Steel coil data multi-dimensional alarm configuration method based on time sequence data and time window

Through a multi-dimensional alarm configuration method based on time series data and time windows, the problem of single and inflexible alarm configuration and poor real-time performance in the existing technology is solved, and flexible alarm configuration and real-time monitoring in industrial production are realized, which improves production safety and stability.

CN119049238AActive Publication Date: 2024-11-29WISDRI ENG & RES INC LTD

Patent Information

Application Number
CN202411144659.4
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-08-20
Publication Date
2024-11-29
Estimated Expiration
2044-08-20

AI Technical Summary

Technical Problem

In the prior art, the alarm configuration in data analysis and diagnosis is single and inflexible, the data source is inaccurate, the real-time performance is poor, the processing time window capability is weak, and it is difficult to meet the safety and stability requirements of the industrial production process.

Method used

Using the multi-dimensional alarm configuration method of steel coil data based on time series data and time windows, collect time series data through sensors, build an alarm configuration module, a time series data logic processing module and a time window processing module to realize custom, limit and third-party alarm rules dynamic configuration and judgment to ensure the timeliness and accuracy of alarms.

Benefits of technology

It realizes flexible alarm configuration and real-time monitoring, improves the safety and stability of the production process, reduces maintenance costs, simplifies the implementation process, and improves the production efficiency and the reliability of the normal operation of the system.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119049238A_ABST
    Figure CN119049238A_ABST
Patent Text Reader

Abstract

The invention discloses a steel coil data multi-dimensional alarm configuration method based on time sequence data and a time window. The method comprises the steps that the time sequence data of all measuring points in the steel coil production process are collected through a sensor; an alarm configuration module is constructed and used for defining and configuring alarm rules, alarm conditions and alarm levels; a time sequence data logic processing module is constructed and used for receiving the time sequence data, distinguishing different alarm types according to alarm rules issued by configuration, analyzing the time sequence data in real time, and matching the data into the alarm rules to obtain alarm result data conforming to the rules; the construction time window processing module is used for starting a timer in a configured time window and continuously receiving the alarm result data sent by the time sequence data logic processing module and the alarm result data generated by the third-party service; and whether an alarm condition is satisfied is determined. According to the invention, multi-dimensional flexible definition and configuration can be carried out on alarm data sources and rules, and the alarm timeliness and accuracy are ensured.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the field of data processing and analysis diagnosis information technology, and in particular to a multi-dimensional alarm configuration method for steel coil data based on time series data and time windows. Background Art

[0002] With the rapid development of modern industrial production, the requirements for production process safety, stability, and efficiency are increasing. Against this backdrop, industrial alarm systems have emerged as a critical tool for ensuring production safety, preventing economic losses, and safeguarding corporate reputation. These automated systems monitor various parameters in the production process in real time and, upon detecting anomalies, issue immediate alarms, prompting operators to take swift action to prevent accidents and potentially serious consequences.

[0003] The flexibility and accuracy of an alarm system are key to its success. Through carefully designed alarm configuration, the system can promptly issue alerts when equipment or system anomalies occur, prompting operators to take emergency measures, prevent accidents, and ensure production safety. Well-designed alarm configuration not only improves the management and maintenance efficiency of the alarm system but also enhances work efficiency by streamlining operational processes. Furthermore, appropriate alarm device and parameter configurations help promptly detect and resolve equipment failures and issues on the production line, ensuring stable operation of the production system. This efficient alarm mechanism not only ensures production safety but also provides a solid foundation for a company's sustainable development and market competitiveness.

[0004] Therefore, it is urgent to design a multi-dimensional alarm configuration method for steel coil data based on time series data and time windows. Summary of the Invention

[0005] The technical problem to be solved by the present invention is to provide a multi-dimensional alarm configuration method for steel coil data based on time series data and time windows, in response to the technical and scenario problems in data analysis and diagnosis in the existing technology, such as the single and inflexible alarm configuration, inaccurate data source, poor real-time performance, and weak time window processing capability.

