Coal preparation plant expert database system based on browser / server B / S architecture

Abstract

The invention discloses a coal preparation plant expert database system based on a browser / server B / S architecture. The coal preparation plant expert database system comprises a data acquisition and integration module, an expert knowledge base construction module, an intelligent analysis and decision support module, a B / S architecture module and a system integration and deployment module. The data utilization efficiency is improved. An expert knowledge base is successfully constructed, and precious expert experience is effectively inherited and utilized. And through intelligent analysis and decision support, the production process is optimized, and the production efficiency and the product quality are improved. And due to the application of the B / S architecture, the system has good cross-platform compatibility and usability. The application of the micro-service and containerization technology improves the maintainability and expandability of the system.

Description

Technical Field

[0001] The present invention belongs to the technical field of coal industry informatization, and in particular relates to a coal preparation plant expert database system based on a browser / server B / S architecture. Background Art

[0002] The traditional coal preparation plant operation model has the following problems:

[0003] Data silos: Information is not interoperable between various production systems and management systems, resulting in inefficient decision-making.

[0004] Dependence on experience: Key production knowledge and experience often rely on a small number of experts, making them difficult to systematically inherit and apply.

[0005] Passive maintenance: Equipment maintenance is mainly based on regular inspections, which makes it difficult to prevent sudden failures and affect production continuity.

[0006] Difficulty in optimization: Process parameter optimization mainly relies on manual experience and is difficult to adapt to complex and changing production conditions.

[0007] Response lag: Traditional management systems are unable to reflect production status in real time, affecting the timeliness and accuracy of decision-making.

[0008] To solve these problems, an integrated and intelligent expert database system is needed that can comprehensively collect and analyze various data of the coal preparation plant and provide intelligent decision support. Summary of the Invention

[0009] Purpose of the invention: The purpose of the present invention is to provide a coal preparation plant expert database system based on browser / server B / S architecture.

[0010] Technical solution: The present invention provides a coal preparation plant expert database system based on a browser / server B / S architecture, comprising a data acquisition and integration module, an expert knowledge base construction module, an intelligent analysis and decision support module, a B / S architecture module, and a system integration and deployment module;

[0011] The data collection and integration module accesses the data source and collects structured data and unstructured data; the data collection and integration module cleans and processes abnormal data on the collected data and performs data conversion on the processed data;

[0012] The expert knowledge base construction module includes a knowledge acquisition module, a knowledge representation module and a knowledge storage module;

[0013] The intelligent analysis and decision support module includes a data mining module, a prediction model module and an optimization algorithm module;

[0014] The B / S architecture module includes a front-end design module, a back-end implementation module and a data interaction module;

[0015] The system integration and deployment module includes a database integration module, a security mechanism module and a containerized deployment module.

[0016] Furthermore, the data acquisition and integration module adopts a standardized RESTful API interface to connect various subsystems within the coal preparation plant, including the intelligent sorting system, the coal slurry water treatment system, and the intelligent loading system; and adopts the OAuth 2.0 protocol for identity authentication and authorization; the data acquisition and integration module uses Java Database Connectivity technology to connect to the relational databases of various systems in the coal preparation plant, including Oracle and MySQL; the data acquisition and integration module processes the surveillance camera images of the coal preparation plant through the OpenCV library, and uses the FFmpeg library to collect and process the audio data of the equipment operation for abnormal sound detection.

[0017] Furthermore, the data cleaning is to use the verification rules written in the Java regular expression library to build a verification framework based on Apache Commons Validator to centrally manage and apply various verification rules; the abnormal data processing builds an anomaly detection model based on the Isolation Forest algorithm to identify abnormal values in coal quality data and equipment operating parameters; the local outlier factor LOF algorithm is used to detect abnormal points in multidimensional data, and the ARIMA model is used to predict equipment operating parameters to identify abnormal data that deviates from the predicted values; the data conversion of the processed data includes ETL process implementation and unstructured data processing; the ETL process implementation uses Apache NiFi to build a visual ETL process to realize the extraction, conversion and loading of various data of the coal preparation plant; the unstructured data processing uses TensorFlow to implement a deep learning-based image recognition model to identify equipment status and safety hazards from monitoring images; the Apache Tika library is used for document parsing to extract text content from various documents.

