Steel production process standard intelligent management system and method based on rule engine

Abstract

The application discloses a kind of steel production process standard intelligent management system and method based on rule engine, belong to steel production control technical field.System includes the full production factor data acquisition module, intelligent rule analysis engine, abnormal classification early warning module and dynamic closed-loop management module connected in turn.Communications corresponding method includes four steps of full production factor data acquisition, intelligent rule analysis, abnormal classification early warning and dynamic closed-loop management.The application realizes process standard automatic unified execution by semantic process standard rule base, completes abnormal risk classification and accurate push based on multi-attribute decision model, and forms "monitoring-early warning-disposal-optimization" full closed loop;Solve the problem of abnormal response lag, process standard execution is not unified and post-control in prior art, realize production whole process initiative intelligent control, improve steel production quality stability and operation efficiency.

Description

Technical Field

[0001] This invention relates to the field of steel production control technology, and in particular to an intelligent governance system and method for steel production process standards based on a rule engine. Background Technology

[0002] In the entire steel production process, the monitoring and management of production process parameters (such as steelmaking temperature and continuous casting speed), safety risk indicators (such as gas concentration), environmental emission data (such as dust concentration), and equipment monitoring status (such as bearing vibration and motor current) currently mainly rely on manual inspection and manual recording.

[0003] This model has significant drawbacks: First, when abnormal events occur, the lengthy transmission process makes it impossible to detect them immediately, resulting in a large average response delay and potentially exacerbating accidents or causing quality fluctuations. Second, the lack of unified standards in process implementation leads to inconsistencies in standard execution due to differences in operator experience, resulting in unstable product quality. Third, the control methods rely excessively on post-event processing, with lagging data collection and analysis, making it impossible to achieve pre-event warnings and in-event intervention, leading to frequent process quality problems and safety and environmental incidents. Summary of the Invention

[0004] To address the technical problems in existing steel production caused by reliance on manual labor, such as untimely acquisition of abnormal information, inconsistent implementation of process standards, and excessive dependence on post-event management, this invention provides a rule-engine-based intelligent governance system and method for steel production process standards. This system enables real-time perception and accurate early warning of multi-dimensional data on production processes, safety risks, environmental emissions, and equipment monitoring, ensuring efficient and unified implementation of process standards and proactive control throughout the entire process.

[0005] The technical solution adopted by the intelligent governance system and method for steel production process standards based on rule engine of this invention is as follows: The rule-based intelligent governance system for steel production process standards includes a data acquisition module for all production elements, an intelligent rule analysis engine, an anomaly classification and early warning module, and a dynamic closed-loop management module, all connected in sequence. The data acquisition module collects multi-source heterogeneous production data from the entire steel production process in real time. The intelligent rule analysis engine has a built-in semantic process standard rule library, which automatically compares the real-time collected production data with standardized thresholds in the rule library to determine process compliance and identify abnormal events. The anomaly classification and early warning module quantifies the risk level of identified abnormal events based on a multi-attribute decision model and accurately pushes the classification and early warning information to the corresponding responsible positions according to the dynamic mapping relationship of role permissions. The dynamic closed-loop management module records the entire handling process of abnormal events and feeds the handling results back to the intelligent rule analysis engine, driving the dynamic iterative optimization of the semantic process standard rule library.

[0006] A further improvement of the technical solution of the present invention is that: the full production factor data acquisition module includes a multi-protocol industrial IoT gateway, an edge computing unit, and a time-series database; the multi-protocol industrial IoT gateway is used to connect to field controllers, sensors, and smart meters to realize the seamless acquisition of production process, safety risk, environmental emission, and equipment status data; the edge computing unit is used to preprocess and synchronize the acquired data in time; and the time-series database is used to store the preprocessed production data without loss.

[0007] A further improvement of the technical solution of the present invention is that the semantic process standard rule base stores structured steel industry process standards, safety specifications and environmental protection requirements, including process parameter thresholds, safety risk thresholds, environmental emission thresholds and equipment operating status thresholds, and supports business personnel to manage the entire life cycle of rules through a visual interface.

[0008] A further improvement of the technical solution of the present invention is that the evaluation dimensions of the multi-attribute decision model include the severity of the abnormal event, the scope of impact on production, and the priority of emergency response, and the abnormal event is divided into three risk levels: mild, moderate, and severe.

[0009] A further improvement of the technical solution of the present invention is that the anomaly classification early warning module supports multiple push methods such as industrial control terminal pop-ups, WeChat for enterprises, work groups and SMS, and different risk levels correspond to different push frequencies and reminder intensities.