[0006] The technical solution adopted by the present invention to solve its technical problem is:

[0007] The present invention provides a method for configuring multi-dimensional alarms for steel coil data based on time series data and a time window, the method comprising the following steps:

[0008] Step 1: Collect time series data of each measuring point in the steel coil production process through sensors;

[0009] Step 2: Build an alarm configuration module to define and configure alarm rules, alarm conditions, and alarm levels;

[0010] Step 3: Build a time series data logic processing module to receive time series data, distinguish different alarm types according to the configured alarm rules, parse the time series data in real time, match the data to the alarm rules, and obtain the alarm result data that meets the rules;

[0011] Step 4. Construct a time window processing module to start the timer within the configured time window, continuously receive the alarm result data sent by the time series data logic processing module and the alarm result data generated by the third-party service; determine whether the alarm conditions are met. If the alarm result data has been in the alarm state within the time range, it is determined that an alarm record is actually generated at this time and sent to the alarm processing business platform; if the alarm result data has been in the normal state within the time range, it is determined to be an alarm reset under the premise that an alarm has been issued.

[0012] Furthermore, the time series data of the present invention includes the name ID of the measuring point, measuring point data, and a timestamp.

[0013] Furthermore, the alarm rules in step 2 of the present invention include:

[0014] Custom alarm: Configure alarm formulas or multiple expressions based on data from multiple measuring points, including: four arithmetic operations (addition, subtraction, multiplication, and division), AND, OR, NOT, greater than, less than, and basic functions (max, min, and diff).

[0015] Limit alarm: including high-high limit, high limit, low limit, low-low limit alarms of numerical type and true / false and 0 / 1 alarms of Boolean type;

[0016] Third-party alarm: The alarm result data of the third-party alarm logic is sent to the time window processing module to further verify whether the alarm is valid within a certain time range and push it to the alarm business platform.

[0017] Furthermore, the specific method of the custom alarm, limit alarm, and third-party alarm in step 2 of the present invention includes:

[0018] The method of customizing alarm includes: judging whether there is speed abnormality in the steel coil processing production process, configuring custom alarm rules for the speed setting value and actual value during steel coil production, and determining that the speed is abnormal when the actual speed value exceeds the setting value by ±10%. The alarm formula or multi-expression configuration based on the actual speed value pv and the speed setting value sv of multiple measuring points is: (pv>sv*1.1)||(pv <sv*0.9);

[0019] The limit alarm method includes: during the steel coil processing and production process, the sensor continuously collects the cylinder / strip deviation distance data, and the threshold alarm configures the high-limit threshold, high-limit threshold, low-limit threshold, and low-low-limit threshold for the deviation distance data value. If the threshold configuration range is exceeded, an alarm message of strip deviation will be generated; for the process switching and flow in the process of entrance loading / exit unloading of steel coil production, the switch status data of various equipment / working conditions generated therein are used for monitoring. The normal state of a certain process should be true. By configuring the switch alarm for the process state, when the state becomes false, a switch alarm for the process is generated.

[0020] The third-party alarm method includes: for the process parameter data generated during the steel coil processing and production, including temperature, liquid level, and pressure, a process parameter alarm is generated when the process parameter exceeds the standard range corresponding to the process parameter; for the process processing production line, the production line is divided into multiple areas and positions, which are distributed with different types of process parameters. Each steel coil has a corresponding process parameter value at the corresponding area position. The process parameter standards required for the processing of steel coils of different brands and different widths and thicknesses are different. This scenario that changes dynamically due to the alarm rules is uniformly handled by the third-party alarm; the third-party alarm obtains the corresponding upper and lower limits of the process parameters based on the steel coil information at each position, and obtains the real-time value of the process parameter at the position. If the real-time value exceeds the upper and lower limits, an alarm is generated and sent to the time window processing module as the preliminary result of the third-party alarm logic.

[0021] Furthermore, the method for parsing data in step 3 of the present invention includes:

[0022] The time series data logic processing module receives the preset alarm configuration message body, obtains the alarm configuration alarmId, alarm name, alarm type, data source eventSourceVOList information, and matches the alarm rules - the formula polynomial for custom alarm type = 0, the point itemTag for limit alarm type = 1, and the upper and lower limit values ​​lower / upper.