[0018] Furthermore, the knowledge acquisition module includes an expert experience collection module and a document knowledge extraction module; the expert experience collection module is based on an intelligent question-answering system of a decision tree, and dynamically generates follow-up questions according to the expert's answers; the document knowledge extraction module uses Python's NLTK library to implement a keyword extraction algorithm, identifies important concepts from the technical documents of the coal preparation plant, and automatically discovers and extracts subject knowledge from the documents based on the topic model of LDA latent Dirichlet allocation; the knowledge representation module includes a knowledge graph construction module and a decision logic representation module, the knowledge graph construction module uses the Protégé tool to design the coal preparation field ontology model, adopts the RDF resource description framework format to represent knowledge triples, and supports OWL reasoning through the Apache Jena reasoning engine; the knowledge storage module includes a graph database application module and a relational database usage module, the graph database application module uses the Neo4j graph database to store and manage the coal preparation field knowledge graph, and the knowledge retrieval interface based on the Cypher query language supports graph traversal and pattern matching; the relational database usage module uses the PostgreSQL database to store structured expert rules and parameters, and implements object-relational mapping based on the MyBatis ORM layer.

[0019] Furthermore, the data mining module includes an equipment operation mode analysis module and a production parameter association analysis module; the equipment operation mode analysis module performs equipment operation status clustering analysis based on the K-means++ algorithm to identify the typical operation mode of coal preparation plant equipment; uses the DBSCAN density clustering algorithm to detect abnormal operation modes and identify potential equipment failures, and uses the t-SNE algorithm to map high-dimensional equipment operation data to a 2D plane; the production parameter association analysis module applies the Apriori algorithm to mine association rules between coal preparation process parameters, mines association rules based on the FP-growth algorithm, and uses the PageRank algorithm to identify key influencing factors; the prediction model module includes an equipment failure prediction module and a product quality prediction module; the equipment failure prediction module uses the ARIMA autoregressive integrated moving average model to analyze equipment vibration time series data of movement and temperature; a deep learning model based on the LSTM long short-term memory network predicts the failure probability of complex equipment; a multi-parameter fault prediction model based on random forest takes into account factors such as equipment operating time, load, and environment; the product quality prediction module uses a deep neural network DNN to construct a coal quality prediction model, predicts the quality of clean coal products based on the characteristics of the raw coal entering the washing process and process parameters, predicts the coal slime water treatment effect based on the XGBoost algorithm, and optimizes the use of flocculants; an image recognition model based on the convolutional neural network CNN predicts ash and moisture indicators through raw coal images; the optimization algorithm module includes a production parameter optimization module and a control strategy optimization module, the production parameter optimization module uses a genetic algorithm to optimize the coal preparation process parameters; the control strategy optimization module uses a deep reinforcement learning algorithm to optimize the coal preparation plant automatic control system to achieve dynamic adjustment of the control strategy.

[0020] Furthermore, the front-end design module includes a responsive interface construction module and a Vue.js single-page application implementation module; the responsive interface construction module uses HTML5 semantic tags to build the page structure, and adopts CSS3 Flexbox and Grid layout to implement page typesetting; the Vue.js single-page application implementation module adopts Vue CLI to build the project scaffolding, and uses Vue Router to implement front-end routing; the back-end implementation module includes a RESTfulAPI construction module and a microservice architecture module; the RESTfulAPI construction module uses the Spring Boot framework to build the back-end service, and adopts Spring Data JPA to simplify the data access layer; the microservice architecture module adopts the Spring CloudNetflix suite to implement the microservice architecture; the data interaction module includes a real-time data push module and an identity authentication module; the real-time data push module uses WebSocket technology to realize real-time data push from the server to the client; the identity authentication module uses Redis to cache the token information of the logged-in user, and supports integration with the enterprise's existing identity system based on the third-party authentication interface of OAuth 2.0.

[0021] Furthermore, the database integration module includes an ORM framework application module and a database connection pool module; the ORM framework application module uses the Hibernate ORM framework to implement object-relational mapping and uses Hibernate's secondary cache mechanism; the database connection pool module adopts HikariCP connection pool technology to implement connection pool monitoring and uses Java Management Extensions to expose key indicators; the security mechanism module includes an access control module and a data encryption module; the access control module uses role-based access control RBAC, defines roles such as administrator, engineer, and operator, and uses Spring Security's expressive security control to achieve fine-grained method-level permission management; the data encryption module uses the SSL / TLS protocol to encrypt all communications between the client and the server, and uses the AES encryption algorithm to protect stored key business data; the containerized deployment module includes a Docker application module and a Kubernetes management module; the Docker application module uses Docker Compose to orchestrate multi-container applications; the Kubernetes management module uses Kubernetes for container orchestration, configures the Horizontal Pod Autoscaler, and automatically adjusts the number of Pods according to CPU usage or custom indicators.

