Production traceability and transaction collaboration method and system of MES (Manufacturing Execution System)

Semantic alignment of heterogeneous data is achieved through distributed data acquisition network and deep learning technology, and a traceable production process knowledge graph is built in combination with graph database and blockchain technology, and intelligent collaborative decision-making is used to solve the problem of traditional MES systems in Inadequate data integration, traceability and decision-making have improved the integrity and efficiency of the production process.

CN119067689AActive Publication Date: 2024-12-03冠骋信息技术(苏州)有限公司

Patent Information

Application Number
CN202411563725.1
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-11-05
Publication Date
2024-12-03
Estimated Expiration
2044-11-05

AI Technical Summary

Technical Problem

Traditional MES systems have problems such as insufficient data integration capabilities, incomplete traceability of production process traceability, and limited collaborative decision-making in dealing with heterogeneous data in complex production environments, realizing comprehensive traceability and collaborative decision-making capabilities.

Method used

Use distributed data acquisition network, edge computing, deep learning and transfer learning methods to perform real-time acquisition and semantic alignment of heterogeneous data, use graph database and blockchain technology to build a traceable production process knowledge graph, combine graph neural network and enhancement Learning algorithms enable intelligent collaborative decision-making.

Benefits of technology

It improves the integrity and credibility of data processing capabilities and production process traceability, enhances intelligent collaborative decision-making capabilities, and significantly improves production efficiency and emergency response capabilities.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119067689A_ABST
    Figure CN119067689A_ABST
Patent Text Reader

Abstract

The invention provides a production traceability and transaction collaboration method and system of an MES system, and relates to the technical field of production traceability, and the method comprises the steps: constructing a production process knowledge graph through employing a graph database technology, and mapping the fused multi-source heterogeneous data into the entity and relation of the knowledge graph; based on a time sequence diagram algorithm, dynamic time sequence association is established in the knowledge graph, and a traceable production process chain is formed; constructing a transaction triggering mechanism based on the knowledge graph and the anomaly detection model, and automatically identifying production transactions needing cooperative processing to obtain a transaction cooperative record; a reinforcement learning algorithm is utilized to establish a multi-agent collaborative decision-making model, and the model can adaptively allocate tasks and resources according to the current production state and historical experience; converting an unstructured collaborative demand into an executable workflow through a sequence-to-sequence model enhanced by an attention mechanism; and automatically generating a multi-dimensional production analysis report based on a traceable production process chain and transaction collaborative record.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to production tracing technology, and in particular to a production tracing and transaction collaboration method and system of an MES system. Background Art

[0002] With the rapid development of industry, the manufacturing industry has placed higher demands on refined management and real-time traceability of production processes. Traditional Manufacturing Execution Systems (MES) face enormous challenges in processing heterogeneous data in complex production environments, achieving comprehensive traceability, and collaborative decision-making.

[0003] The existing MES system has the following main defects:

[0004] Inadequate data integration and processing capabilities: Traditional MES systems often utilize a centralized data processing architecture, making it difficult to effectively handle massive amounts of heterogeneous data from various sources. Data silos and inconsistent information often arise during data collection, cleaning, and integration, hindering real-time monitoring and accurate traceability of the production process.

[0005] The integrity and credibility of production process traceability are low: Existing systems typically use relational databases to store production data, making it difficult to effectively represent and analyze complex production process relationships. Furthermore, the lack of reliable anti-tampering mechanisms makes it impossible to guarantee the authenticity and integrity of traceability data, creating potential risks in quality traceability and accountability.

[0006] Limited collaborative decision-making capabilities: Traditional MES systems rely heavily on pre-set rules and manual intervention for exception handling and resource scheduling, lacking intelligent collaborative decision-making mechanisms. In complex and volatile production environments, rapid and accurate transaction collaboration and optimal resource allocation are difficult to achieve, impacting production efficiency and emergency response capabilities.

[0007] Therefore, there is an urgent need for a new MES system that can effectively integrate multi-source heterogeneous data, achieve comprehensive and reliable traceability, and have intelligent collaborative decision-making capabilities to meet the needs of modern manufacturing for efficient, flexible and intelligent production management. Summary of the Invention

[0008] The embodiments of the present invention provide a production tracing and transaction collaboration method and system for an MES system, which can solve the problems in the prior art.

[0009] According to a first aspect of the embodiments of the present invention,

[0010] Provides production traceability and transaction collaboration methods for the MES system, including:

[0011] Deploy a distributed data acquisition network to collect heterogeneous data in real time from production equipment, environmental sensors, human-computer interaction terminals, and enterprise information systems; classify and annotate the collected heterogeneous data to establish a unified data model architecture; use edge computing nodes to perform preliminary cleaning, denoising, and standardization on the heterogeneous data to generate structured data streams; transmit the structured data streams to the central data processing unit, use deep learning algorithms to extract data features and understand semantics, and construct multidimensional data feature vectors; based on these multidimensional data feature vectors, use transfer learning methods to establish a cross-domain data association model to achieve semantic alignment and fusion of heterogeneous data;

[0012] Leveraging graph database technology, we construct a knowledge graph for the production process, mapping the fused multi-source heterogeneous data to the entities and relationships within the knowledge graph. Based on a time-series graph algorithm, we establish dynamic time-series associations within the knowledge graph, forming a traceable production process chain. Leveraging blockchain technology, we create tamper-proof digital fingerprints for key nodes and transactions in the production process and associate these fingerprints with the corresponding entities within the knowledge graph. We also construct an anomaly detection model based on a graph neural network to monitor data flows and relationship changes within the knowledge graph in real time, identifying potential production anomalies and quality risks.

[0013] Based on the knowledge graph and anomaly detection model, a transaction trigger mechanism is constructed to automatically identify production transactions that require collaborative processing and obtain transaction collaboration records; using the reinforcement learning algorithm, a multi-agent collaborative decision-making model is established, which can adaptively allocate tasks and resources based on the current production status and historical experience; through the sequence-to-sequence model enhanced by the attention mechanism, unstructured collaboration requirements are converted into executable workflows; based on the traceable production process chain and transaction collaboration records, multi-dimensional production analysis reports are automatically generated.

[0014] In an optional embodiment,

[0015] Based on the multi-dimensional data feature vector, a transfer learning method is used to establish a cross-domain data association model to achieve semantic alignment and fusion of heterogeneous data, including:

[0016] Use transfer learning methods to align feature vectors from different domains, including:

[0017] Select feature vectors for the source and target domains; construct a domain adversarial neural network, including a feature extractor; train the domain adversarial neural network to minimize the difference in feature distribution between the source and target domains; and use the trained feature extractor to obtain aligned feature vectors.

[0018] The cosine similarity is used to calculate the similarity between the aligned feature vectors, and the local sensitive hashing algorithm is used to accelerate the similarity calculation between large-scale vectors to obtain the similarity matrix;

[0019] Based on the similarity matrix, a spectral clustering algorithm is used to cluster the cross-domain data, including:

[0020] Construct an affinity matrix; calculate a Laplace matrix; calculate the eigenvectors of the Laplace matrix; perform k-means clustering on the eigenvectors, establish correlations between data, and achieve heterogeneous data fusion.

[0021] In an optional embodiment,

[0022] Using graph database technology, we build a knowledge graph of the production process and map the fused multi-source heterogeneous data to the entities and relationships of the knowledge graph. Based on the time sequence graph algorithm, we establish dynamic time sequence associations in the knowledge graph to form a traceable production process chain, including:

[0023] Design a knowledge graph model and define the entity types and relationship types of the production process knowledge graph. The entity types include products, processes, equipment, raw materials, quality indicators, and operators. The relationship types include contain, use, produce, monitor, operate, and precedent.

[0024] Select and configure a graph database platform, including installing a database server, creating a database instance, configuring memory and storage parameters, and enabling necessary plugins. Next, preprocess the fused multi-source heterogeneous data, including data cleaning, standardization, entity recognition, and relationship extraction.

[0025] The preprocessed data is then imported into the graph database using a graph database query language to create nodes and relationships and build an initial knowledge graph. In this initial knowledge graph, timestamp attributes are added to the nodes and relationships, and "happens after" temporal relationships are created based on the timestamps to establish dynamic temporal associations.

[0026] Based on the established temporal relationship, a temporal traversal algorithm is developed for temporal path query in the knowledge graph; by utilizing the constructed knowledge graph and temporal relationship, forward tracing, reverse tracing and quality problem tracing functions are realized to form a traceable production process chain.