[0023] Furthermore, the method of step 3 of the present invention further includes:

[0024] Step 3.1: The time series data logic processing module processes and analyzes the input time series data, obtains the data source by subscribing to the topic and measurement point information configured in the alarm configuration, and preprocesses the original time series data;

[0025] Step 3.2: Filter, calculate, match, and screen the data according to the alarm conditions set in the alarm configuration module;

[0026] Step 3.3: Summarize, count and forward the data that meets the alarm conditions, and send it to the time window processing module for analysis and judgment.

[0027] Furthermore, the specific methods of steps 3.1-3.3 of the present invention include:

[0028] By subscribing to the topic list in the alarm configuration, all measurement point data can be obtained in real time. The data includes the timestamp of receiving the message and all measurement point data information values. Among them, each measurement point data information includes id (measurement point identifier), v (data value), q (validity), and t (timestamp). First, check whether the timestamp is within a certain time range of nearly a certain number of minutes, and filter out dirty data caused by data retransmission due to various reasons. Then, parse the message body according to the rules of id, v, q, and t, filter out data with q being false, and assign the parsed v data value to the point in the alarm configuration.

[0029] Furthermore, the method of step 4 of the present invention includes:

[0030] Step 4.1: Divide the time series data into several time windows according to the time window size set by the user;

[0031] Step 4.2: Perform statistical analysis on the data within each time window. The data comes from the result data processed by the time series data logic processing module and the alarm result data generated by the third-party service.

[0032] Step 4.3: Based on the alarm conditions set in the alarm configuration module and the alarm rules associated with the alarm result data, determine whether the data in the current time window meets the alarm conditions;

[0033] Step 4.4: If the alarm condition is met, the corresponding alarm information is generated, prioritized according to the alarm level, and sent to relevant personnel or systems.

[0034] Furthermore, the specific methods of steps 4.1-4.4 of the present invention include:

[0035] The specific method for statistical analysis of data in each time window is:

[0036] The time series data logic processing module sends the alarm event / reset event results to the time window processing module. Within a certain time window period, if the alarm event recurs for a certain number of consecutive seconds or a certain number of times, that is, if the data of type 2 in the event message body of the same alarm configuration ID is continuously tracked and the time window configuration conditions in the alarm configuration are met, an alarm will be generated. The specific configuration conditions include the following three situations:

[0037] Continuous N seconds: Starting from the first alarm event triggering, no alarm event reset occurs within N consecutive seconds;

[0038] Consecutive N times: Starting from the first alarm event trigger, the same alarm event trigger occurs N times in succession, and no alarm event reset occurs;

[0039] Occurs N times or more within N consecutive seconds: Starting from the first alarm event trigger, the same alarm event trigger occurs N times or more within N consecutive seconds, and no alarm event reset occurs.

[0040] Furthermore, the time series data logic processing module of the present invention and the third-party service transmit data with the time window processing module via MQTT / HTTP.

[0041] The beneficial effects of the present invention are as follows: By dynamically configuring data sources, time windows, and alarm rules, and standardizing alarm configuration methods, the present invention is flexibly applicable to factory production systems in various fields. On-site implementation is easy, with a low learning curve and minimal complexity. Only the relevant configuration is required, significantly shortening the implementation cycle and improving debugging efficiency. Alarm configuration information can be flexibly added or modified without modifying any code to complete relevant alarm judgments, effectively reducing maintenance and operation costs and enabling timely and accurate alarm configuration. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] The present invention will be further described below with reference to the accompanying drawings and embodiments, in which:

[0043] Figure 1 is a system module framework diagram of an embodiment of the present invention;

[0044] Figure 2 It is an alarm configuration module of an embodiment of the present invention. DETAILED DESCRIPTION

[0045] In order to make the purpose, technical solutions and advantages of the present invention more clearly understood, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0046] Example 1

[0047] like Figure 1As shown, the embodiment of the present invention implements multi-dimensional definition and configuration of data sources and rules through an alarm configuration module, which mainly includes three types of alarms: custom alarms, limit alarms, and third-party alarms. Among them, real-time time series data with timestamps is sent to the logic processing module for alarm logic judgment. The results of the judgment are transferred to the time window processing module to verify the validity and level of the alarm, and the alarm information is pushed to the alarm business platform. The present invention can alarm faults in a timely and accurate manner, improve the timeliness of emergency response, greatly improve production efficiency, and ensure the normal operation of the system.