[0022] Beneficial effects: Compared with the prior art, the present invention has the following significant advantages:

[0023] (1) Intelligent fusion of multi-source heterogeneous data: The innovative combination of machine learning algorithms (such as anomaly detection) and semantic analysis technology realizes the intelligent cleaning and fusion of multi-source heterogeneous data in coal preparation plants.

[0024] (2) Dynamically evolving knowledge graph: A dynamic knowledge graph construction method based on a graph database (Neo4j) was designed, which can automatically extract knowledge from documents and expert questions and answers, and continuously update and expand the knowledge base.

[0025] (3) Collaborative application of deep learning and reinforcement learning: Deep learning is innovatively used for product quality prediction, and deep reinforcement learning is used to optimize control strategies, achieving a close combination of prediction and control.

[0026] (4) Adaptive optimization algorithm: An adaptive optimization strategy was developed that can automatically adjust the optimization objectives and constraints according to the characteristics of different coal types, significantly improving the applicability and effectiveness of the optimization algorithm.

[0027] (5) Real-time data processing under microservice architecture: Innovatively integrate a real-time data push mechanism based on WebSocket into the microservice architecture, combined with containerization technology, to achieve a high-performance and scalable real-time data processing framework.

[0028] (6) This invention solves the problem of heterogeneous data integration: by developing standardized RESTful APIs and WebService interfaces, it achieves unified data collection from various subsystems within the coal preparation plant (such as the intelligent sorting system and the coal slurry water treatment system). At the same time, the use of ETL tools and natural language processing technology solves the problem of integrating structured and unstructured data.

[0029] (7) The present invention solves the problem of large-scale real-time data processing: Apache NiFi is used to build a visual ETL process, and WebSocket technology is combined to achieve real-time data push, which solves the problem of real-time processing and transmission of large-scale production data.

[0030] (8) The present invention solves the problem of expert knowledge representation and reasoning: by constructing a knowledge graph based on the ontology model and combining it with the Drools rule engine, the formal representation and automatic reasoning of complex coal preparation expert knowledge are realized.

[0031] (9) The present invention solves intelligent prediction and optimization problems: it applies deep learning models (such as LSTM) to predict equipment failures, and uses genetic algorithms and deep reinforcement learning algorithms to optimize production parameters and control strategies, solving complex prediction and optimization problems that are difficult to handle with traditional methods.

[0032] (10) The present invention solves the system scalability and performance problems: it adopts microservice architecture and containerization technology (Docker+Kubernetes) to solve the problems of poor system scalability and complex deployment, and at the same time improves system performance through database connection pool and caching mechanism. BRIEF DESCRIPTION OF THE DRAWINGS

[0033] Figure 1 It is a structural schematic diagram of the present invention. DETAILED DESCRIPTION

[0034] The technical solution of the present invention is further described below.

[0035] The present invention adopts the following technical solution: a coal preparation plant expert database system based on B / S architecture, implemented according to the following contents:

[0036] S1, data collection and integration module

[0037] 1.1 Data Source Access

[0038] a) Subsystem interface integration

[0039] Develop a standardized RESTful API interface to connect various subsystems within the coal preparation plant, such as the intelligent sorting system, coal slime water treatment system, intelligent loading system, etc.

[0040] Implement WebService based on SOAP protocol to ensure compatibility with traditional systems.

[0041] The OAuth 2.0 protocol is used for identity authentication and authorization to ensure the security of data transmission.

[0042] b) Structured data collection

[0043] Use JDBC (Java Database Connectivity) technology to directly connect to the relational databases of various systems in the coal preparation plant, such as Oracle, MySQL, etc.

[0044] Implement a scheduled task scheduler (such as Quartz) to regularly extract structured data from production control systems, business management systems, etc.

[0045] Adopting an incremental synchronization strategy, only synchronizing data that has changed since the last synchronization improves efficiency.

[0046] c) Unstructured data collection

[0047] Develop a specialized image acquisition module to process surveillance camera images from a coal preparation plant using the OpenCV library for equipment status monitoring and safety management.

