A method and system for generating multi-dimensional alarm light signs

Abstract

This invention relates to the field of sequential controllers or logic controllers, and particularly to the use of digital processing devices. The technical solution is as follows: A method for generating multi-dimensional alarm light signs, comprising: configuring alarm configurations based on tag numbers and alarm information, and setting the priority of each alarm configuration; reading the alarm configurations and acquiring all alarms in the system; determining dimensions and filtering; calculating the similarity of alarms within a time window; processing the filtered alarms; comparing the time window with a similarity model to determine if there are identical alarm events; for identical alarm events, triggering an alarm according to the information of the high-priority alarm configuration, with the information of the low-priority alarm configuration serving as a supplement. This method solves the problems of large configuration workload, error-proneness, low efficiency, and one-sided processing dimensions in existing technologies, and proposes a multi-dimensional alarm light sign generation method and system, achieving the goals of multi-dimensional processing, high stability, high efficiency, and small configuration workload.

Description

Technical Field

[0001] This invention relates to the field of sequential controllers or logic controllers, and more particularly to the use of digital processing devices. Background Technology

[0002] In medium to large-scale industrial DCS control systems, the number of tag numbers, including both hard and soft points, often reaches tens or even hundreds of thousands, and each tag number often has multiple alarm types. When designing alarm indicator panels, directly selecting from so many tag number alarm types would result in a very large configuration workload and be prone to errors. For example, Chinese patent CN118584882A discloses a high-precision integration method and system for alarm event signals in a distributed control system, providing the following technical solution: This invention discloses a high-precision integration method and system for alarm event signals in a distributed control system, relating to the field of signal data integration technology, including storing real-time data and generated alarm event information on a DCS database server; constructing a physical data link to transmit real-time data generated in the DCS and corresponding alarm signal information from the security I zone to the management information zone; and calling real-time measurement point data in the real-time library database and alarm event message queue data in the corresponding InfluxDB database. The high-precision integration method for alarm event signals in a distributed control system provided by this invention integrates the Distributed Control System (DCS) and the Plant-Level Monitoring System (SIS), ensuring that the alarm event information of the measurement points in the SIS remains consistent with that in the DCS. This avoids the workload of repeatedly performing alarm monitoring condition judgment and alarm level judgment in the plant-level monitoring system, significantly reducing implementation costs. However, the aforementioned high-precision integration method and system for alarm event signals in a distributed control system involves a large configuration workload, is prone to errors, and lacks statistical functions based on alarm partitions, flowcharts, alarm statistical tag numbers, alarm levels, and alarm types. Furthermore, it cannot effectively handle the same alarm event, which can easily lead to duplicate alarms and significantly reduce efficiency. Summary of the Invention

[0003] This invention solves the problems of large configuration workload, easy error, low efficiency and one-sided processing in the prior art. It proposes a multi-dimensional alarm light sign generation method and system, which achieves the goals of multi-dimensional processing, high stability, high efficiency and small configuration workload.

[0004] To achieve the above objectives, the present invention adopts the following technical solution:

[0005] A method for generating multi-dimensional alarm light signs includes the following steps:

[0006] S1: Configure the alarm configuration according to the tag number and alarm information, and set the priority of each alarm configuration;

[0007] S2: Read alarm configuration and obtain all system alarms;

[0008] S3: Determine the dimension and perform filtering, calculate the similarity of alarms within the time window, process the filtered alarms, and determine whether there are the same alarm events by comparing the time window with the similarity model.

[0009] S4: For the same alarm event, the alarm will be triggered according to the information configured for the high-priority alarm, and the information configured for the low-priority alarm will be used as a supplement.

[0010] A multi-dimensional alarm light sign system includes: a data acquisition module, a dimension judgment module, and an alarm module. The data acquisition module collects alarm status and tag number information, as well as other real-time data, and transmits them to the dimension judgment module. The dimension judgment module receives configuration information transmitted by the configuration module and transmits the filtered alarm data to the alarm module. The alarm module filters and analyzes the alarm data and generates the final result to show the display status of the alarm light sign, which is then sent to the display module.