[0048] The present invention can be used as a data analysis and diagnosis module, and its implementation system (see Figure 1 ) includes alarm configuration module, time series data logic processing module, and time window processing module.

[0049] Step 1: Collect time series data of each measuring point in the steel coil production process through sensors;

[0050] Step 2: Build an alarm configuration module to define and configure alarm rules, alarm conditions, and alarm levels;

[0051] like Figure 2 FIG. 1 shows an alarm configuration module according to an embodiment of the present invention, which is mainly responsible for defining and configuring alarm rules.

[0052] Custom alarms: Alarm formulas or multi-expressions can be flexibly configured based on data from multiple measuring points, including the four arithmetic operations of addition, subtraction, multiplication, and division, logical operators such as AND, OR, NOT, greater than, and less than, and basic functions such as max / min / diff.

[0053] In a preferred embodiment of the present invention, it is determined whether there is a speed anomaly in the steel coil processing production process. By configuring a custom alarm rule for the speed setting value and the actual value during steel coil production, it is determined that the speed is abnormal when the actual speed value exceeds the setting value by ±10%. The alarm formula or multi-terminal expression based on the actual speed value (pv) and the speed setting value (sv) of multiple measuring points can be configured as: (pv>sv*1.1)||(pv <sv*0.9)。

[0054] Limit alarm: mainly divided into numerical type high-high limit, high limit, low limit, low-low limit alarms and Boolean type switch true / false, 0 / 1 alarms.

[0055] In a preferred embodiment of the present invention, during the production process of steel coils, the strip may deviate abnormally. In severe cases, this may lead to production obstruction or even shutdown, affecting production efficiency and quality. In response to the deviation of the strip, the sensor continuously collects the cylinder / strip deviation distance data during the production process, and the threshold alarm configures the deviation distance data value with an upper limit of 100mm, an upper limit of 60mm, a lower limit of -60mm, and a lower limit of -100mm. If the threshold configuration range is exceeded, an alarm message indicating the strip has deviated will be generated. During the process of loading at the entrance / unloading at the exit of steel coil production, there will be a large number of process switches and flows, in which the switch status data of various equipment / working conditions are generated for monitoring. The normal state of a certain process should be true. By configuring the switch alarm for the process state, a switch alarm for the process will be generated when the state becomes false.

[0056] Third-party alarm: The preliminary result data of the third-party alarm logic can be sent to the time window processing module to further verify whether the alarm is valid within a certain time range and push it to the alarm business platform.

[0057] In a preferred embodiment of the present invention, the coil processing process generates a large amount of process parameter data, including common parameters such as temperature, liquid level, and pressure. These parameters often have corresponding standard ranges, and any deviations from these ranges trigger process parameter alarms. Production lines focused on process processing can reach hundreds or even thousands of meters in length and are typically divided into zones and locations. These zones contain different types of process parameters, and each coil has a corresponding process parameter value at its corresponding zone location. For example, coil A has a process parameter temperature of 800°C in heating section 1 of the furnace, while coil B has a process parameter temperature of 300°C in cooling section 3 of the furnace. The process parameter standards required for processing coils of different grades, widths, and thicknesses vary. Therefore, scenarios requiring dynamic changes in alarm rules are handled uniformly by a third-party alarm system. The third-party alarm system determines the corresponding upper and lower limits for the process parameters based on the coil information at each location, as well as the real-time value of the process parameter at that location. If the real-time value exceeds the upper and lower limits, an alarm is generated and sent to the time window processing module as the preliminary result of the third-party alarm logic.