[0010] A further improvement of the technical solution of the present invention is that: the dynamic closed-loop management module includes a handling process tracking unit, a result verification unit, and a rule optimization unit; the handling process tracking unit records the entire handling process information of abnormal events; the result verification unit performs quantitative evaluation of the handling effect; and the rule optimization unit generates rule base adjustment suggestions based on the evaluation results.

[0011] The intelligent governance method for steel production process standards based on a rule engine, applying the aforementioned system, includes the following steps: S1. Data collection of all production factors: Real-time collection of production process, safety risks, environmental emissions and equipment status data of the entire steel production process through industrial Internet of Things technology; S2, Intelligent Rule Analysis: Automatically compares the collected real-time data with the built-in semantic process standard rule library to complete the process compliance judgment and identify abnormal events; S3. Anomaly Classification Early Warning: Based on a multi-attribute decision model, the risk level of abnormal events is quantitatively classified, and the classification early warning information is accurately pushed to the corresponding responsible positions according to the dynamic mapping relationship of role permissions. S4. Dynamic closed-loop management: Record the entire handling process of abnormal events, verify and evaluate the handling effect, and feed the evaluation results back to the semantic process standard rule base to realize dynamic iterative optimization of rules.

[0012] A further improvement of the above technical solution of the present invention is that: in step S1, the collected raw data is first cleaned, deduplicated, format converted and time-series synchronized preprocessed by the edge computing unit, and then transmitted to the time-series database for lossless storage, with a data acquisition frequency of 1 time / second.

[0013] A further improvement of the above technical solution of the present invention is that: in step S3, mild anomalies are only pushed to the field operator, moderate anomalies are pushed to the operator and process supervisor, and severe anomalies are pushed to the operator, process supervisor and middle and senior management personnel.

[0014] A further improvement of the above technical solution of the present invention is that: in step S4, the rule optimization unit performs statistical analysis on historical processing data every month, automatically identifies rules with excessively high false alarm rates or missed alarms, generates threshold adjustment suggestions, and completes the rule base update after manual confirmation.

[0015] The technological advancements achieved by this invention due to the adoption of the above technical solutions are as follows: This invention achieves automatic and unified execution of process standards through a semantic process standard rule base, eliminating differences in human experience, ensuring consistent quality across processes, and improving product qualification rates; it establishes a real-time anomaly perception and hierarchical early warning mechanism, shortening the average response time for abnormal events and preventing the escalation of accidents; it effectively reduces quality defects and cost waste caused by human error; it promotes the transformation of production management from passive post-event handling to proactive pre-event prevention, achieving closed-loop management throughout the entire process; and it strengthens the real-time monitoring capabilities of safety risks and environmental emissions, reducing the incidence of safety and environmental incidents and ensuring production safety and compliance levels. Detailed Implementation

[0016] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to specific embodiments. In the following description, descriptions of well-known structures and technologies are omitted to avoid unnecessarily obscuring the concepts of this invention. Example 1

[0017] This embodiment discloses an intelligent governance system for steel production process standards based on a rule engine, which is deployed on an artificial intelligence training and promotion platform. The various modules of the system communicate with each other via Ethernet.

[0018] This embodiment includes a full production factor data acquisition module, an intelligent rule analysis engine, an anomaly classification and early warning module, and a dynamic closed-loop management module that are connected in sequence via communication.

[0019] One data acquisition server is deployed in each of the key production processes, including steelmaking, continuous casting, and rolling. Each server is equipped with a multi-protocol industrial IoT gateway and edge computing software. The multi-protocol industrial IoT gateway is connected to thermocouple temperature sensors and diffused silicon pressure sensors via RS485 interface, and to PLC controllers, infrared gas detectors, laser dust analyzers, and motor vibration monitors via Ethernet. It collects data in real time, such as the final temperature of steelmaking, continuous casting speed, converter gas concentration, sintering machine head dust emission concentration, main motor current, and bearing vibration acceleration. The data acquisition frequency is 1 time per second.

[0020] The edge computing unit preprocesses the raw data: it uses the 3σ criterion to remove outliers exceeding the physical range, performs millisecond-level time-series synchronization of data from different devices based on the NTP network time protocol, converts the data into a unified JSON format, and then transmits it to the cloud-based time-series database via the MQTT protocol. The time-series database uses InfluxDB 2.0, deployed on a 3-node server cluster, supporting millions of data points written per second. The data is stored using the LZ4 compression algorithm and has a retention period of 5 years.

[0021] In this embodiment, the intelligent rule analysis engine has a built-in semantic process standard rule library. Process experts break down documents such as the "Iron and Steel Metallurgical Industry Process Standard Specification" into more than 1,200 key parameters and thresholds, and use DRL language to convert them into executable rules. For example, "When the final temperature of the steelmaking converter is lower than 1600℃ and the duration exceeds 5 minutes, it is determined to be a moderately abnormal process parameter." The rule library establishes a version management mechanism to support rule rollback and comparison operations.