[0011] The advantage of this design is that it enables multi-dimensional management and integration of alarms, reduces duplicate alarms, improves alarm identification and accuracy, and achieves priority management, making it easier for operators to quickly locate key issues.

[0012] Preferably, step S3 further includes the following steps:

[0013] S3.1: Perform corresponding filtering based on the filtering type, such as tag number, tag number alarm type, alarm level, or alarm type;

[0014] S3.2: Set the required time window and obtain the timestamp of each alarm message;

[0015] S3.3: Sort the received alarm information according to the timestamp, and check the difference between the timestamp of the alarm record and the timestamp of the next alarm record in sequence;

[0016] S3.4: Determine whether the timestamp difference is within the time window range. If yes, proceed to step S3.5; otherwise, return to step S3.3 and calculate the next timestamp difference.

[0017] S3.5: Process the alarm text and calculate its similarity.

[0018] The advantage of this design is that it can effectively filter and classify alarm data through time windows and similarity calculations, avoid repeated alarms for the same event, and improve the level of intelligence in alarm management.

[0019] Preferably, step S3.5 specifically includes the following steps:

[0020] S3.51: Perform text preprocessing on the alarm content, removing stop words and punctuation, and extracting stems;

[0021] S3.52: Convert each alarm message into a word set, and calculate the size of the intersection and union of the two sets;

[0022] S3.53: Calculate the value of the first similarity and determine whether the first similarity is less than the first similarity threshold. If yes, proceed to step S3.54; otherwise, determine that they are the same alarm event.

[0023] S3.54: Convert the alarm content into a vector representation to obtain the feature vector of each alarm;

[0024] S3.55: Calculate the dot product and magnitude of the two vectors, and calculate the second similarity value;

[0025] S3.56: Determine whether the second similarity is less than the second similarity threshold. If yes, then it is not the same alarm event and return to step S3.3. If no, then it is determined to be the same alarm event.

[0026] The advantage of this design is that, through text preprocessing and similarity calculation, it is possible not only to analyze alarm content based on semantics, but also to further improve the accuracy of the analysis through vector calculation, ensuring accurate identification of the same alarm event.

[0027] Preferably, in step S4, if the alarm event is determined to be the same, the alarm information with the highest priority is selected and an alarm is triggered. It is then determined whether the alarm information with lower priority has any additional information compared to the alarm information with the highest priority. If it does, it is retained as supplementary information; if it does not, the alarm information with lower priority is deleted.

[0028] The advantage of this design is that by prioritizing information, key alarm information can be retained, while improving information processing efficiency, making alarm content more concise and focused, and additional information can also be retained as supplementary information to prevent omissions.

[0029] Preferably, in step S1, configuring the alarm configuration according to the tag number includes configuring according to the alarm partition, flowchart and direct tag number. The alarm partition is a tag number partition or a custom alarm partition. The custom alarm partition classifies the tag number into various custom partition names according to actual needs. One or more alarm partitions can be selected. The flowchart includes several graphical objects, and the graphical objects are associated with several tag numbers.

[0030] The advantage of this design is that it provides a flexible configuration method for alarm configuration. Users can customize partitions and flowcharts according to actual needs, thereby meeting the diverse needs of complex control scenarios.

[0031] Preferably, the direct tag number is selected by the tag number selector of the system to select one or more tag numbers. If the alarm type corresponding to the tag number is selected, it is the tag number alarm type dimension. The alarm configuration according to the tag number also includes alarm tag numbers. The alarm tag numbers group the related tag number alarm types together, and the alarm tag numbers handle the alarm status of the entire group.

