the data point in the isolated tree. When the probability exceeds the preset confidence level, the data is determined to be anomalous and is triggered to be added to the chain.

4. The method according to claim 1, characterized in that, Step S103 includes: Based on access frequency and value density, the initial data containing anomalous data is stored in different off-chain storage media using a distributed storage architecture and hierarchical data lifecycle management. The cryptographic hash value of the abnormal data is calculated, and the hash value is packaged with key metadata including timestamp, device identifier and abnormal type label to form a data packet containing key digest information, which is then sent to the blockchain network for on-chain storage.

5. The method according to claim 1, characterized in that, Step S104 specifically includes: In response to the event of the evidence storage data packet being uploaded to the blockchain, predefined modular smart contracts on the blockchain are automatically triggered, wherein the smart contracts include work order management contracts, resource optimization contracts and supply chain collaboration contracts; The work order management contract calls the declared event result judgment logic to perform conflict judgment on the event data in the evidence data packet. If there is a conflict, the final result of the physical event is determined based on the preset consensus protocol, and the operation and maintenance work order is updated and its status is tracked according to the final result. The resource optimization contract, based on the maintenance work order information generated by the work order management contract, executes a multi-factor evaluation algorithm to select the best engineer from the maintenance technical service personnel database for work order assignment. The supply chain collaboration contract automatically generates spare parts demand orders based on the abnormal events or fault types identified by the work order management contract, and transmits the demand to the external inventory management system through the oracle service to track the outbound and logistics status of spare parts.

6. The method according to claim 1, characterized in that, The multi-level verification mechanism includes: Digital signature verification involves using the sender's public key to decrypt the digital signature attached to the received data to obtain a first hash value H1, independently calculating a second hash value H2 of the received data, and comparing the first hash value H1 with the second hash value H2 to verify the data source and integrity. Merkel proof involves recursively calculating a root hash value based on the data to be verified, the Merkel path corresponding to the data to be verified, and the trusted Merkel root stored on the chain. The calculated root hash value is then matched with the trusted Merkel root to verify that the data to be verified belongs to a specific data set. Hash value comparison involves recalculating the hash value of the abnormal data contained in the initial data corresponding to the abnormal data on the chain, and comparing the recalculated hash value with the initial hash value stored on the chain. If they match, it is determined that the data has not been tampered with.

7. An IoT data uplink system for equipment operation and maintenance, characterized in that, The system uses the method as described in any one of claims 1-6, the system comprising mutually communicatively connected components: The edge computing module is configured to collect device operation data and perform edge preprocessing. The intelligent data filtering module is configured to filter out abnormal data. The blockchain evidence storage module is configured to receive and store the hash value and metadata of abnormal data. The smart contract module is configured to trigger the execution of predefined business logic in response to evidence storage events; The business integration module is configured to transmit the execution results of smart contracts to external business systems; The data verification and auditing module is configured to provide data credibility verification and end-to-end audit tracing capabilities.

8. The system according to claim 7, characterized in that, The edge computing module supports multiple industrial communication protocols and integrates a stream processing engine; the blockchain evidence storage module adopts a permissioned consortium blockchain network and a practical Byzantine fault-tolerant consensus mechanism.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the program, it implements the method as described in any one of claims 1-6.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the program is executed by the processor, it implements the method as described in any one of claims 1-6.

Images