[0048] Use the FFmpeg library to collect and process audio data from device operation for abnormal sound detection.

[0049] Implement file system monitoring and automatically collect new unstructured data such as text documents and Excel spreadsheets.

[0050] 1.2 Data Cleaning

[0051] a) Data format verification

[0052] Use the Java regular expression library (java.util.regex) to write complex verification rules, such as format checks for coal quality parameters and equipment numbers.

[0053] Implement custom data validation annotations for data validation at the JavaBean level.

[0054] Develop a validation framework based on Apache Commons Validator to centrally manage and apply various validation rules.

[0055] b) Abnormal data processing

[0056] Implement an anomaly detection model based on the Isolation Forest algorithm to identify outliers in coal quality data and equipment operating parameters.

[0057] The local outlier factor (LOF) algorithm is used to detect outliers in multidimensional data (such as multiple coal quality index combinations).

[0058] Develop an anomaly detection module based on time series analysis, use the ARIMA model to predict equipment operating parameters, and identify abnormal data that deviates from the predicted values.

[0059] Implement data repair strategies, such as using moving averages or interpolation to repair missing data and smooth outliers.

[0060] 1.3 Data Conversion

[0061] a) ETL process implementation

[0062] Use Apache NiFi to build a visual ETL process to extract, transform, and load various types of data in the coal preparation plant.

[0063] Develop custom data converters to handle specific data formats, such as unifying coal quality data from different sources into a standard format.

[0064] Realize data aggregation functions, such as calculating the average coal quality indicators and equipment operating efficiency per hour.

[0065] b) Unstructured data processing

[0066] Use TensorFlow to implement a deep learning-based image recognition model to identify device status, safety hazards, etc. from monitoring images.

[0067] Apache Tika library is used for document parsing to extract text content from various documents.

[0068] Implement a text analysis pipeline based on NLTK (Natural Language Toolkit) to perform word segmentation, part-of-speech tagging, named entity recognition and other processing.

[0069] S2. Expert knowledge base construction module

[0070] 2.1 Knowledge Acquisition

[0071] a) Expert experience collection

[0072] Design a structured online questionnaire system and set professional questions on coal preparation technology, equipment maintenance, safety management and other aspects.

[0073] Implement an intelligent question-answering system based on decision trees, dynamically generate follow-up questions based on expert answers, and deeply explore expert knowledge.

[0074] Develop a mobile app to facilitate experts to record and upload their experience and knowledge at any time.

[0075] b) Document knowledge extraction

[0076] The keyword extraction algorithm is implemented using Python's NLTK library to identify important concepts from the technical documents of the coal preparation plant.

[0077] The Stanford NER (Named Entity Recognition) tool is used to identify professional terms such as device names, parameter indicators, etc. in documents.

[0078] Implement a topic model based on LDA (Latent Dirichlet Allocation) to automatically discover and extract topic knowledge from a large number of documents.

[0079] 2.2 Knowledge Representation

[0080] a) Knowledge graph construction

[0081] Use Protégé tool to design coal preparation domain ontology model and define core concepts, relationships and attributes.

[0082] The RDF (Resource Description Framework) format is used to represent knowledge triples, such as "flotation machine-processing-coal slime water".

[0083] Implement an inference engine based on Apache Jena to support OWL (Web Ontology Language) reasoning and expand implicit knowledge.

[0084] b) Decision logic representation

[0085] Use the Drools rule engine to represent complex coal preparation decision logic, such as "if the slime concentration > X and the pH value < Y, then increase the dosage of reagent Z".

[0086] Implement a process parameter optimization model based on decision trees to guide the adjustment of coal preparation processes.

[0087] Develop a control rule system based on fuzzy logic to handle uncertain factors in the coal preparation process.

[0088] 2.3 Knowledge storage

[0089] a) Application of graph databases

[0090] Use the Neo4j graph database to store and manage the knowledge graph in the field of coal preparation.

[0091] Implement a knowledge retrieval interface based on the Cypher query language to support complex graph traversal and pattern matching.

[0092] Develop a knowledge graph visualization module to intuitively display the association relationships between knowledge.

[0093] b) Use of relational databases

[0094] Adopt the PostgreSQL database to store structured expert rules and parameters.

[0095] Implement a database partitioning strategy to partition data according to knowledge domains (such as processes, equipment, safety, etc.) to improve query efficiency.