[0032] Preferably, in step S1, configuring the alarm configuration according to the alarm information includes configuring according to the alarm level and alarm category. Each alarm type has a corresponding alarm level; a tag number has several alarm types. Tag numbers of the same type have the same alarm type, and tag numbers of different types can also have the same alarm type. One or more of the alarm level and the alarm type can be selected.

[0033] Preferably, in step S3, if the dimension is an alarm partition, the alarm partition is obtained, and all relevant tag numbers are obtained through the alarm partition; if the dimension is a flowchart, the flowchart name is extracted and the flowchart configuration file is parsed, and then the tag numbers are extracted. If multiple tag numbers are extracted from the flowchart, all extracted tag numbers are integrated into a list, duplicate tag numbers are removed, and all relevant tag numbers are obtained; if the dimension is a tag number, the selected tag number is directly obtained; if the dimension is a tag number alarm type, the tag number alarm type is directly obtained; if the dimension is an alarm tag number, all tag number alarm types are obtained through alarm statistics tag numbers; if the dimension is an alarm level, the selected alarm level is obtained; if the dimension is an alarm type, the selected alarm type is obtained.

[0034] The advantage of this design is that it provides a wealth of filtering dimensions and data extraction methods, thereby meeting the alarm statistics needs of different scenarios and improving the applicability and flexibility of the system.

[0035] Preferably, in step S3.53, the specific method for calculating the value of the first similarity is: the first similarity is the quotient of the size of the intersection of the two sets and the size of the size of the union of the two sets; the specific method for calculating the value of the second similarity is: the second similarity is the quotient of the dot product of the two vectors and the product of the magnitudes of the two vectors.

[0036] The advantage of this design is that by combining two similarity calculation methods, the similarity between alarm content can be analyzed more accurately, thereby improving the efficiency and accuracy of identifying the same alarm event.

[0037] Compared with the prior art, the beneficial effects of the present invention are as follows.

[0038] 1. This invention significantly improves the flexibility and accuracy of alarm management by statistically analyzing and filtering alarm information from multiple dimensions. These dimensions include alarm partitions, flowcharts, tag numbers, tag number alarm types, alarm levels, and alarm types, adaptable to the needs of various industrial control scenarios. This design allows users to flexibly select the required dimensions for alarm configuration based on different statistical needs, improving the system's configurability and scalability. Furthermore, by combining alarm information from different dimensions, alarm data can be filtered and integrated more efficiently, simplifying the management of complex systems, reducing error rates, and enhancing system adaptability and efficiency.

[0039] 2. This invention uses time windows and similarity calculations to determine the same alarm event, effectively avoiding interference from multiple duplicate alarms. The time window ensures that the same alarm event is identified within a short period, preventing missed alarms caused by the same alarm event being classified as the same for an extended period. The similarity model identifies whether it is the same event at the semantic level. Combining these two methods for comprehensive analysis significantly reduces the possibility of false alarms and missed alarms. Simultaneously, the introduction of a priority mechanism allows high-priority alarms to be presented to users more quickly and comprehensively, while low-priority alarms are retained or removed as supplementary information, achieving precise management of unified alarm events. This design helps operators focus on key issues and improves response efficiency to emergencies.

[0040] 3. This invention significantly improves the efficiency of alarm configuration through modular design and multi-dimensional options, especially when dealing with complex industrial processes. Users can quickly filter and configure the required alarm rules based on multiple dimensions such as alarm zone, tag number, and alarm type, eliminating the need for manual setting and greatly reducing configuration time. Furthermore, this invention supports batch operations and template-based management, further optimizing the configuration process and making it more efficient and convenient for large-scale system applications. This design can flexibly adapt to various complex scenarios while ensuring the accuracy and consistency of configuration, providing strong support for the rapid deployment and operation of industrial control systems. Attached Figure Description

[0041] Figure 1 This is an overall flowchart of a multi-dimensional alarm light sign generation method and system according to the present invention.