[0022] The intelligent rule analysis engine reads the latest production data every second and compares it in parallel with the rule base. The response time for a single rule judgment does not exceed 100ms. When an abnormal event is detected, the abnormal parameter name, abnormal value, standard threshold, occurrence time, device number, and location are encapsulated into a JSON object and sent to the abnormal classification and early warning module.

[0023] The anomaly classification and early warning module uses a multi-attribute decision model to quantify the risk level of identified abnormal events and accurately pushes the classified early warning information to the corresponding responsible positions based on the dynamic mapping relationship of role permissions. The weights of the multi-attribute decision model are set as follows: anomaly severity 60%, production impact range 30%, and emergency response priority 10%. For example, if the converter gas concentration reaches 200ppm (100% exceeding the standard, severity 90 points), affecting the entire converter workshop (impact range 80 points), and requires immediate evacuation of personnel (emergency priority 100 points), the weighted score is 88 points, which is judged as a severe risk.

[0024] Based on the role and permission mapping relationship: minor anomalies (overall score < 60 points) are pushed to the industrial control terminal of the on-site operator; moderate anomalies (60 points ≤ overall score < 80 points) are pushed to the WeChat of the operator and process supervisor; severe anomalies (overall score ≥ 80 points) are pushed to the WeChat of the operator, process supervisor, workshop director and safety and environmental protection manager and SMS, and trigger an audible and visual alarm of 85 decibels or higher on the industrial control terminal.

[0025] After an anomaly is reported, the system automatically generates a unique work order and assigns it to the corresponding personnel. The work order status is displayed as "Pending Handling". After receiving the work order, the personnel handling the incident fill in the handling measures and start time, and submit an acceptance application upon completion. The work order status is then updated to "Pending Acceptance". The process supervisor conducts acceptance within 2 hours. If the anomaly is confirmed to be eliminated, the work order is marked as "Completed"; otherwise, it is returned for re-handling.

[0026] The results verification unit compares production data one hour before and after the intervention, calculates parameter recovery time, fluctuation range, and pass rate, and quantifies the intervention effect. The rule optimization unit compiles historical data monthly, and automatically generates adjustment suggestions when a certain parameter threshold causes the monthly false alarm rate to exceed 10%. After review and confirmation by process experts, the system automatically updates the rule base and generates a change log. Example 2

[0027] This embodiment provides a smart governance method for steel production process standards based on a rule engine. The specific steps are as follows: S1. Data collection of all production factors: Through multi-protocol industrial IoT gateways deployed in various processes, more than 1,200 production data items such as steelmaking end temperature and continuous casting speed are collected in real time at a frequency of 1 time / second; the raw data is preprocessed by the edge computing unit and then transmitted to the time-series database for storage.

[0028] S2. Intelligent rule analysis: The intelligent rule analysis engine reads the latest production data every second, compares it with the semantic process standard rule library, and automatically determines the process compliance; when the final temperature of the steelmaking converter is lower than 1600℃ and lasts for 5 minutes, it is identified as a moderate abnormal event.

[0029] S3. Anomaly Classification Early Warning: The multi-attribute decision model calculates the comprehensive score of the anomaly to be 72 points, which is judged as a medium risk. The system will push the early warning information to the enterprise WeChat of the on-site steelmaking operator and the steelmaking process supervisor, and trigger the audible and visual alarm of the industrial control terminal.

[0030] S4. Dynamic Closed-Loop Management: The system automatically generates a disposal work order and assigns it to the operator. The operator adjusts the oxygen supply flow of the converter to increase the temperature and submits it for acceptance after completion. The process supervisor accepts and confirms that the temperature has returned to normal and marks the work order as completed. The rule optimization unit subsequently collects statistics on this type of abnormal disposal data and optimizes the temperature threshold and duration parameters.

[0031] In the above embodiments, a rule engine-based intelligent governance system and method for steel production process standards is provided. This invention achieves automatic and unified execution of process standards through a semantic process standard rule base, eliminating differences in human experience, ensuring consistency in quality across processes, and improving product qualification rate; it establishes a real-time anomaly perception and hierarchical early warning mechanism, shortening the average response time for abnormal events and preventing the escalation of accidents; it effectively reduces quality defects and cost waste caused by human error; it promotes the transformation of production control from passive post-event processing to proactive pre-event prevention, achieving closed-loop management throughout the entire process; and it strengthens the real-time monitoring capabilities of safety risks and environmental emissions, reducing the incidence of safety and environmental incidents and ensuring production safety and compliance levels.