[0058] Step 3: Build a time series data logic processing module to receive time series data, distinguish different alarm types according to the configured alarm rules, parse the time series data in real time, match the data to the alarm rules, and obtain the alarm result data that meets the rules;

[0059] Time series data logic processing module. The time series data collector sends various time series data collected from sensors such as PLCs and switch stations to the logic processing module. This time series data includes the measurement point ID, measurement point data value, and timestamp. The logic processing module distinguishes different alarm types based on the configured alarm rules, analyzes the time series data in real time, matches the data to the alarm rules, and sends the recorded data that meets the alarm rules to the time window processing module.

[0060] In a preferred embodiment of the present invention, the time series data logic processing module receives the pre-agreed alarm configuration message body, obtains the alarm configuration alarmId, alarm name name, alarm type type, data source eventSourceVOList and other information, and matches the alarm rules - the formula polynomial of the custom alarm type = 0, the point itemTag of the limit alarm type = 1, and the upper and lower limit values ​​lower / upper.

[0061] The specific implementation method of the time series data logic processing module includes:

[0062] Step 3.1: The time series data logic processing module processes and analyzes the input time series data, obtains the data source by subscribing to the topic and measurement point information configured in the alarm configuration, and preprocesses the original time series data;

[0063] Step 3.2: Filter, calculate, match, and screen the data according to the alarm conditions set in the alarm configuration module;

[0064] Step 3.3: Summarize, count and forward the data that meets the alarm conditions, and send it to the time window processing module for analysis and judgment.

[0065] By subscribing to the topic list in the alarm configuration, all measurement point data is obtained in real time. This data includes the timestamp of the message receipt and all measurement point data values. Each measurement point data value includes id (measurement point identifier), v (data value), q (validity), and t (timestamp). First, the timestamp is verified to be within the last 5 minutes, filtering out dirty data caused by retransmission due to various reasons. Then, the message body is parsed according to the id, v, q, and t rules, filtering out data where q is false. The parsed v data value is assigned to the point in the alarm configuration.

[0066] Step 4. Construct a time window processing module to start the timer within the configured time window, continuously receive the alarm result data sent by the time series data logic processing module and the alarm result data generated by the third-party service; determine whether the alarm conditions are met. If the alarm result data has been in the alarm state within the time range, it is determined that an alarm record is actually generated at this time and sent to the alarm processing business platform; if the alarm result data has been in the normal state within the time range, it is determined to be an alarm reset under the premise that an alarm has been issued.

[0067] Time window processing module. This module starts the timer within the configured time window, continuously receives the recorded data that meets the alarm rules thrown by the time series data logic processing module, and determines whether the recorded data has been in the alarm state within the time range. If the recorded data has been in the alarm state, it is determined that an alarm record has actually been generated at this time, and can be sent to the alarm processing business platform through other communication methods such as HTTP / MQTT to prevent false alarms. Similarly, if the recorded data has always been in a normal state, it can be determined that the alarm has been reset under the premise that an alarm has been issued. Among them, whether continuous judgment is required within the time window range can be configured. If non-continuous alarm judgment is configured, certain normal state data within the time window range can be automatically ignored, and ultimately timely and accurate production alarms are generated.

[0068] The specific implementation method of the time window processing module includes:

[0069] Step 4.1: Divide the time series data into several time windows according to the time window size set by the user;

[0070] Step 4.2: Perform statistical analysis on the data within each time window. The data comes from the result data processed by the time series data logic processing module and the alarm result data generated by the third-party service.

[0071] Step 4.3: Based on the alarm conditions set in the alarm configuration module and the alarm rules associated with the alarm result data, determine whether the data in the current time window meets the alarm conditions;

[0072] Step 4.4: If the alarm condition is met, the corresponding alarm information is generated, prioritized according to the alarm level, and sent to relevant personnel or systems.

[0073] The time series data logic processing module sends the alarm event / reset event results to the time window processing module. Within a certain time window period, an alarm event is generated if a certain number of consecutive seconds or a certain number of consecutive times occur, that is, if the data type 2 in the event message body of the same alarm configuration ID is continuously tracked and the time window configuration conditions in the alarm configuration are met, an alarm is finally generated. The specific configuration conditions are divided into the following three cases:

[0074] 1. Continuous N seconds: Starting from the first alarm event trigger, no alarm event reset occurs within N consecutive seconds.