[0096] Develop an ORM layer based on MyBatis to achieve object-relational mapping and simplify data access operations.

[0097] S3, Intelligent analysis and decision support module

[0098] 3.1 Data mining

[0099] a) Analysis of equipment operation modes

[0100] Implement a clustering analysis of equipment operation status based on the K-means++ algorithm to identify the typical operation modes of key equipment in coal preparation plants (such as flotation machines, centrifuges).

[0101] Use the DBSCAN (density clustering) algorithm for detecting abnormal operation modes to identify potential equipment failures.

[0102] An equipment energy consumption analysis module based on hierarchical clustering was developed to classify and evaluate the energy consumption characteristics of various equipment in a coal preparation plant.

[0103] Implement a visual analysis tool that uses the t-SNE algorithm to map high-dimensional device operation data to a 2D plane, intuitively displaying the distribution of operation modes.

[0104] b) Production parameter correlation analysis

[0105] The Apriori algorithm is used to mine association rules between coal preparation process parameters, such as the relationship between "increase in ash content of raw coal entering washing" and "increase in flotation agent dosage".

[0106] Implement efficient association rule mining based on FP-growth algorithm to process large-scale production data.

[0107] Develop a parameter association network based on a graph model and use the PageRank algorithm to identify key influencing factors.

[0108] Implement a dynamic association rule update mechanism to automatically adjust rule confidence and support over time.

[0109] 3.2 Prediction Model

[0110] a) Equipment failure prediction

[0111] The ARIMA (Autoregressive Integrated Moving Average) model is used to analyze time series data such as equipment vibration and temperature to predict potential failures.

[0112] Implement a deep learning model based on LSTM (Long Short-Term Memory Network) to predict the failure probability of complex equipment (such as heavy medium cyclones).

[0113] Develop a multi-parameter fault prediction model based on random forest, taking into account factors such as equipment operating time, load, and environment.

[0114] Implement online learning mechanism to continuously optimize the prediction model and improve prediction accuracy.

[0115] b) Product quality prediction

[0116] A coal quality prediction model was constructed using a deep neural network (DNN) to predict the quality of clean coal products based on the characteristics of the raw coal entering the washing process and process parameters.

[0117] Realize the prediction of coal slime water treatment effect based on XGBoost algorithm and optimize the use of flocculants.

[0118] Develop an image recognition model based on convolutional neural network (CNN) to predict ash content, moisture and other indicators through raw coal images.

[0119] Implement ensemble learning methods to combine multiple basic models (such as SVM, decision tree, neural network) to improve prediction stability.

[0120] 3.3 Optimization Algorithm

[0121] a) Production parameter optimization

[0122] Genetic algorithms are used to optimize coal preparation process parameters, such as flotation reagent ratio and heavy medium suspension concentration.

[0123] Implement equipment operating parameter adjustment based on particle swarm optimization (PSO) algorithm, such as centrifuge speed and feed rate optimization.

[0124] Develop multi-objective optimization models that simultaneously consider factors such as product quality, production costs, and energy consumption.

[0125] Implement adaptive optimization strategy and automatically adjust optimization objectives and constraints according to the characteristics of different coal types.

[0126] b) Control strategy optimization

[0127] Use deep reinforcement learning algorithms (such as DQN) to optimize the automatic control system of coal preparation plants and realize dynamic adjustment of control strategies.

[0128] It realizes continuous control optimization based on policy gradient, which is suitable for continuous process control such as flotation and jigging.

[0129] Develop a multi-agent reinforcement learning system to coordinate and optimize the control strategies of multiple coal preparation units.

[0130] Implement a simulation environment to perform large-scale policy exploration and evaluation using the Monte Carlo Tree Search (MCTS) algorithm.

[0131] S4, B / S architecture implementation

[0132] 4.1 Front-end Design

[0133] a) Responsive interface construction

[0134] Use HTML5 semantic tags to build a clear page structure, such as <header> 、 <nav> 、 <main>wait.

[0135] Use CSS3 Flexbox and Grid layout to achieve flexible page layout and adapt to device screens of different sizes.

[0136] Implement CSS3 media queries to provide optimized display effects for mobile, tablet and desktop.

[0137] Use JavaScript ES6+ features, such as arrow functions and destructuring assignment, to improve code readability and development efficiency.

[0138] b) Vue.js single page application implementation

[0139] Use Vue CLI to build project scaffolding and standardize project structure and construction process.