[0027] In an optional embodiment,

[0028] Based on blockchain technology, tamper-proof digital fingerprints are created for key nodes and transactions in the production process, and the digital fingerprints are associated with the corresponding entities in the knowledge graph, including:

[0029] Design and deploy a permissioned blockchain network based on an open-source blockchain distributed ledger, define the network topology, configure channels corresponding to production lines or product types, and write smart contracts for digital fingerprint creation, storage, and verification. Use a multi-hash algorithm to generate digital fingerprints, including combining production process data, timestamps, and random numbers, performing a first hash operation using the SHA256 algorithm, and then performing a second hash operation on the first hash result using the SHA3-256 algorithm to obtain the final digital fingerprint. Implement the digital fingerprint storage function in the smart contract, encapsulating the entity ID, digital fingerprint, and timestamp into a structure, serializing it, and storing it in the blockchain world state.

[0030] Add digital fingerprint attributes to entities in the knowledge graph, create new nodes representing blockchain transactions, and establish the relationship between entity nodes and blockchain transaction nodes; develop a digital fingerprint verification process, first retrieve the digital fingerprint information of the target entity from the knowledge graph, then call the verification function of the blockchain smart contract, pass in the entity ID and digital fingerprint, and finally return the verification result; implement an integrated query interface for the knowledge graph and blockchain, and support simultaneous acquisition of entity information in the knowledge graph and digital fingerprint verification results in the blockchain based on the entity ID.

[0031] In an optional embodiment,

[0032] Using reinforcement learning algorithms, a multi-agent collaborative decision-making model is established. This model can adaptively allocate tasks and resources based on the current production status and historical experience. Through a sequence-to-sequence model enhanced by an attention mechanism, unstructured collaborative requirements are transformed into executable workflows, including:

[0033] A multi-agent deep deterministic policy gradient model based on the actor-critic architecture is constructed, in which each agent represents a workstation or piece of equipment on the production line. Each agent consists of an actor network for generating actions and a critic network for evaluating the value of actions. A state space is defined, which includes the current workstation's task queue, resource utilization, and the states of neighboring workstations, as well as an action space that includes task allocation decisions and resource scheduling strategies. A reward function is designed that comprehensively considers production efficiency, resource utilization, and task completion quality.

[0034] Utilizing the multi-agent deep deterministic policy gradient model, selecting the optimal action for each workstation based on the observed current state, and performing corresponding task allocation and resource scheduling operations; calculating the reward value after executing the action, and using the gained experience to update the multi-agent deep deterministic policy gradient model;

[0035] Build an attention-enhanced sequence-to-sequence model consisting of a bidirectional LSTM long short-term memory encoder, an attention layer, and a unidirectional LSTM decoder; convert the unstructured collaboration requirement text into a word sequence and encode it, inputting it into the attention-enhanced sequence-to-sequence model to generate the corresponding workflow sequence; and decode the generated workflow sequence into a readable workflow description.

[0036] According to the decoded workflow description, the updated multi-agent deep deterministic policy gradient model is used for task allocation and resource scheduling. During the execution of the production process, the latest production status information and historical experience data are continuously obtained, and the multi-agent deep deterministic policy gradient model is dynamically updated to achieve adaptive task allocation and resource scheduling optimization.

[0037] In an optional embodiment,

[0038] Based on the knowledge graph and anomaly detection model, a transaction trigger mechanism is built to automatically identify production transactions that require collaborative processing. The transaction collaboration records obtained include:

[0039] Design a multi-layer autoencoder network structure as an anomaly detection model, train the autoencoder using normal production data, calculate the reconstruction error, and set a threshold for anomaly determination; collect sensor data and production status information from the production system in real time, and input the collected data into the trained anomaly detection model for analysis;

[0040] When an anomaly is detected, context analysis is performed using the knowledge graph to determine the impact and possible causes of the anomaly. Based on the anomaly and the knowledge graph analysis results, a decision is made as to whether collaborative processing needs to be initiated. For transactions requiring collaborative processing, a transaction collaboration record is generated containing a timestamp, anomaly information, affected equipment, affected processes, and collaboration requirements.

[0041] The generated transaction collaboration records are input into the multi-agent collaborative decision-making system for task allocation and resource scheduling; the knowledge graph and anomaly detection model are continuously updated during the production process, and the anomaly detection threshold and collaborative processing strategy are dynamically adjusted; the knowledge graph is supplemented and corrected based on the collaborative processing results and production effects, and the anomaly detection model is incrementally learned using newly added anomaly samples.

[0042] According to a second aspect of the embodiments of the present invention,

[0043] Provides MES system production traceability and transaction collaboration system, including:

[0044] The first unit is used to deploy a distributed data acquisition network to collect heterogeneous data in real time from production equipment, environmental sensors, human-computer interaction terminals, and enterprise information systems; classify and annotate the collected heterogeneous data to establish a unified data model architecture; use edge computing nodes to perform preliminary cleaning, denoising, and standardization on the heterogeneous data to generate a structured data stream; transmit the structured data stream to the central data processing unit, use deep learning algorithms to extract data features and understand semantics, and construct a multidimensional data feature vector; based on the multidimensional data feature vector, use transfer learning methods to establish a cross-domain data association model to achieve semantic alignment and fusion of heterogeneous data;

[0045] The second unit uses graph database technology to build a production process knowledge graph, mapping the fused multi-source heterogeneous data to the entities and relationships of the knowledge graph. Based on a time-series graph algorithm, dynamic time-series associations are established in the knowledge graph to form a traceable production process chain. Based on blockchain technology, tamper-proof digital fingerprints are created for key nodes and transactions in the production process and associated with the corresponding entities in the knowledge graph. An anomaly detection model based on a graph neural network is constructed to monitor the data flow and relationship changes in the knowledge graph in real time, identifying potential production anomalies and quality risks.

[0046] The third unit is used to build a transaction trigger mechanism based on knowledge graphs and anomaly detection models, automatically identify production transactions that require collaborative processing, and obtain transaction collaboration records; use reinforcement learning algorithms to establish a multi-agent collaborative decision-making model, which can adaptively allocate tasks and resources based on current production status and historical experience; through a sequence-to-sequence model enhanced by an attention mechanism, unstructured collaboration requirements are converted into executable workflows; based on traceable production process chains and transaction collaboration records, multi-dimensional production analysis reports are automatically generated.

[0047] According to a third aspect of the embodiments of the present invention,

[0048] An electronic device is provided, comprising:

[0049] processor;

[0050] a memory for storing processor-executable instructions;

[0051] The processor is configured to call the instructions stored in the memory to execute the aforementioned method.

[0052] According to a fourth aspect of the embodiments of the present invention,

[0053] A computer-readable storage medium is provided, on which computer program instructions are stored. When the computer program instructions are executed by a processor, the method described above is implemented.

[0054] Beneficial effects:

[0055] 1. Improve data integration and processing capabilities:

[0056] This invention utilizes a distributed data acquisition network and edge computing technology to achieve real-time acquisition and preliminary processing of heterogeneous data. By using deep learning algorithms and transfer learning methods, it constructs multidimensional data feature vectors and cross-domain data association models, effectively resolving the data silo problem and achieving semantic alignment and fusion of heterogeneous data. This approach significantly improves the system's ability to process massive amounts of heterogeneous data, laying a solid data foundation for real-time monitoring and precise analysis of production processes.

[0057] 2. Enhance the integrity and credibility of production process traceability:

[0058] This invention utilizes graph database technology to construct a production process knowledge graph and, combined with a time-series graph algorithm, establishes dynamic time-series associations, creating a complete traceable production process chain. Furthermore, by introducing blockchain technology to create tamper-proof digital fingerprints for key nodes and transactions, the authenticity and integrity of traceability data are significantly improved. This innovative traceability mechanism not only enhances the accuracy of quality management but also provides reliable technical support for accountability determination, effectively reducing production risks.