[0075] 2. Consecutive N times: Starting from the first alarm event trigger, the same alarm event trigger occurs N times in succession, and no alarm event reset occurs.

[0076] 3. Occurs N times or more within N consecutive seconds: Starting from the first alarm event trigger, the same alarm event trigger occurs N times or more within N consecutive seconds, and no alarm event reset occurs.

[0077] Example 2

[0078] The embodiment of the present invention is mainly used in a multi-dimensional alarm configuration method based on time series data and time windows in data processing and analysis diagnosis, and has functional modules including: an alarm configuration module, a time series data logic processing module, and a time window processing module.

[0079] The multi-dimensional alarm configuration method using the method of the present invention mainly includes the following steps:

[0080] Alarm Configuration: This module is primarily used to set alarm rules, alarm conditions, and alarm levels. Users can flexibly customize alarm rules across multiple dimensions, such as setting thresholds and comparison operators, based on actual needs. Multiple alarm conditions can be set to meet the needs of complex scenarios. Furthermore, each alarm condition can be assigned a different alarm level, allowing for appropriate response measures based on severity when an alarm occurs.

[0081] Time Series Data Logic Processing Module: This module is primarily responsible for processing and analyzing input time series data. First, it obtains data sources by subscribing to the topics and measurement point information configured in the alarm configuration. It then preprocesses and cleans the raw data, removing noise and filling missing values ​​and non-production dirty data. Then, based on the alarm conditions set in the alarm configuration module, it filters, calculates, matches, and selects the data. Finally, data that meets the alarm conditions is aggregated, counted, and forwarded for analysis and judgment by the subsequent time window processing module.

[0082] Time Window Processing Module: This module is primarily used to divide and analyze time series data into time windows. First, the time series data is divided into several time windows based on the time window size set by the user. Next, a statistical analysis is performed on the data within each time window. This data is derived from the result data processed by the time series data logic processing module and the alarm result data generated by third-party services. Next, based on the alarm conditions set in the alarm configuration module and the alarm rules associated with the alarm result data, a determination is made as to whether the data within the current time window meets the alarm conditions. If the alarm conditions are met, the corresponding alarm information is generated and prioritized according to the alarm level. Finally, the alarm information is sent to the relevant personnel or system so that timely measures can be taken to handle and resolve the situation.

[0083] During specific implementation, first configure the alarm-related information in multiple dimensions in the alarm configuration module, mainly including: alarm name, alarm type, alarm level, alarm rules, alarm points, alarm description, etc. The time series data collected by the time series data logic processing module is transmitted via MQTT / HTTP, and the time series data logic processing module judges the received data according to the preset alarm rules and thresholds. If the data exceeds the threshold or meets the alarm rules, an alarm is triggered. At the same time, the time window processing module will continue to receive the alarm result data sent by the logic processing module within the specified time range according to the set time window, and perform timed or fixed-length processing on the alarm results to ensure the accuracy and timeliness of the alarm. Finally, the alarm result data that meets the alarm configuration is sent to the business platform by the time window processing module, and the alarm information is pushed to the relevant departments and personnel on the system platform for timely processing.

[0084] Practice has proven that the multi-dimensional alarm method of this invention addresses the technical and scenario-specific issues in existing data analysis and diagnosis, such as the single, inflexible alarm configuration, inaccurate data sources, poor real-time performance, and weak time window processing capabilities. Flexible multi-dimensional definition and configuration of alarm data sources and rules, and alarm judgment based on a time series data logic processing module and a time window processing module, ensure the timeliness and accuracy of alarms, ensuring the safety and smooth progress of the production process, and avoiding significant economic losses and reputational damage to the factory.

[0085] It should be understood that the size of the serial numbers of the steps in the above embodiments does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of this application.

[0086] It should be understood that those skilled in the art can make improvements or changes based on the above description, and all such improvements and changes should fall within the scope of protection of the appended claims of the present invention.