[0140] Use Vue Router to implement front-end routing and switch between different functional modules (such as data analysis, device monitoring, knowledge base, etc.) without refreshing.

[0141] Implement Vuex state management and centrally handle complex data flows, such as user authentication status, global configuration, etc.

[0142] Develop custom Vue components, such as data visualization charts, real-time monitoring panels, etc., to improve code reusability.

[0143] 4.2 Backend Implementation

[0144] a) RESTful API construction

[0145] Use the Spring Boot framework to quickly build backend services and use its automatic configuration features to simplify development.

[0146] Implement a RESTful API based on Spring MVC, and design URL structures and HTTP methods that conform to REST principles.

[0147] Use Spring Data JPA to simplify the data access layer and implement the Repository pattern.

[0148] Use Swagger UI to automatically generate API documentation to facilitate collaboration between front-end and back-end developers.

[0149] b) Microservices Architecture

[0150] The Spring CloudNetflix suite is used to implement the microservice architecture, including Eureka service registration and discovery, Zuul gateway, etc.

[0151] Implement Docker-based microservice container deployment to improve system portability and scalability.

[0152] Use Spring Cloud Config to implement centralized configuration management, facilitating dynamic adjustment of system parameters.

[0153] Implement a Hystrix-based service circuit breaking and degradation mechanism to improve system availability.

[0154] 4.3 Data Interaction

[0155] a) Real-time data push

[0156] Use WebSocket technology to push real-time data from the server to the client, such as device status updates, alarm information, etc.

[0157] Implements a message subscription mechanism based on STOMP (Simple Text Oriented Messaging Protocol), allowing clients to selectively receive data.

[0158] SockJS is used as a fallback option for WebSocket to ensure that similar functions can be achieved in environments that do not support WebSocket.

[0159] b) Identity Authentication

[0160] Implement a token refresh mechanism to balance security and user experience.

[0161] Use Redis to cache the token information of logged-in users to achieve fast verification and single sign-on.

[0162] Develop a third-party authentication interface based on OAuth 2.0 to support integration with the company's existing identity system.

[0163] S5. System Integration and Deployment

[0164] 5.1 Database Integration

[0165] a) ORM framework application

[0166] Use Hibernate ORM framework to implement object-relational mapping and support multiple databases such as MySQL, Oracle, PostgreSQL, etc.

[0167] Implement JPA (Java Persistence API) specifications to improve code portability.

[0168] Use Hibernate's secondary cache mechanism to optimize the query performance of frequently accessed data.

[0169] Realize dynamic SQL generation to meet the complex data query needs of coal preparation plants.

[0170] b) Database connection pool

[0171] Adopt HikariCP connection pool technology to optimize database connection management and improve system throughput.

[0172] Implement connection pool monitoring and expose key metrics using JMX (Java Management Extensions).

[0173] Configure automatic tuning of connection pool parameters to dynamically adjust the number of connections based on system load.

[0174] Implement support for multiple data sources and flexibly handle connection requirements of different databases.

[0175] 5.2 Security Mechanism

[0176] a) Access Control

[0177] Implement role-based access control (RBAC) and define roles such as administrator, engineer, and operator.

[0178] Use Spring Security's expressive security control to implement fine-grained method-level permission management.

[0179] Develop a dynamic permission assignment mechanism that allows adjusting user permissions at runtime.

[0180] Implement operation log recording to track the access and modification history of key data.

[0181] b) Data encryption

[0182] All client-server communications are encrypted using SSL / TLS protocols.

[0183] Implement one-way encrypted storage of sensitive data (such as user passwords) using algorithms such as bcrypt.

[0184] AES encryption algorithm is used to protect stored critical business data.

[0185] Implement data desensitization and hide sensitive information in non-essential scenarios.

[0186] 5.3 Containerized Deployment

[0187] a)Docker Application

[0188] Create an independent Docker image for each microservice component (such as data collection service, analysis service, etc.).

[0189] Use Docker Compose to orchestrate multi-container applications and simplify deployment in development and test environments.

[0190] Implement a Docker-based continuous integration / continuous deployment (CI / CD) process to automate the build and deployment process.

[0191] Optimize the size of Docker images and use multi-stage builds to reduce the final image size.

[0192] b) Kubernetes Management

[0193] Use Kubernetes for container orchestration to achieve elastic scaling of the coal preparation plant expert system.