[0059] 3. Realize intelligent collaborative decision-making:

[0060] This invention builds an intelligent transaction triggering mechanism based on knowledge graphs and graph neural networks, capable of automatically identifying production transactions requiring collaborative processing. A multi-agent collaborative decision-making model, combined with a reinforcement learning algorithm, enables adaptive allocation of tasks and resources. A sequence-to-sequence model enhanced with an attention mechanism transforms unstructured collaborative requirements into executable workflows. This series of innovative technologies significantly enhances the system's intelligent collaborative decision-making capabilities, significantly improving production efficiency and emergency response capabilities, and providing strong support for enterprises to achieve flexible production and intelligent manufacturing. BRIEF DESCRIPTION OF THE DRAWINGS

[0061] Figure 1 Schematic diagram of the process of production tracing and transaction collaboration method of the MES system according to an embodiment of the present invention;

[0062] Figure 2 This is a structural diagram of the production traceability and transaction collaboration system of the MES system according to an embodiment of the present invention. DETAILED DESCRIPTION

[0063] To make the objectives, technical solutions, and advantages of the embodiments of the present invention more clear, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts shall fall within the scope of protection of the present invention.

[0064] The following specific embodiments are used to describe the technical solution of the present invention in detail. The following specific embodiments can be combined with each other, and the same or similar concepts or processes may not be described in detail in some embodiments.

[0065] Figure 1 FIG. 1 is a flow chart of a production traceability and transaction collaboration method of an MES system according to an embodiment of the present invention. Figure 1 As shown, the method includes:

[0066] S101. Deploy a distributed data acquisition network to collect heterogeneous data in real time from production equipment, environmental sensors, human-computer interaction terminals, and enterprise information systems; classify and annotate the collected heterogeneous data to establish a unified data model architecture; utilize edge computing nodes to perform preliminary cleaning, denoising, and standardization on the heterogeneous data to generate a structured data stream; transmit the structured data stream to a central data processing unit, utilize deep learning algorithms to extract data features and perform semantic understanding, and construct a multidimensional data feature vector; based on the multidimensional data feature vector, employ transfer learning methods to establish a cross-domain data association model to achieve semantic alignment and fusion of heterogeneous data;

[0067] S102. Utilize graph database technology to construct a production process knowledge graph, mapping the fused multi-source heterogeneous data to the entities and relationships within the knowledge graph. Based on a time-series graph algorithm, establish dynamic time-series associations within the knowledge graph to form a traceable production process chain. Leverage blockchain technology to create tamper-proof digital fingerprints for key nodes and transactions in the production process and associate these fingerprints with corresponding entities within the knowledge graph. Build an anomaly detection model based on a graph neural network to monitor data flows and relationship changes within the knowledge graph in real time, identifying potential production anomalies and quality risks.

[0068] S103. Based on the knowledge graph and anomaly detection model, a transaction trigger mechanism is constructed to automatically identify production transactions that require collaborative processing and obtain transaction collaboration records. A multi-agent collaborative decision-making model is established using a reinforcement learning algorithm. This model can adaptively allocate tasks and resources based on the current production status and historical experience. Through a sequence-to-sequence model enhanced by an attention mechanism, unstructured collaboration requirements are converted into executable workflows. Based on the traceable production process chain and transaction collaboration records, a multi-dimensional production analysis report is automatically generated.

[0069] In this embodiment, the production traceability and transaction collaboration method of the MES (Manufacturing Execution System) system includes the following specific steps:

[0070] First, a distributed data acquisition network is deployed. This network consists of multiple edge nodes, located at key locations throughout the production line. Each edge node connects to a variety of data sources, including PLC controllers for production equipment, environmental monitoring sensors, and human-machine interface terminals used by operators. Raw data is collected from these heterogeneous data sources in real time using standardized industrial communication protocols (such as OPC UA and Modbus TCP).

[0071] Next, the collected heterogeneous data is categorized and annotated. Based on characteristics such as data source type, collection location, and data format, the raw data is divided into categories such as device data, environmental data, manual operation data, and management system data. A unified metadata model is developed for each data type, including attribute fields such as data description, collection time, source device, and unit. Based on this foundation, a unified data model architecture for cross-domain data association is constructed.

[0072] Next, edge computing nodes are used to perform preliminary processing on heterogeneous data. Filtering algorithms are used to remove noise and outliers, such as median filtering for temperature data and wavelet denoising for vibration data. Data cleaning rules are designed based on the characteristics of different data sources, such as removing invalid zero values ​​in device data and filling in missing values ​​in environmental data. Finally, the processed data is converted into a unified JSON format to generate a structured data stream.

[0073] The preprocessed structured data stream is transmitted to a central data processing unit. This unit utilizes a distributed architecture, comprising multiple parallel computing nodes. First, deep learning algorithms are used to extract data features, such as convolutional neural networks (CNNs) for visual features from image data and long-short-term memory (LSTM) networks for temporal features from time series data. Natural language processing techniques are then used to perform semantic understanding of textual data and extract key entities and relationships. Ultimately, multidimensional vectors representing the characteristics of different data sources are generated.

[0074] Based on the obtained multidimensional feature vectors, a cross-domain data association model is constructed using transfer learning. Specifically, a base model is first trained in the source domain (e.g., equipment data). This model is then transferred to the target domain (e.g., quality data) and fine-tuned to adapt to the new data distribution. This approach effectively leverages the commonalities between different data sources to achieve semantic alignment and fusion of heterogeneous data. For example, equipment vibration data can be correlated with product quality data to uncover potential causal relationships.

[0075] Next, we leverage graph database technology to construct a knowledge graph for the production process. First, we define the graph's ontology model, including entity types such as equipment, materials, personnel, and processes, as well as relationship types such as processing, testing, and transportation. We then map the fused, multi-source, heterogeneous data to the entities and relationships within the graph. For example, we map equipment operating data to attributes of the equipment entity, and operation records to relationships between personnel and equipment entities. This creates a comprehensive semantic representation of the production process.

[0076] Based on the constructed knowledge graph, a time-series graph algorithm is used to establish dynamic temporal associations. Specifically, this involves introducing a time dimension into the graph, annotating each entity and relationship with a timestamp. Graph traversal algorithms are then used to track changes in entity states and the evolution of relationships at different points in time. For example, the complete flow of a batch of materials, from raw material warehousing, processing, inspection, to finished product shipment, can be tracked, forming a traceable production process chain.

[0077] To ensure data authenticity and immutability, blockchain technology is used to create digital fingerprints for key nodes and transactions in the production process. Specifically, for important production events (such as critical quality inspections and process parameter adjustments), the relevant data is hashed to generate a unique digital digest. This digest is then recorded on the blockchain network, forming an immutable chain of evidence. Furthermore, transaction hashes on the blockchain are linked to corresponding entities in the knowledge graph to achieve data traceability.

[0078] On this basis, an anomaly detection model based on a graph neural network was constructed. This model uses entities in the knowledge graph as nodes and relationships as edges, and performs feature learning through multi-layer graph convolution. The model is trained using normal production data to learn the normal association patterns between entities during the production process. In practical applications, the model monitors the flow of data and changes in relationships within the knowledge graph in real time, identifying anomalies that deviate from normal patterns. For example, it can detect potential production anomalies and quality risks, such as abnormal fluctuations in equipment parameters and deviations from process paths.

[0079] Based on the constructed knowledge graph and anomaly detection model, a transaction triggering mechanism is designed to automatically identify production transactions requiring collaborative processing. Specifically, when the anomaly detection model identifies a potential issue, the system automatically analyzes the impact and severity of the anomaly to determine whether collaborative processing is necessary. The system also monitors pre-set key performance indicators (KPIs), such as production progress and quality pass rate, and triggers collaborative transactions when these indicators fluctuate abnormally. The system records automatically identified collaborative needs, creating a transaction collaboration record.

[0080] To efficiently handle collaborative tasks, a multi-agent collaborative decision-making model was established using a reinforcement learning algorithm. This model abstracts different production units (such as process, quality, and equipment departments) into agents and adaptively allocates tasks and resources based on the current production status and historical experience. The model's reward function takes into account multiple objectives, including production efficiency, quality level, and cost control. Through continuous learning and optimization, the model is able to make reasonable collaborative decisions in complex and changing production environments.

[0081] For collaborative requirements identified by the system, a sequence-to-sequence model enhanced with an attention mechanism is used to transform unstructured collaborative requirements into executable workflows. The model first encodes the collaborative requirement text and then, based on the learned context, gradually generates standardized workflow steps. Introducing an attention mechanism during the decoding process enables the model to focus on key information in the input text. For example, for a collaborative requirement such as "Equipment A has experienced abnormal fluctuations, requiring analysis by the process department," the model can generate a detailed workflow including steps such as shutdown, inspection, analysis, and adjustment.