[0032] The embodiments described above are merely preferred embodiments of the present invention and are not intended to limit the concept and scope of the present invention. Various modifications and improvements made to the technical solutions of the present invention by those skilled in the art without departing from the inventive concept should fall within the protection scope of the present invention. All technical contents for which protection is sought in this invention are fully described in the claims.

Claims

1. A rule-based intelligent management system for steel production process standards, characterized by: It includes a data acquisition module for all production factors, an intelligent rule analysis engine, an anomaly classification and early warning module, and a dynamic closed-loop management module that are connected in sequence. The full production element data acquisition module is used to collect multi-source heterogeneous production data of the entire steel production process in real time; the intelligent rule analysis engine has a built-in semantic process standard rule library, which is used to automatically compare the real-time collected production data with the standardized thresholds in the rule library to complete the process compliance judgment and abnormal event identification. The anomaly classification and early warning module quantifies the risk level of identified abnormal events based on a multi-attribute decision model, and accurately pushes the classification and early warning information to the corresponding responsible positions according to the dynamic mapping relationship of role permissions. The dynamic closed-loop management module is used to record the entire handling process of abnormal events and feed the handling results back to the intelligent rule analysis engine to drive the dynamic iterative optimization of the semantic process standard rule base.

2. The intelligent management system for steel production process standards based on a rule engine as described in claim 1, characterized in that: The full production factor data acquisition module includes a multi-protocol industrial IoT gateway, an edge computing unit, and a time-series database. The multi-protocol industrial IoT gateway is used to connect to field controllers, sensors, and smart meters to achieve seamless acquisition of production process, safety risk, environmental emission, and equipment status data. The edge computing unit is used to preprocess and synchronize the collected data in time series; the time series database is used to store the preprocessed production data without loss.

3. The intelligent management system for steel production process standards based on a rule engine as described in claim 1, characterized in that: The semantic process standard rule base stores structured steel industry process standards, safety specifications, and environmental protection requirements, including process parameter thresholds, safety risk thresholds, environmental emission thresholds, and equipment operating status thresholds. It also supports business personnel to manage the entire lifecycle of rules through a visual interface.

4. The intelligent management system for steel production process standards based on a rule engine according to claim 1, characterized in that: The evaluation dimensions of the multi-attribute decision model include the severity of the abnormal event, the scope of its impact on production, and the priority of emergency response, classifying abnormal events into three risk levels: mild, moderate, and severe.

5. The intelligent management system for steel production process standards based on a rule engine according to claim 1, characterized in that: The anomaly classification and early warning module supports multiple push methods, including pop-up windows on industrial control terminals, WeChat for enterprises, work groups, and SMS, with different push frequencies and reminder intensities corresponding to different risk levels.

6. The intelligent management system for steel production process standards based on a rule engine according to claim 1, characterized in that: The dynamic closed-loop management module includes a handling process tracking unit, a result verification unit, and a rule optimization unit; the handling process tracking unit records information on the entire handling process of abnormal events. The result verification unit performs a quantitative evaluation of the treatment effect; The rule optimization unit generates rule base adjustment suggestions based on the evaluation results.

7. A rule-based intelligent management method for steel production process standards, characterized in that: The system according to any one of claims 1-6 includes the following steps: S1. Data collection of all production factors: Real-time collection of production process, safety risks, environmental emissions and equipment status data of the entire steel production process through industrial Internet of Things technology; S2, Intelligent Rule Analysis: Automatically compares the collected real-time data with the built-in semantic process standard rule library to complete the process compliance judgment and identify abnormal events; S3. Anomaly Classification Early Warning: Based on a multi-attribute decision model, the risk level of abnormal events is quantitatively classified, and the classification early warning information is accurately pushed to the corresponding responsible positions according to the dynamic mapping relationship of role permissions. S4. Dynamic closed-loop management: Record the entire handling process of abnormal events, verify and evaluate the handling effect, and feed the evaluation results back to the semantic process standard rule base to realize dynamic iterative optimization of rules.

8. The intelligent management method for steel production process standards based on a rule engine according to claim 7, characterized in that: In step S1, the collected raw data is first cleaned, deduplicated, format converted, and time-series synchronized preprocessed by the edge computing unit, and then transmitted to the time-series database for lossless storage. The data collection frequency is 1 time / second.

9. The intelligent management method for steel production process standards based on a rule engine according to claim 7, characterized in that: In step S3, minor anomalies are pushed to the field operator only, moderate anomalies are pushed to the operator and process supervisor, and severe anomalies are pushed to the operator, process supervisor, and middle and senior management personnel.

10. The intelligent management method for steel production process standards based on a rule engine according to claim 7, characterized in that: In step S4, the rule optimization unit performs monthly statistical analysis on historical processing data, automatically identifies rules with excessively high false alarm rates or missed alarms, generates threshold adjustment suggestions, and completes the rule base update after manual confirmation.