[0194] Configure the Horizontal PodAutoscaler to automatically adjust the number of pods based on CPU usage or custom metrics.

[0195] Implement Ingress-based intelligent routing to optimize external access traffic distribution.

[0196] Use Kubernetes Secrets to manage sensitive configuration information such as database passwords, API keys, etc.

[0197] Deploy Prometheus and Grafana to implement monitoring and alerting in containerized environments.

[0198] Through the detailed technical solution described above, the coal preparation plant expert database system, based on a B / S architecture, not only provides powerful data analysis and decision support capabilities, but also boasts high scalability, security, and ease of maintenance. The system's front-end interface is responsive, and back-end processing is efficient, meeting the complex and ever-changing business needs of coal preparation plants and providing strong support for their intelligent and digital transformation.

[0199] This invention achieves unified management and deep mining of various types of data in a coal preparation plant, improving data utilization efficiency. It successfully builds an expert knowledge base, effectively inheriting and leveraging valuable expert experience. Through intelligent analysis and decision support, it optimizes production processes, improving production efficiency and product quality. The application of a B / S architecture ensures excellent cross-platform compatibility and ease of use. The application of microservices and containerization technologies enhances the system's maintainability and scalability.< / main> < / nav> < / header>

Claims

1. A coal preparation plant expert database system based on browser / server B / S architecture, characterized in that: Including data collection and integration module, expert knowledge base construction module, intelligent analysis and decision support module, B / S architecture module, system integration and deployment module; The data collection and integration module accesses the data source and collects structured data and unstructured data; the data collection and integration module cleans and processes abnormal data on the collected data and performs data conversion on the processed data; The expert knowledge base construction module includes a knowledge acquisition module, a knowledge representation module and a knowledge storage module; The intelligent analysis and decision support module includes a data mining module, a prediction model module and an optimization algorithm module; The B / S architecture module includes a front-end design module, a back-end implementation module and a data interaction module; The system integration and deployment module includes a database integration module, a security mechanism module and a containerized deployment module.

2. A coal preparation plant expert database system based on browser / server B / S architecture according to claim 1, characterized in that: The data acquisition and integration module uses a standardized RESTful API interface to connect to various subsystems within the coal preparation plant, including the intelligent sorting system, the coal slurry water treatment system, and the intelligent loading system; and adopts the OAuth 2.0 protocol for identity authentication and authorization; the data acquisition and integration module uses Java Database Connectivity technology to connect to the relational databases of various systems in the coal preparation plant, including Oracle and MySQL; the data acquisition and integration module processes the coal preparation plant's surveillance camera images through the OpenCV library, and uses the FFmpeg library to collect and process audio data from equipment operation for abnormal sound detection.

3. The coal preparation plant expert database system based on browser / server B / S architecture according to claim 1 is characterized in that: The data cleaning is to use the validation rules written in the Java regular expression library to build a validation framework based on Apache Commons Validator to centrally manage and apply various validation rules; the abnormal data processing is to build an anomaly detection model based on the Isolation Forest algorithm to identify abnormal values in coal quality data and equipment operating parameters; The local outlier factor (LOF) algorithm is used to detect outliers in multidimensional data, and the ARIMA model is used to predict equipment operating parameters and identify abnormal data that deviates from the predicted values. The data conversion of the processed data includes ETL process implementation and unstructured data processing. The ETL process implementation uses Apache NiFi to build a visual ETL process to realize the extraction, conversion and loading of various types of data in the coal preparation plant. The unstructured data processing uses TensorFlow to implement a deep learning-based image recognition model to identify equipment status and safety hazards from monitoring images. The Apache Tika library is used for document parsing to extract text content from various documents.

4. The coal preparation plant expert database system based on browser / server B / S architecture according to claim 1, characterized in that: The knowledge acquisition module includes an expert experience collection module and a document knowledge extraction module; the expert experience collection module is based on an intelligent question-answering system of a decision tree, and dynamically generates follow-up questions according to the expert's answers; the document knowledge extraction module uses Python's NLTK library to implement a keyword extraction algorithm, identifies important concepts from the technical documents of the coal preparation plant, and automatically discovers and extracts subject knowledge from the documents based on the LDA latent Dirichlet allocation topic model; the knowledge representation module includes a knowledge graph construction module and a decision logic representation module. The knowledge graph construction module uses the Protégé tool to design an ontology model for the coal preparation field, adopts the RDF resource description framework format to represent knowledge triples, and supports OWL reasoning through the Apache Jena reasoning engine; the knowledge storage module includes a graph database application module and a relational database usage module. The graph database application module uses the Neo4j graph database to store and manage the coal preparation field knowledge graph, and uses the knowledge retrieval interface based on the Cypher query language to support graph traversal and pattern matching; the relational database usage module uses the PostgreSQL database to store structured expert rules and parameters, and implements object-relational mapping based on the MyBatis ORM layer.