[0082] Finally, based on the traceable production process chain and transaction collaboration records, a multi-dimensional production analysis report is automatically generated. This report includes production efficiency analysis, quality trend analysis, abnormality cause analysis, and collaborative effect evaluation. The system extracts relevant data from the knowledge graph and conducts in-depth analysis using statistical analysis and data mining methods. For example, association rule mining algorithms can be used to identify key factors affecting product quality from historical data, while time series analysis can be used to predict the probability of equipment failure. The generated analysis report is presented in a visual format, supporting managers in making data-driven decisions.

[0083] Through the above steps, this method realizes the full process management from data collection, processing, analysis to collaborative decision-making, providing manufacturing enterprises with an efficient and intelligent production traceability and transaction collaboration solution.

[0084] In an optional embodiment, based on the multidimensional data feature vector, a transfer learning method is used to establish a cross-domain data association model to achieve semantic alignment and fusion of heterogeneous data, including:

[0085] Use transfer learning methods to align feature vectors from different domains, including:

[0086] Select feature vectors for the source and target domains; construct a domain adversarial neural network, including a feature extractor; train the domain adversarial neural network to minimize the difference in feature distribution between the source and target domains; and use the trained feature extractor to obtain aligned feature vectors.

[0087] The cosine similarity is used to calculate the similarity between the aligned feature vectors, and the local sensitive hashing algorithm is used to accelerate the similarity calculation between large-scale vectors to obtain the similarity matrix;

[0088] Based on the similarity matrix, a spectral clustering algorithm is used to cluster the cross-domain data, including:

[0089] Construct an affinity matrix; calculate a Laplace matrix; calculate the eigenvectors of the Laplace matrix; perform k-means clustering on the eigenvectors, establish correlations between data, and achieve heterogeneous data fusion.

[0090] In this implementation, based on multidimensional data feature vectors, a transfer learning method is used to establish a cross-domain data association model to achieve semantic alignment and fusion of heterogeneous data. The specific steps are as follows:

[0091] First, transfer learning methods are used to align feature vectors from different domains. Feature vectors for the source and target domains are selected. For example, the source domain can be word vectors for text data, and the target domain can be visual feature vectors for image data. A domain adversarial neural network is constructed, consisting of a feature extractor, a domain classifier, and a label predictor. The feature extractor, consisting of a multi-layer, fully connected network, is used to extract common feature representations for the source and target domain data. The domain classifier is used to determine the domain from which the feature originates, and the label predictor is used to predict the class label of the data.

[0092] The domain adversarial neural network is then trained to minimize the difference in feature distribution between the source and target domains while maintaining the discriminative nature of the features. During training, the feature extractor aims to deceive the domain classifier, preventing it from distinguishing the domain origin of the features, thereby making the extracted features domain-invariant. For example, a gradient reversal layer can be used to implement adversarial training. After multiple rounds of iterative training, the trained feature extractor is used to obtain aligned feature vectors for the source and target domain data.

[0093] Next, cosine similarity is used to calculate the similarity between the aligned feature vectors. Cosine similarity ranges from -1 to 1, with 1 indicating the same direction, 0 indicating orthogonal, and -1 indicating opposite directions. For two n-dimensional vectors a and b, cosine similarity can be calculated by dividing the dot product of the two vectors by the product of their moduli. To accelerate similarity calculations between large-scale vectors, the locality-sensitive hashing (LSH) algorithm is employed. The core idea of ​​LSH is to map similar data into the same "bucket," requiring only the similarity of vectors within the same bucket to be calculated, significantly reducing the computational effort. In practice, random projection can be used to generate multiple hash functions, mapping feature vectors to different buckets. This ultimately results in an N×N similarity matrix, where N is the number of data samples.

[0094] Based on the obtained similarity matrix, a spectral clustering algorithm is used to cluster cross-domain data. First, an affinity matrix is ​​constructed. This can be directly used as the similarity matrix or by transforming the similarity using a Gaussian kernel function. The Laplacian matrix is ​​then calculated using various methods, such as an undirected normalized Laplacian matrix. The eigenvectors of the Laplacian matrix are then calculated, and the eigenvectors corresponding to the first k smallest eigenvalues ​​are selected to form an N × k matrix. Finally, k-means clustering is performed on each row of this matrix (representing a data point) to obtain the data clustering results.

[0095] Through the above steps, we can establish associations between cross-domain data and achieve heterogeneous data fusion. For example, we can cluster text and image data into the same category and discover their semantic connections. This approach can effectively handle heterogeneous data with different modalities and structures, overcoming the limitations of traditional methods that make it difficult to directly compare heterogeneous data.

[0096] In practical applications, appropriate feature extraction methods and network architectures can be selected based on specific tasks. For example, for text data, a pre-trained BERT (Bidirectional Encoder Representations from Transformers) model can be used as a feature extractor; for image data, a pre-trained CNN (convolutional neural network) model can be used. During domain adversarial network training, the weights of various loss functions must be carefully adjusted to balance domain invariance and discriminability of features. When using the LSH (Locality Sensitive Hashing) algorithm, the number of hash functions and projection dimensions are important hyperparameters that require a trade-off based on data size and computational resources. The number of clusters k in spectral clustering is also a key parameter. The optimal k value can be determined by observing the eigenvalue distribution or using metrics such as the silhouette coefficient.

[0097] This approach can effectively uncover potential connections between cross-domain data, providing crucial support for subsequent data analysis and decision-making. For example, in a recommendation system, users' textual reviews and product images can be correlated to uncover the relationship between user preferences and product visual features, enabling more accurate personalized recommendations. In medical diagnosis, patients' medical records, medical imaging, and genetic data can be integrated to fully understand disease characteristics and assist doctors in making more accurate diagnoses.

[0098] In an optional implementation, graph database technology is used to construct a production process knowledge graph, mapping the fused multi-source heterogeneous data to the entities and relationships of the knowledge graph. Based on a time sequence graph algorithm, dynamic time sequence associations are established in the knowledge graph to form a traceable production process chain, including:

[0099] Design a knowledge graph model and define the entity types and relationship types of the production process knowledge graph. The entity types include products, processes, equipment, raw materials, quality indicators, and operators. The relationship types include contain, use, produce, monitor, operate, and precedent.

[0100] Select and configure a graph database platform, including installing a database server, creating a database instance, configuring memory and storage parameters, and enabling necessary plugins. Next, preprocess the fused multi-source heterogeneous data, including data cleaning, standardization, entity recognition, and relationship extraction.

[0101] The preprocessed data is then imported into the graph database using a graph database query language to create nodes and relationships and build an initial knowledge graph. In this initial knowledge graph, timestamp attributes are added to the nodes and relationships, and "happens after" temporal relationships are created based on the timestamps to establish dynamic temporal associations.

[0102] Based on the established temporal relationship, a temporal traversal algorithm is developed for temporal path query in the knowledge graph; by utilizing the constructed knowledge graph and temporal relationship, forward tracing, reverse tracing and quality problem tracing functions are realized to form a traceable production process chain.

[0103] To implement a knowledge graph-based production process traceability method, the first step is to design a knowledge graph schema. Define the entity types of the production process knowledge graph, including products, processes, equipment, raw materials, quality indicators, and operators. Also define relationship types, including include, use, produce, monitor, operate, and precedent. For example, you can define the "product" entity type as containing attributes such as product ID, name, and specifications; the "process" entity type as containing attributes such as process ID, name, and description; and the "equipment" entity type as containing attributes such as equipment ID, name, and model. Relationship types such as "include" are used to indicate the processes included in a product, "use" represents the equipment and raw materials used in a process, and "produce" represents the product produced by a process.

[0104] Next, select and configure a graph database platform. You can choose Neo4j as the graph database platform, install the Neo4j server, create a database instance named "ProductionKG," and configure it with 8GB of memory and 100GB of storage. Enable the APOC plug-in to support advanced graph algorithms. Preprocess the fused multi-source heterogeneous data, including removing duplicates and outliers, standardizing date formats, identifying entities such as products and processes, and extracting relationships between these entities. For example, identify product entities from order data extracted from the ERP system, identify process and equipment entities from production records extracted from the MES system, and extract the contextual relationships between processes.

[0105] Use the Cypher query language to import the pre-processed data into the Neo4j database. Build an initial knowledge graph containing various entities and relationships through batch import.