[0042] Figure 2 This invention provides a flowchart of a multi-dimensional alarm light sign generation method and system for judging the same alarm event. Detailed Implementation

[0043] To make the objectives, technical solutions, and advantages of this disclosure clearer, the embodiments of this disclosure will be further described in detail below with reference to the accompanying drawings. The proportions of the components are not drawn to scale, and the proportions and dimensions shown in the drawings should not limit the essential technical solutions of the invention. These embodiments do not exhaustively describe all details, nor do they limit the invention to merely the specific embodiments described.

[0044] See Figure 1-2 As shown, a method for generating multi-dimensional alarm light signs includes the following steps:

[0045] S1: Configure the alarm configuration according to the tag number and alarm information, and set the priority of each alarm configuration;

[0046] S2: Read alarm configuration and obtain all system alarms;

[0047] S3: Determine the dimension and perform filtering, calculate the similarity of alarms within the time window, process the filtered alarms, and determine whether there are the same alarm events by comparing the time window with the similarity model.

[0048] S4: For the same alarm event, the alarm will be triggered according to the information configured for the high-priority alarm, and the information configured for the low-priority alarm will be used as a supplement.

[0049] A multi-dimensional alarm light sign system includes: a data acquisition module, a dimension judgment module, and an alarm module. The data acquisition module collects alarm status and tag number information, as well as other real-time data, and transmits them to the dimension judgment module. The dimension judgment module receives configuration information transmitted by the configuration module and transmits the filtered alarm data to the alarm module. The alarm module filters and analyzes the alarm data and generates the final result to show the display status of the alarm light sign, which is then sent to the display module.

[0050] This invention can manage and integrate alarms from multiple dimensions, reduce duplicate alarms, improve alarm identification and accuracy, and at the same time realize priority management, making it easier for operators to quickly locate key issues.

[0051] like Figure 1 In one embodiment shown, Figure 1This invention presents an overall flowchart of a multi-dimensional alarm light sign generation method and system. The multi-dimensional alarm light sign generation method of this invention includes the following main steps: First, alarm configuration is configured based on tag numbers and alarm information. In this stage, the tag numbers to be monitored are associated with their corresponding alarm information, and the priority of each alarm configuration is set. The priority setting ensures that among multiple alarm messages, users focus on the most important information first, while additional information about the same alarm event can be treated as supplementary or secondary information. The alarm information configuration can be flexibly selected based on multiple dimensions such as alarm partitions, flowcharts, or specific tag numbers. For example, alarm partition configuration can be tag number partitioning or custom partitioning. Users can classify different tag numbers into specific partition names according to actual needs, thereby achieving more efficient alarm management.

[0052] Next, the system reads the alarm configuration set above and acquires all alarm data from the system. This step is the core of the entire process, classifying and managing all possible alarms through the configured logic. Acquiring alarm data ensures the system's real-time performance, capturing all alarm events and providing fundamental data support for subsequent filtering and processing.

[0053] Next, the system performs dimensional judgment and filtering on the alarm data. In this step, the system filters the data based on the alarm dimensions configured in the configuration. For example, if the dimension is alarm partition, the system will extract all relevant tag numbers within the alarm partition; if the dimension is flowchart, it will parse the flowchart configuration file and extract the relevant tag number information; if the dimension is tag number or tag number alarm type, it will directly extract the corresponding tag number or alarm type; if the dimension is alarm level or alarm type, it will extract the corresponding alarm level and type. This multi-dimensional filtering method can meet the diverse needs of different users and prepare for subsequent alarm event processing.

[0054] After filtering, the system further analyzes the selected alarm information, including calculating alarm similarity within a time window. To avoid duplicate alarms for the same event, the system compares the alarm data against a time window and a similarity model to determine if they are the same alarm event. Specifically, the system preprocesses the alarm content, such as removing stop words and punctuation, converting the alarm text into a word set or vector representation, and uses a similarity calculation model (such as the ratio of intersection to union or vector dot product calculation) to determine whether two alarms are the same event.