5. The coal preparation plant expert database system based on browser / server B / S architecture according to claim 1, characterized in that: The data mining module includes an equipment operation mode analysis module and a production parameter association analysis module; the equipment operation mode analysis module performs equipment operation status clustering analysis based on the K-means++ algorithm to identify the typical operation mode of coal preparation plant equipment; uses the DBSCAN density clustering algorithm to detect abnormal operation modes, identifies potential equipment failures, and uses the t-SNE algorithm to map high-dimensional equipment operation data to a 2D plane; the production parameter association analysis module applies the Apriori algorithm to mine association rules between coal preparation process parameters, mines association rules based on the FP-growth algorithm, and uses the PageRank algorithm to identify key influencing factors; the prediction model module includes an equipment failure prediction module and a product quality prediction module; the equipment failure prediction module uses the ARIMA autoregressive integrated moving average model to analyze equipment vibration, Time series data of temperature; a deep learning model based on the LSTM long short-term memory network to predict the failure probability of complex equipment; a multi-parameter fault prediction model based on random forest, taking into account factors such as equipment operating time, load, and environment; the product quality prediction module uses a deep neural network DNN to build a coal quality prediction model, predicts the quality of clean coal products based on the characteristics of the raw coal entering the washing process and process parameters, predicts the coal slime water treatment effect based on the XGBoost algorithm, and optimizes the use of flocculants; an image recognition model based on the convolutional neural network CNN predicts ash and moisture indicators through raw coal images; the optimization algorithm module includes a production parameter optimization module and a control strategy optimization module, the production parameter optimization module uses a genetic algorithm to optimize the coal preparation process parameters; the control strategy optimization module uses a deep reinforcement learning algorithm to optimize the coal preparation plant automatic control system to achieve dynamic adjustment of the control strategy.

6. The coal preparation plant expert database system based on browser / server B / S architecture according to claim 1, characterized in that: The front-end design module includes a responsive interface construction module and a Vue.js single-page application implementation module; the responsive interface construction module uses HTML5 semantic tags to build the page structure and adopts CSS3 Flexbox and Grid layout to realize page typesetting; the Vue.js single-page application implementation module adopts Vue CLI to build the project scaffolding and uses VueRouter to realize the front-end routing; the back-end implementation module includes a RESTful API construction module and a microservice architecture module; the RESTful API construction module uses the Spring Boot framework to build the back-end service and adopts Spring Data JPA to simplify the data access layer; the microservice architecture module adopts the Spring Cloud Netflix suite to realize the microservice architecture; the data interaction module includes a real-time data push module and an identity authentication module; the real-time data push module uses WebSocket technology to realize real-time data push from the server to the client; the identity authentication module uses Redis to cache the token information of the logged-in user and supports integration with the enterprise's existing identity system based on the third-party authentication interface of OAuth 2.

0.

7. The coal preparation plant expert database system based on browser / server B / S architecture according to claim 1, characterized in that: The database integration module includes an ORM framework application module and a database connection pool module; the ORM framework application module uses the Hibernate ORM framework to implement object-relational mapping and uses Hibernate's secondary cache mechanism; the database connection pool module adopts HikariCP connection pool technology to implement connection pool monitoring and uses Java Management Extensions to expose key indicators; the security mechanism module includes an access control module and a data encryption module; the access control module uses role-based access control RBAC, defines roles such as administrator, engineer, and operator, and uses Spring Security's expressive security control to achieve fine-grained method-level permission management; the data encryption module uses the SSL / TLS protocol to encrypt all communications between the client and the server, and uses the AES encryption algorithm to protect stored key business data; the containerized deployment module includes a Docker application module and a Kubernetes management module; the Docker application module uses Docker Compose to orchestrate multi-container applications; the Kubernetes management module uses Kubernetes for container orchestration, configures the Horizontal Pod Autoscaler, and automatically adjusts the number of pods according to CPU usage or custom indicators.

Images