[0106] In the initial knowledge graph, add timestamp attributes to nodes and relationships. For example, add start time and end time attributes to the process node:

[0107] MATCH (s:Step {id:'S001'});

[0108] SET s.startTime = '2023-05-01 09:00:00', s.endTime = '2023-05-01 11:00:00';

[0109] Create a "happens after" temporal relationship based on timestamps and establish dynamic temporal associations:

[0110] MATCH (s1:Step), (s2:Step);

[0111] WHERE s1.endTime < s2.startTime;

[0112] CREATE (s1)-[:HAPPENED_BEFORE]->(s2);

[0113] Based on the established temporal relationships, a temporal traversal algorithm is developed for temporal path query. A depth-first search algorithm can be used, starting from the starting node and traversing backward along the "HAPPENED_BEFORE" relationship, recording the traversal path until the termination condition is reached.

[0114] Leveraging the constructed knowledge graph and temporal relationships, we implement forward tracing, reverse tracing, and quality issue tracing. Forward tracing begins with raw materials, traversing forward along the "use" and "production" relationships, tracking the raw material's use process all the way to the final product. Reverse tracing begins with the final product, tracing backward along the "include" and "use" relationships, tracing the product's production process all the way back to the raw materials. Quality issue tracing begins with the problem product, combining forward and reverse tracing to locate the possible source of the problem and analyze the scope of impact.

[0115] This query returns the complete production path from raw materials to the problem product, forming a traceable production process chain. This allows you to quickly locate the source of the problem and assess the scope of impact, effectively supporting quality management and problem resolution.

[0116] In an optional embodiment, based on blockchain technology, tamper-proof digital fingerprints are created for key nodes and transactions in the production process, and the digital fingerprints are associated with corresponding entities in the knowledge graph, including:

[0117] Design and deploy a permissioned blockchain network based on an open-source blockchain distributed ledger, define the network topology, configure channels corresponding to production lines or product types, and write smart contracts for digital fingerprint creation, storage, and verification. Use a multi-hash algorithm to generate digital fingerprints, including combining production process data, timestamps, and random numbers, performing a first hash operation using the SHA256 algorithm, and then performing a second hash operation on the first hash result using the SHA3-256 algorithm to obtain the final digital fingerprint. Implement the digital fingerprint storage function in the smart contract, encapsulating the entity ID, digital fingerprint, and timestamp into a structure, serializing it, and storing it in the blockchain world state.

[0118] Add digital fingerprint attributes to entities in the knowledge graph, create new nodes representing blockchain transactions, and establish the relationship between entity nodes and blockchain transaction nodes; develop a digital fingerprint verification process, first retrieve the digital fingerprint information of the target entity from the knowledge graph, then call the verification function of the blockchain smart contract, pass in the entity ID and digital fingerprint, and finally return the verification result; implement an integrated query interface for the knowledge graph and blockchain, and support simultaneous acquisition of entity information in the knowledge graph and digital fingerprint verification results in the blockchain based on the entity ID.

[0119] The specific implementation of creating tamper-proof digital fingerprints for key nodes and transactions in the production process based on blockchain technology and associating the digital fingerprints with corresponding entities in the knowledge graph is as follows:

[0120] First, design and deploy a permissioned blockchain network based on an open-source blockchain distributed ledger. Define the network topology, including multiple organizations (such as manufacturers, suppliers, and regulators), multiple nodes (such as production line nodes and warehouse nodes), and corresponding channels. For example, in a smart manufacturing production line, separate channels can be configured for different product types (such as mobile phones and televisions) to achieve data isolation. For example, a mobile phone production channel could include nodes from related organizations such as the phone manufacturer and key component suppliers.

[0121] Next, we'll write a smart contract to implement the creation, storage, and verification of digital fingerprints. This smart contract, developed in the Go language, primarily includes the following functions: createFingerprint for creating the digital fingerprint, storeFingerprint for storing the digital fingerprint, and verifyFingerprint for verifying the digital fingerprint. The createFingerprint function first obtains production process data (such as product ID, batch number, production parameters, etc.), the current timestamp, and a random number. This data is then concatenated into a string and the SHA256 algorithm is used to calculate the first hash value. This first hash value is then hashed again using the SHA3-256 algorithm to produce the final digital fingerprint.

[0122] Taking mobile phone production as an example, assume the following production data: product ID is "PHONE001", batch number is "BATCH20230501", production parameters are "{"temperature": 25, "pressure": 1.0}", timestamp is "1682899200", and random number is "123456". After concatenating these data, we get the string:

[0123] "PHONE001BATCH20230501

[0124] {"temperature": 25, "pressure": 1.0}1682899200123456".

[0125] Perform SHA256 hashing on the string to get the first hash value:

[0126] "8f4e8d9a1b3c2d5e6f7a8b9c0d1e2f3a4b5c6d7e8f9a0b1c2d3e4f5a6b7c8d9".

[0127] Then perform SHA3-256 hashing on the first hash value to finally get the digital fingerprint:

[0128] "2a1b3c4d5e6f7a8b9c0d1e2f3a4b5c6d7e8f9a0b1c2d3e4f5a6b7c8d9e0f1".

[0129] In the storeFingerprint function, the entity ID, digital fingerprint, and timestamp are encapsulated into a structure, and then serialized and stored in the blockchain world state.

[0130] In the knowledge graph, digital fingerprint attributes are added to the entity and a new node representing the blockchain transaction is created. This adds digital fingerprint attributes to the product entity in the knowledge graph, creates a new node representing the blockchain transaction, and establishes a relationship between the product entity and the blockchain transaction node.

[0131] Next, we develop the digital fingerprint verification process. First, we retrieve the digital fingerprint information of the target entity from the knowledge graph. This can be achieved through the following Cypher query statement:

[0132] MATCH (p:Product {id: 'PHONE001'});

[0133] RETURN p.digitalFingerprint;

[0134] After obtaining the digital fingerprint, the blockchain smart contract's verifyFingerprint function is called for verification. In the verifyFingerprint function, the stored digital fingerprint data is first retrieved from the world state based on the entity ID, and then compared with the passed digital fingerprint. If a match is found, a verification success result is returned; otherwise, a verification failure result is returned.

[0135] Finally, we implement an integrated query interface for the knowledge graph and blockchain. We develop a RESTful API that supports obtaining both entity information in the knowledge graph and digital fingerprint verification results in the blockchain based on the entity ID. The API process is as follows:

[0136] 1. Receive an API request containing an entity ID.

[0137] 2. Query the knowledge graph to obtain the attribute information of the entity.

[0138] 3. Obtain the digital fingerprint of the entity from the knowledge graph.

[0139] 4. Call the verifyFingerprint function of the blockchain smart contract to verify the digital fingerprint.

[0140] 5. Combine the knowledge graph query results and blockchain verification results and return them to the client.

[0141] Through the above steps, we have achieved a complete process of creating tamper-proof digital fingerprints for key nodes and transactions in the production process based on blockchain technology, and then associating these digital fingerprints with corresponding entities in the knowledge graph. This approach combines the tamper-proof nature of blockchain with the semantic association capabilities of the knowledge graph, providing strong guarantees for data authenticity and traceability in the intelligent manufacturing process.

[0142] In an optional implementation, a multi-agent collaborative decision-making model is established using a reinforcement learning algorithm. This model can adaptively allocate tasks and resources based on the current production status and historical experience. A sequence-to-sequence model enhanced with an attention mechanism is used to transform unstructured collaborative requirements into executable workflows, including:

[0143] A multi-agent deep deterministic policy gradient model based on the actor-critic architecture is constructed, in which each agent represents a workstation or piece of equipment on the production line. Each agent consists of an actor network for generating actions and a critic network for evaluating the value of actions. A state space is defined, which includes the current workstation's task queue, resource utilization, and the states of neighboring workstations, as well as an action space that includes task allocation decisions and resource scheduling strategies. A reward function is designed that comprehensively considers production efficiency, resource utilization, and task completion quality.

[0144] Utilizing the multi-agent deep deterministic policy gradient model, selecting the optimal action for each workstation based on the observed current state, and performing corresponding task allocation and resource scheduling operations; calculating the reward value after executing the action, and using the gained experience to update the multi-agent deep deterministic policy gradient model;

[0145] Build an attention-enhanced sequence-to-sequence model consisting of a bidirectional LSTM (Long Short-Term Memory) encoder, an attention layer, and a unidirectional LSTM decoder. Convert the unstructured collaboration requirement text into a word sequence and encode it. Input the word sequence into the attention-enhanced sequence-to-sequence model to generate the corresponding workflow sequence. Decode the generated workflow sequence into a readable workflow description.