[0055] Finally, for cases identified as the same alarm event, the system will filter and process the alarm information according to priority. The highest priority alarm information will be displayed as the primary alarm, while lower priority alarm information will be compared and analyzed. If different additional information exists, this information will be retained as supplementary information; if the additional information does not have new value, the lower priority information will be directly deleted. This ensures that the system focuses on processing alarm information, reduces operator interference, and improves the efficiency of responding to emergencies.

[0056] In the multi-dimensional alarm light sign system designed in this invention, the data acquisition module is responsible for collecting alarm status, tag number information and other real-time data, and transmitting these data to the dimension judgment module; the dimension judgment module filters and judges the alarm data according to the configuration information transmitted by the configuration module; the alarm module analyzes the final alarm data, generates the display status of the alarm light sign and transmits it to the display module.

[0057] In one embodiment, the present invention is implemented in the following manner:

[0058] First, enter the configuration phase and classify the alarm statistics dimensions:

[0059] 1. Alarm Zoning: A DCS control system is often divided into several sections, each of which can be one or more zones. When configuring tag numbers, tag number zoning (also called tag number grouping) attributes need to be set. Alarm zoning can be tag number zoning or custom alarm zoning, categorizing tag numbers into various custom zone names according to actual needs. Alarm statistics configuration allows selection of one or more alarm zones.

[0060] 2. Flowchart: The flowchart is the most important monitoring and operation interface in a control system. It is used to display the overall process and working status of the controlled equipment and to manipulate related data. A flowchart often contains many graphical objects, and each graphical object is associated with many tag numbers. Alarm statistics configuration can select one flowchart.

[0061] 3. Tag number: One or more tag numbers can be selected through the system's tag number selector.

[0062] 4. Track Number Alarm Type: The track number alarm type is an extension of the track number dimension and is also the data source for underlying alarm statistics. To configure alarm statistics, first select a set of track numbers, then select the corresponding alarm type.

[0063] 5. Alarm statistics tag number: An alarm statistics tag number is a host computer tag number that is defined by engineers in the upper-level configuration. It groups related tag numbers and alarm types together, and a single alarm statistics tag number manages the alarm status of the entire group (whether there is an alarm, alarm level, etc.). To configure alarm statistics, you can first select an alarm statistics tag number.

[0064] 6. Alarm Levels: A DCS system can be configured with multiple alarm levels, such as low, medium, high, emergency, and security-related. Each alarm type will be assigned a corresponding alarm level. Alarm statistics configuration allows for the selection of one or more alarm levels.

[0065] 7. Alarm Types: A tag number can have multiple alarm types. Tag numbers of the same type will have the same alarm type, and tag numbers of different types may also have the same alarm type. Alarm statistics configuration allows you to select one or more alarm types.

[0066] After further classifying the alarm statistics dimensions, the process enters the runtime phase. The multi-dimensional alarm statistics data flow during runtime is as follows: Read the alarm statistics configuration, obtain all system alarms, and determine the statistical dimension: If it is an alarm partition, obtain the alarm partition and retrieve all related tag numbers through the alarm partition. If it is a flowchart, obtain the flowchart name and retrieve all related tag numbers through the flowchart parsing configuration; if it is a tag number, directly obtain the selected tag number. If it is a tag number alarm type, directly obtain the tag number alarm type. If it is an alarm statistics tag number, obtain all tag number alarm types through the alarm statistics tag number.

[0067] If it is an alarm level, then retrieve the selected alarm level. If it is an alarm type, then retrieve the selected alarm type.

[0068] If the alarm filtering type is tag number, then filter according to the tag number obtained above to filter out alarms that meet the conditions. If the alarm filtering type is tag number alarm type, then filter according to the tag number alarm type obtained above to filter out alarms that meet the conditions. If the alarm filtering type is alarm level, then filter according to the alarm level obtained above to filter out alarms that meet the conditions. If the alarm filtering type is alarm type, then filter according to the alarm type obtained above to filter out alarms that meet the conditions.