Claims

1. A multi-dimensional alarm configuration method for steel coil data based on time series data and time windows, It is characterized in that The method comprises the following steps: Step 1: Collecting time series data of each measuring point in the steel coil production process through sensors; Step 2: Build an alarm configuration module to define and configure alarm rules, alarm conditions, and alarm levels; Step 3: Build a time series data logic processing module to receive time series data, distinguish different alarm types according to the configured alarm rules, parse the time series data in real time, match the data to the alarm rules, and obtain the alarm result data that meets the rules; Step 4: Construct a time window processing module to start the timer within the configured time window, continuously receive the alarm result data sent by the time series data logic processing module and the alarm result data generated by the third-party service; determine whether the alarm conditions are met. If the alarm result data has been in the alarm state within the time range, it is determined that an alarm record is actually generated at this time and sent to the alarm processing business platform; if the alarm result data has been in the normal state within the time range, it is determined to be an alarm reset under the premise that an alarm has been issued.

2. According to claim 1, the multi-dimensional alarm configuration method for steel coil data based on time series data and time window, It is characterized in that The time series data includes the name ID of the measuring point, measuring point data, and a timestamp.

3. According to claim 2, the multi-dimensional alarm configuration method for steel coil data based on time series data and time window, It is characterized in that The alarm rules in step 2 include: Custom alarm: configure alarm formulas or multiple expressions based on multiple measuring point data, including: four arithmetic operations of addition, subtraction, multiplication and division, AND, OR, NOT, GREATER THAN, LESS THAN logical operators, and basic functions max / min / diff; Limit alarm: including high-high limit, high limit, low limit, low-low limit alarms of numerical type and switch value true / false, 0 / 1 alarms of Boolean type; Third-party alarm: The alarm result data of the third-party alarm logic is sent to the time window processing module to further verify whether the alarm is valid within a certain time range and push it to the alarm business platform.

4. According to claim 3, the multi-dimensional alarm configuration method for steel coil data based on time series data and time window, It is characterized in that The specific methods of customizing alarm, limit alarm, and third-party alarm in step 2 include: The method of customizing alarm includes: judging whether there is speed abnormality in the steel coil processing production process, configuring custom alarm rules for the speed setting value and actual value during steel coil production, and judging that the speed is abnormal when the actual speed value exceeds the setting value by ±10%. The alarm formula or multiple expressions based on the actual speed value pv and the speed setting value sv of multiple measuring points are configured as follows: (pv>sv*1.1)||(pv <sv*0.9); The limit alarm method includes: during the steel coil processing and production process, the sensor continuously collects the cylinder / strip deviation distance data, and the threshold alarm configures the high-limit threshold, high-limit threshold, low-limit threshold, and low-limit threshold for the deviation distance data value. If the threshold configuration range is exceeded, the alarm information of the strip deviation is produced; for the process switching and flow existing in the process of entrance loading / exit unloading of steel coil production, the switch quantity status data of various equipment / working conditions generated therein are used for monitoring. The normal state of a certain process should be true. By configuring the switch quantity alarm for the process state, when the state becomes false, the switch quantity alarm for the process is generated. The third-party alarm method includes: for the process parameter data generated during the steel coil processing and production, including temperature, liquid level, and pressure, when the process parameter exceeds the standard range corresponding to the process parameter, an alarm is generated; for the process processing production line, the production line is divided into multiple areas and positions, where different types of process parameters are distributed, and each steel coil has a corresponding process parameter value at the corresponding area position. The process parameter standards required for processing steel coils of different brands and different widths and thicknesses are different. This scenario that changes dynamically due to the alarm rules is uniformly handled by the third-party alarm; the third-party alarm obtains the corresponding upper and lower limits of the process parameters based on the steel coil information at each position, and obtains the real-time value of the process parameter at the position. If the real-time value exceeds the upper and lower limits, an alarm is generated and sent to the time window processing module as the preliminary result of the third-party alarm logic.