[0146] According to the decoded workflow description, the updated multi-agent deep deterministic policy gradient model is used for task allocation and resource scheduling. During the execution of the production process, the latest production status information and historical experience data are continuously obtained, and the multi-agent deep deterministic policy gradient model is dynamically updated to achieve adaptive task allocation and resource scheduling optimization.

[0147] According to the described method, a multi-agent deep deterministic policy gradient model based on an actor-critic architecture is first constructed. Each agent in this model represents a workstation or piece of equipment on a production line and includes an actor network for generating actions and a critic network for evaluating the value of actions. In specific implementations, deep neural networks can be used as function approximators for both the actor and critic. For example, the actor network can be a three-layer fully connected network, with the input layer containing the same number of neurons as the state space dimension, the hidden layer containing 64 neurons, and the output layer containing the same number of neurons as the action space dimension. The critic network also uses a three-layer fully connected network, with the input layer containing the state and action, the hidden layer containing 64 neurons, and the output layer containing a single neuron representing the state-action value function.

[0148] Next, we define the state space and action space. The state space contains the current workstation's task queue, resource utilization, and neighboring workstation status. For example, the task queue can be represented by a vector of length 10, with each element representing the number of tasks of a certain type; resource utilization can be represented by a scalar between 0 and 1; and the neighboring workstation status can be represented by a vector of the same dimension as the current workstation. The action space contains task allocation decisions and resource scheduling strategies and can be designed as a discrete action space, for example, containing a combination of 10 predefined task allocation schemes and 5 resource scheduling strategies.

[0149] Then, a reward function is designed that comprehensively considers production efficiency, resource utilization, and task completion quality. The reward function can be defined as a weighted sum of these three metrics, with the weights adjusted based on actual production needs. For example, production efficiency can be expressed as the number of tasks completed per unit time, resource utilization can be expressed as the average utilization of various resources, and task completion quality can be expressed as the pass rate.

[0150] Using a constructed multi-agent deep deterministic policy gradient model, the optimal action is selected for each workstation based on the observed current state. Specifically, the current workstation state information is first obtained and input into the corresponding agent's actor network, resulting in a deterministic action output. This action is then executed, performing the corresponding task allocation and resource scheduling operations. For example, if the selected action is to assign three Class A tasks to workstation number one and increase resource allocation by 50%, workstation number one's task queue and resource allocation will be updated accordingly.

[0151] After executing an action, the reward is calculated. Based on the previously defined reward function, the productivity, resource utilization, and task completion quality after the action are calculated, and a composite reward is derived based on the set weights. Information such as the state before the action, the action executed, the reward obtained, and the new state after the action are stored as a single piece of experience data. This collected experience data is used to update the multi-agent deep deterministic policy gradient model. The update process involves updating the critic network to reduce the TD error, and then using the policy gradient method to update the actor network to maximize the expected reward.

[0152] Next, we built an attention-enhanced sequence-to-sequence model to transform unstructured collaboration requirements into actionable workflows. This model consists of a bidirectional LSTM encoder, an attention layer, and a unidirectional LSTM decoder. The encoder uses a bidirectional LSTM to capture contextual information about the input sequence, while the decoder uses a unidirectional LSTM to predict the next output based on the generated output sequence. The attention layer dynamically focuses on different parts of the input sequence during the decoding process.

[0153] Unstructured collaborative requirements are converted into word sequences and encoded. For example, a requirement like "100 Class A products and 50 Class B products need to be produced within 24 hours" can be converted into a word sequence like ["need", "in", "24", "hours", "complete", "100", "Class A", "products", and", "50", "Class B", "products", "of", "production"]. A pre-trained word embedding model (such as Word2Vec) is then used to convert the word sequence into a vector sequence, which serves as the encoder input.

[0154] The encoded vector sequence is fed into the attention-enhanced sequence-to-sequence model to generate the corresponding workflow sequence. During decoding, the model predicts the next output word based on the partial output generated and the context vector calculated by the attention mechanism. The generated workflow sequence might look like ["assign","50","pieces","Category A","task","give","Workstation 1","assign","25","pieces","Category B","task","give","Workstation 2",...].

[0155] Decode the generated workflow sequence into a readable workflow description. This step can be achieved through a simple template filling method, filling the key information from the generated sequence into a predefined workflow description template. For example, a workflow description such as "First, assign 50 Class A tasks to workstation number one, then assign 25 Class B tasks to workstation number two,..." can be obtained.

[0156] Based on the decoded workflow description, the updated multi-agent deep deterministic policy gradient model is used for task allocation and resource scheduling. Specifically, each task allocation step in the workflow is converted into a decision cycle of the model. In each decision cycle, the optimal action is selected based on the current state, the corresponding task allocation and resource scheduling operations are executed, and the system state is updated.

[0157] During production execution, the latest production status information and historical experience data are continuously collected to dynamically update the multi-agent deep deterministic policy gradient model. A fixed update cycle can be set, for example, after every 100 decision cycles. During updates, the critic network and actor network are updated using the newly collected experience data, following the previously described method. In this way, the model can continuously adapt to the changing production environment, achieving adaptive task allocation and resource scheduling optimization.

[0158] In an optional implementation, a transaction trigger mechanism is constructed based on a knowledge graph and anomaly detection model to automatically identify production transactions that require collaborative processing. The resulting transaction collaboration records include:

[0159] Design a multi-layer autoencoder network structure as an anomaly detection model, train the autoencoder using normal production data, calculate the reconstruction error, and set a threshold for anomaly determination; collect sensor data and production status information from the production system in real time, and input the collected data into the trained anomaly detection model for analysis;

[0160] When an anomaly is detected, context analysis is performed using the knowledge graph to determine the impact and possible causes of the anomaly. Based on the anomaly and the knowledge graph analysis results, a decision is made as to whether collaborative processing needs to be initiated. For transactions requiring collaborative processing, a transaction collaboration record is generated containing a timestamp, anomaly information, affected equipment, affected processes, and collaboration requirements.

[0161] The generated transaction collaboration records are input into the multi-agent collaborative decision-making system for task allocation and resource scheduling; the knowledge graph and anomaly detection model are continuously updated during the production process, and the anomaly detection threshold and collaborative processing strategy are dynamically adjusted; the knowledge graph is supplemented and corrected based on the collaborative processing results and production effects, and the anomaly detection model is incrementally learned using newly added anomaly samples.

[0162] In order to build a transaction trigger mechanism based on knowledge graph and anomaly detection model to automatically identify production transactions that need to be collaboratively processed, the following specific implementation methods can be adopted:

[0163] First, a multi-layer autoencoder network structure was designed as an anomaly detection model. This autoencoder consists of an input layer, multiple hidden layers, and an output layer. The number of neurons in the input layer is equal to the total number of sensors and state variables in the production system, for example, 100 neurons. The hidden layers are three layers, with 80, 50, and 80 neurons, respectively. The output layer has the same number of neurons as the input layer. Reluctant Lu (ReLU) is selected as the activation function, and mean squared error is used as the loss function.

[0164] Next, train the autoencoder using normal production data. Filter out normal operating data from the production system's historical data for a period of time, for example, the last three months. Divide this data into a training set and a validation set with an 80% to 20% split. Train the autoencoder using the stochastic gradient descent algorithm, setting a learning rate of 0.001, a batch size of 64, and 100 training epochs. Calculate the reconstruction error on the validation set, using the 99.9% quantile as the threshold for abnormality detection, for example, 0.05.

[0165] Next, sensor data and production status information are collected from the production system in real time. Data can be collected every minute, including sensor values ​​such as temperature, pressure, and flow, as well as information such as equipment operating status and production parameters. The collected data is normalized and fed into the trained anomaly detection model to calculate the reconstruction error. If the reconstruction error exceeds a preset threshold of 0.05, it is considered an anomaly.

[0166] When an anomaly is detected, the system leverages a pre-built knowledge graph for contextual analysis to determine the anomaly's scope and possible causes. The knowledge graph includes entities such as equipment, processes, parameters, and faults, as well as the relationships between them. For example, if a temperature sensor's value is abnormal, the knowledge graph can quickly locate the device where the sensor is located, the associated processes, and the type of fault that may have caused the temperature anomaly. The system then tracks the anomaly's propagation path and assesses its potential impact on downstream processes and product quality.