5. According to claim 1, the multi-dimensional alarm configuration method for steel coil data based on time series data and time window, It is characterized in that The method for parsing data in step 3 includes: The time series data logic processing module receives the preset alarm configuration message body, obtains the alarm configuration alarmId, alarm name name, alarm type type, data source eventSourceVOList information, and matches the alarm rules - the formula polynomial of the custom alarm type = 0, the point itemTag of the limit alarm type = 1, and the upper and lower limits lower / upper.

6. The method for configuring multi-dimensional alarms for coil data based on time series data and time windows according to claim 1, It is characterized in that The method of step 3 further includes: Step 3.1, the time series data logic processing module processes and analyzes the input time series data, obtains the data source by subscribing to the topic and measurement point information configured in the alarm configuration, and pre-processes the original time series data; Step 3.2: Filter, calculate, match and screen the data according to the alarm conditions set in the alarm configuration module; Step 3.3: Summarize, count and forward the data that meets the alarm conditions, and send it to the time window processing module for analysis and judgment.

7. The method for configuring multi-dimensional alarms for coil data based on time series data and time windows according to claim 6, It is characterized in that The specific method of steps 3.1-3.3 includes: By subscribing to the topic list in the alarm configuration, all the measuring point data can be obtained in real time; the data includes the timestamp of receiving the message and all the measuring point data information values; each measuring point data information includes id-measuring point identification, v-data value, q-whether it is valid, and t-timestamp; first check whether the timestamp is within a certain time range of nearly a certain number of minutes, and filter out dirty data caused by data retransmission due to various reasons; then parse the message body according to the rules of id, v, q, t, filter out the data with q being false, and assign the parsed v data value to the point in the alarm configuration.

8. The method for configuring multi-dimensional alarms for coil data based on time series data and time windows according to claim 5, It is characterized in that The method of step 4 comprises: Step 4.1: Divide the time series data into several time windows according to the time window size set by the user; Step 4.2: Perform statistical analysis on the data in each time window. The data comes from the result data processed by the time series data logic processing module and the alarm result data generated by the third-party service. Step 4.3: According to the alarm conditions set in the alarm configuration module and the alarm rules associated with the alarm result data, determine whether the data in the current time window meets the alarm conditions; Step 4.4: If the alarm condition is met, the corresponding alarm information is generated, and the alarm level is prioritized and sent to relevant personnel or systems.

9. The method for configuring multi-dimensional alarms for coil data based on time series data and time windows according to claim 8, It is characterized in that The specific methods of steps 4.1-4.4 include: The specific method for statistical analysis of the data in each time window is: The time series data logic processing module sends the alarm event / reset event results to the time window processing module; within a certain time window period, the alarm event repeats for a certain number of seconds or a certain number of times, that is, the data of type 2 in the event message body of the same alarm configuration ID is continuously tracked, and the time window configuration conditions in the alarm configuration are met, then an alarm is actually generated; the specific configuration conditions include the following three situations: Continuous N seconds: Starting from the first alarm event triggering, no alarm event reset occurs within N consecutive seconds; N consecutive times: Starting from the first alarm event trigger, the same alarm event trigger occurs N times in succession, and no alarm event reset occurs; Occurs N times or more within N consecutive seconds: Starting from the first alarm event trigger, the same alarm event trigger occurs N times or more within N consecutive seconds, and no alarm event reset occurs.

10. The method for configuring multi-dimensional alarms for coil data based on time series data and time windows according to claim 1, It is characterized in that The time series data logic processing module and third-party services transmit data to the time window processing module through MQTT / HTTP.

Citation Information

Patent Citations

  • DECT handle with intelligent automatic alarm functions

    CN111862543A

  • Alarm diagnosis method and system based on real-time data monitoring

    CN112071031A

  • Alarm analysis method, device and equipment based on time sequence data and medium

    CN115952217A

  • Multi-stage early warning method based on rationality pre-judgment of planned coil division of cold-rolled product

    CN116798200A

  • Industrial alarm analysis method and system

    CN118155389A

Cited By

  • Full-automatic steel drum production system

    CN119512016A

  • Off-line chemical process alarm management diagnosis service method

    CN120494509A

  • Off-line chemical process alarm management diagnostic service method

    CN120494509B