[0167] Based on the anomaly and knowledge graph analysis results, the system determines whether collaborative processing is necessary. Criteria include the severity of the anomaly, the scope of impact, and whether collaboration between multiple departments is required. For transactions requiring collaborative processing, the system automatically generates a transaction collaboration record containing the following information: the timestamp of the occurrence, the anomaly sensor or device information, the anomaly value, the scope of impact, possible causes, affected processes, and a recommended collaborative processing solution.

[0168] For example, if temperature sensor A detects an abnormal temperature rise of 95°C (normal range is 60-80°C), the transaction collaboration record generated by the system may be:

[0169] {

[0170] "Timestamp": "2023-06-01 10:30:15",

[0171] "Abnormal device": "Temperature sensor A",

[0172] "Outlier": "95°C",

[0173] "Affected Scope": ["Reactor No. 1", "Cooling System No. 1"],

[0174] "Possible causes": ["Cooling system failure", "Reaction out of control"],

[0175] "Affected Process": ["Polymerization", "Product Cooling"],

[0176] "Collaboration Requirements": ["Process Engineer", "Equipment Maintenance Personnel", "Quality Control Personnel"]

[0177] }.

[0178] The resulting transaction collaboration records are fed into a multi-agent collaborative decision-making system, which is responsible for task allocation and resource scheduling. The collaborative decision-making system assigns tasks to appropriate individuals or intelligent agents based on factors such as the urgency of the transaction and the required expertise. For example, in the case of the aforementioned abnormal temperature event, the system might simultaneously notify a process engineer to assess the response, dispatch maintenance personnel to inspect the cooling system, and assign quality control personnel to monitor changes in product quality.

[0179] During the production process, the system continuously updates the knowledge graph and anomaly detection model. After each collaborative process, relevant experiences and solutions are added to the knowledge graph, enriching the knowledge base for anomaly handling. Simultaneously, the system regularly (e.g., weekly) uses newly added production data to incrementally learn the anomaly detection model to adapt to changes in the production environment. Furthermore, the system dynamically adjusts anomaly detection thresholds based on actual production conditions. For example, if a high false alarm rate is detected for a particular anomaly, the threshold for that type of anomaly can be appropriately increased.

[0180] Finally, the system supplements and amends the knowledge graph based on the results of collaborative processing and production outcomes. For example, if a collaborative process successfully resolves a problem, the relevant processing steps and experience are added to the knowledge graph. If it is discovered that certain pre-specified fault causes do not match the actual situation, the relationships in the knowledge graph are revised. Furthermore, the system uses these newly added anomaly samples for incremental learning in the anomaly detection model, continuously improving the model's accuracy and generalization capabilities.

[0181] Through the above steps, this method can effectively build a transaction trigger mechanism, realize timely discovery and intelligent collaborative processing of production anomalies, and thus improve the stability and efficiency of the production system.

[0182] Figure 2 FIG. 1 is a structural diagram of the production traceability and transaction collaboration system of the MES system according to an embodiment of the present invention. Figure 2 As shown, the system includes:

[0183] The first unit is used to deploy a distributed data acquisition network to collect heterogeneous data in real time from production equipment, environmental sensors, human-computer interaction terminals, and enterprise information systems; classify and annotate the collected heterogeneous data to establish a unified data model architecture; use edge computing nodes to perform preliminary cleaning, denoising, and standardization on the heterogeneous data to generate a structured data stream; transmit the structured data stream to the central data processing unit, use deep learning algorithms to extract data features and understand semantics, and construct a multidimensional data feature vector; based on the multidimensional data feature vector, use transfer learning methods to establish a cross-domain data association model to achieve semantic alignment and fusion of heterogeneous data;

[0184] The second unit uses graph database technology to build a production process knowledge graph, mapping the fused multi-source heterogeneous data to the entities and relationships of the knowledge graph. Based on a time-series graph algorithm, dynamic time-series associations are established in the knowledge graph to form a traceable production process chain. Based on blockchain technology, tamper-proof digital fingerprints are created for key nodes and transactions in the production process and associated with the corresponding entities in the knowledge graph. An anomaly detection model based on a graph neural network is constructed to monitor the data flow and relationship changes in the knowledge graph in real time, identifying potential production anomalies and quality risks.

[0185] The third unit is used to build a transaction trigger mechanism based on knowledge graphs and anomaly detection models, automatically identify production transactions that require collaborative processing, and obtain transaction collaboration records; use reinforcement learning algorithms to establish a multi-agent collaborative decision-making model, which can adaptively allocate tasks and resources based on current production status and historical experience; through a sequence-to-sequence model enhanced by an attention mechanism, unstructured collaboration requirements are converted into executable workflows; based on traceable production process chains and transaction collaboration records, multi-dimensional production analysis reports are automatically generated.

[0186] According to a third aspect of the embodiments of the present invention,

[0187] An electronic device is provided, comprising:

[0188] processor;

[0189] a memory for storing processor-executable instructions;

[0190] The processor is configured to call the instructions stored in the memory to execute the aforementioned method.

[0191] According to a fourth aspect of the embodiments of the present invention,

[0192] A computer-readable storage medium is provided, on which computer program instructions are stored. When the computer program instructions are executed by a processor, the method described above is implemented.

[0193] The present invention may be a method, an apparatus, a system and / or a computer program product. The computer program product may include a computer-readable storage medium carrying computer-readable program instructions for executing various aspects of the present invention.

[0194] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the above embodiments, or replace some or all of the technical features therein with equivalents. However, these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. The production traceability and transaction collaboration method of the MES system is characterized by: include: Deploy a distributed data collection network to collect heterogeneous data in real time from production equipment, environmental sensors, human-computer interaction terminals, and enterprise information systems; Classify and annotate the collected heterogeneous data and establish a unified data model architecture; Use edge computing nodes to perform preliminary cleaning, denoising, and standardization on heterogeneous data to generate structured data streams; Transmit the structured data stream to the central data processing unit, use the deep learning algorithm to extract data features and semantic understanding, and construct a multi-dimensional data feature vector; based on the multi-dimensional data feature vector, use the transfer learning method to establish a cross-domain data association model to achieve semantic alignment and fusion of heterogeneous data; Using graph database technology, we build a knowledge graph of the production process and map the fused multi-source heterogeneous data to the entities and relationships of the knowledge graph. Based on the time-series graph algorithm, dynamic time-series associations are established in the knowledge graph to form a traceable production process chain; Based on blockchain technology, we create tamper-proof digital fingerprints for key nodes and transactions in the production process, and associate the digital fingerprints with the corresponding entities in the knowledge graph. We build an anomaly detection model based on graph neural networks to monitor the data flow and relationship changes in the knowledge graph in real time, and identify potential production anomalies and quality risks. Based on the knowledge graph and anomaly detection model, a transaction trigger mechanism is built to automatically identify production transactions that need to be processed collaboratively and obtain transaction collaboration records; Using reinforcement learning algorithms, a multi-agent collaborative decision-making model is established, which can adaptively allocate tasks and resources based on current production status and historical experience; Through the sequence-to-sequence model enhanced by the attention mechanism, unstructured collaboration requirements are transformed into executable workflows; based on the traceable production process chain and transaction collaboration records, multi-dimensional production analysis reports are automatically generated.

2. The method according to claim 1, characterized in that Based on the multi-dimensional data feature vector, a transfer learning method is used to establish a cross-domain data association model to achieve semantic alignment and fusion of heterogeneous data, including: Use transfer learning methods to align feature vectors from different domains, including: Select feature vectors of the source domain and the target domain; construct a domain adversarial neural network, including a feature extractor; train the domain adversarial neural network to minimize the difference in feature distribution between the source domain and the target domain; use the trained feature extractor to obtain the aligned feature vectors; The cosine similarity is used to calculate the similarity between the aligned feature vectors, and the local sensitive hashing algorithm is used to accelerate the similarity calculation between large-scale vectors to obtain the similarity matrix; Based on the similarity matrix, a spectral clustering algorithm is used to cluster the cross-domain data, including: Construct an affinity matrix; calculate a Laplace matrix; calculate the eigenvector of the Laplace matrix; perform k-means clustering on the eigenvector, establish associations between data, and achieve heterogeneous data fusion.

3. The method according to claim 1, characterized in that: Using graph database technology, we build a knowledge graph of the production process and map the fused multi-source heterogeneous data to the entities and relationships of the knowledge graph. Based on the time series graph algorithm, dynamic time series associations are established in the knowledge graph to form a traceable production process chain including: Design a knowledge graph model and define the entity types and relationship types of the production process knowledge graph. The entity types include products, processes, equipment, raw materials, quality indicators, and operators. The relationship types include inclusion, use, production, monitoring, operation, and predecessor. Select and configure the graph database platform, including installing the database server, creating a database instance, configuring memory and storage parameters, and enabling necessary plug-ins; then pre-process the fused multi-source heterogeneous data, including data cleaning, standardization, entity recognition, and relationship extraction; Then, the preprocessed data is imported into the graph database using the graph database query language to create nodes and relationships and construct an initial knowledge graph. In the constructed initial knowledge graph, timestamp attributes are added to nodes and relationships, and a "happens after" temporal relationship is created based on the timestamp to establish a dynamic temporal association. Based on the established temporal relationship, a temporal traversal algorithm is developed for temporal path query in the knowledge graph. By utilizing the constructed knowledge graph and temporal relationship, forward tracing, reverse tracing and quality problem tracing functions are realized to form a traceable production process chain.

4. The method according to claim 1, characterized in that: Based on blockchain technology, tamper-proof digital fingerprints are created for key nodes and transactions in the production process, and the digital fingerprints are associated with the corresponding entities in the knowledge graph, including: Design and deploy a permissioned blockchain network based on an open source blockchain distributed ledger, define the network topology, configure channels corresponding to production lines or product types, and write smart contracts for digital fingerprint creation, storage, and verification; use a multi-hash algorithm to generate digital fingerprints, including combining production process data, timestamps, and random numbers, first performing a first hash operation using the SHA256 algorithm, and then performing a second hash operation on the first hash result using the SHA3-256 algorithm to obtain the final digital fingerprint; implement the digital fingerprint storage function in the smart contract, encapsulate the entity ID, digital fingerprint, and timestamp into a structure, and store them in the blockchain world state after serialization; Add digital fingerprint attributes to entities in the knowledge graph, create new nodes representing blockchain transactions, and establish the relationship between entity nodes and blockchain transaction nodes; develop a digital fingerprint verification process, first retrieve the digital fingerprint information of the target entity from the knowledge graph, then call the verification function of the blockchain smart contract, pass in the entity ID and digital fingerprint, and finally return the verification result; implement an integrated query interface for the knowledge graph and blockchain, and support simultaneous acquisition of entity information in the knowledge graph and digital fingerprint verification results in the blockchain based on the entity ID.

5. The method according to claim 1, characterized in that Using reinforcement learning algorithms, a multi-agent collaborative decision-making model is established, which can adaptively allocate tasks and resources based on current production status and historical experience; The sequence-to-sequence model enhanced by the attention mechanism transforms unstructured collaborative requirements into executable workflows including: Construct a multi-agent deep deterministic policy gradient model based on the Actor-Critic architecture, where each agent represents a workstation or equipment on the production line. Each agent includes an actor network for generating actions and a critic network for evaluating the value of actions. Define a state space containing the current workstation task queue, resource utilization, and the status of adjacent workstations, as well as an action space containing task allocation decisions and resource scheduling strategies. Design a reward function that comprehensively considers production efficiency, resource utilization, and task completion quality. Using the multi-agent deep deterministic policy gradient model, select the optimal action for each workstation according to the observed current state, and perform corresponding task allocation and resource scheduling operations; calculate the reward value after executing the action, and use the gained experience to update the multi-agent deep deterministic policy gradient model; Construct an attention mechanism-enhanced sequence-to-sequence model including a bidirectional LSTM long short-term memory encoder, an attention layer, and a unidirectional LSTM decoder; convert the unstructured collaborative requirement text into a word sequence and encode it, input it into the attention mechanism-enhanced sequence-to-sequence model, and generate the corresponding workflow sequence; decode the generated workflow sequence into a readable workflow description; According to the decoded workflow description, the updated multi-agent deep deterministic policy gradient model is used for task allocation and resource scheduling. During the execution of the production process, the latest production status information and historical experience data are continuously obtained, and the multi-agent deep deterministic policy gradient model is dynamically updated to achieve adaptive task allocation and resource scheduling optimization.

6. The method according to claim 1, characterized in that Based on the knowledge graph and anomaly detection model, a transaction trigger mechanism is built to automatically identify production transactions that need to be processed collaboratively. The transaction collaboration records obtained include: Design a multi-layer autoencoder network structure as an anomaly detection model, use normal production data to train the autoencoder, calculate the reconstruction error and set a threshold for anomaly determination; collect sensor data and production status information from the production system in real time, and input the collected data into the trained anomaly detection model for analysis; When an anomaly is detected, the knowledge graph is used for context analysis to determine the impact range and possible causes of the anomaly. Based on the anomaly and the results of the knowledge graph analysis, it is determined whether collaborative processing needs to be initiated. For transactions that require collaborative processing, a transaction collaborative record containing timestamps, anomaly information, affected equipment, affected processes, and collaborative requirements is generated. The generated transaction collaboration records are input into the multi-agent collaborative decision-making system for task allocation and resource scheduling. The knowledge graph and anomaly detection model are continuously updated during the production process, and the anomaly detection threshold and collaborative processing strategy are dynamically adjusted. The knowledge graph is supplemented and corrected based on the collaborative processing results and production effects, and the anomaly detection model is incrementally learned using the newly added abnormal samples.

7. A production traceability and transaction collaboration system of an MES system, used to implement the method described in any one of claims 1 to 6, characterized in that: include: The first unit is used to deploy a distributed data acquisition network to collect heterogeneous data in real time from production equipment, environmental sensors, human-computer interaction terminals, and enterprise information systems; Classify and annotate the collected heterogeneous data and establish a unified data model architecture; Use edge computing nodes to perform preliminary cleaning, denoising, and standardization on heterogeneous data to generate structured data streams; Transmit the structured data stream to the central data processing unit, use the deep learning algorithm to extract data features and semantic understanding, and construct a multi-dimensional data feature vector; based on the multi-dimensional data feature vector, use the transfer learning method to establish a cross-domain data association model to achieve semantic alignment and fusion of heterogeneous data; The second unit is used to build a knowledge graph of the production process using graph database technology, and map the fused multi-source heterogeneous data to the entities and relationships of the knowledge graph; Based on the time-series graph algorithm, dynamic time-series associations are established in the knowledge graph to form a traceable production process chain; Based on blockchain technology, we create tamper-proof digital fingerprints for key nodes and transactions in the production process, and associate the digital fingerprints with the corresponding entities in the knowledge graph. We build an anomaly detection model based on graph neural networks to monitor the data flow and relationship changes in the knowledge graph in real time, and identify potential production anomalies and quality risks. The third unit is used to build a transaction trigger mechanism based on the knowledge graph and anomaly detection model, automatically identify production transactions that need to be processed collaboratively, and obtain transaction collaboration records; Using reinforcement learning algorithms, a multi-agent collaborative decision-making model is established, which can adaptively allocate tasks and resources based on current production status and historical experience; Through the sequence-to-sequence model enhanced by the attention mechanism, unstructured collaboration requirements are transformed into executable workflows; based on the traceable production process chain and transaction collaboration records, multi-dimensional production analysis reports are automatically generated.

8. An electronic device, characterized in that: include: processor; a memory for storing processor-executable instructions; The processor is configured to call the instructions stored in the memory to execute the method according to any one of claims 1 to 6.

9. A computer-readable storage medium having computer program instructions stored thereon, characterized in that: When the computer program instructions are executed by a processor, the method according to any one of claims 1 to 6 is implemented.

Citation Information

Patent Citations

  • Parameter adjustment method and device, electronic equipment and storage medium

    CN115660312A

  • Auditing knowledge mapping system based on digital auditing platform

    CN117932074A

  • Method and system for implementing suite mechanism based on low-code platform

    CN118312167A

  • Power grid fault power protection maintenance plan making method, device, equipment and medium

    CN118504939A

  • Office cooperation system and method based on artificial intelligence

    CN118521272A

Cited By

  • Production process data set construction method and system for food quality traceability edge calculation

    CN119313361A

  • Data exception traceability tracking and positioning method, system and device and storage medium

    CN119416131A

  • Model optimization method and system based on large and small model collaborative intelligence

    CN119476445A

  • Big data driven network operation state monitoring management method and system

    CN119675988A

  • Application system integration method and system based on machine intelligence

    CN119760237A