[0069] The filtered alarms are statistically analyzed to obtain information such as whether there is an alarm, the highest alarm level, the lowest alarm level, whether all alarms have been confirmed, and whether there is the same alarm event.

[0070] In another embodiment, the system portion of the present invention comprises the following parts:

[0071] 1. Data Acquisition Module: Responsible for acquiring real-time data from field I / O devices, including the status and alarm information of various sensors and actuators. Main components include:

[0072] The I / O module collects raw signals from the device, including data such as pressure, temperature, and current.

[0073] Data transmission interface, used to transmit acquired data to the DCS host (e.g., via communication bus or industrial Ethernet).

[0074] Sensors and actuators connected to the industrial site serve as inputs, and the collected alarm status, tag number information, and other real-time data are transmitted to the DCS host module as outputs.

[0075] 2. The statistical dimension module is responsible for determining the statistical dimensions of alarm data based on the rules configured on the engineering workstation (partitions, flowcharts, tag numbers, alarm levels, etc.), and filtering alarm information sets that meet the criteria. It mainly includes:

[0076] The dimension parsing unit parses the alarm configuration and determines the alarm statistics dimensions (such as partitions, flowcharts, etc.).

[0077] The data filtering unit filters real-time alarm data based on the dimensional analysis results.

[0078] The data output interface transmits the filtered alarm dataset to subsequent modules.

[0079] It receives the statistical dimension configuration provided by the engineer station module or the host and the input of real-time alarm data transmitted by the data acquisition module, and then passes the filtered alarm data to the alarm statistics module.

[0080] 3. The alarm statistics module performs statistical analysis on the alarm data filtered by the statistics dimension module, generating summary information on alarm status (such as whether an alarm exists, the highest level alarm, etc.). This mainly includes:

[0081] The data statistics unit performs statistics on alarm data, calculating the number of alarms, confirmation status, etc.

[0082] The status generation unit generates the display status of the alarm light panel (such as fast flashing, slow flashing, alarm level, etc.) based on statistical results; the data output interface outputs statistical results for the operator station or other modules to use.

[0083] It receives alarm data filtered by the judgment and statistical dimension module and transmits the alarm statistical results to the display module (operator station).

[0084] 4. The Engineer's Station module provides configuration tools, allowing users to define alarm statistics rules and dimensions. Engineers can use this module to configure alarm zones, flowcharts, tag numbers, alarm levels, etc. It mainly includes:

[0085] The configuration interface provides a user interface that supports the definition and grouping of alarm dimensions.

[0086] The data storage unit stores the configuration rules into the system and passes them to the statistical dimension module;

[0087] The data interaction interface communicates with the DCS host and other modules to update configuration settings.

[0088] 5. Display module (operator station): Displays the real-time status of alarm indicator panels in the control room for operator monitoring and rapid response. Supports displaying alarm level, alarm quantity, confirmation status, and other information. Main components include:

[0089] The display interface is a graphical interface that shows the alarm light bar, including alarm status, color, flashing mode, etc.

[0090] The user interaction unit allows operators to confirm alarms, view historical alarm records, etc.

[0091] The data receiving interface receives alarm statistics results from the alarm statistics module.

[0092] While receiving the alarm status calculated by the alarm statistics module, the operator can also provide feedback to the system through the operation interface (such as confirming the alarm).

[0093] 6. Data storage module: Stores the system's historical alarm data, configuration rules, and statistical results for subsequent querying, analysis, and tracing. Main components include:

[0094] Historical data storage unit records alarm information (such as alarm time, level, confirmation status, etc.);

[0095] The configuration rule storage unit saves the alarm dimension configurations defined by the engineer station or host.

[0096] The interface module provides data access services for operator stations, engineer stations, and other similar facilities.

[0097] This invention presents a multi-dimensional alarm light bar generation method and system with significant advantages. It can greatly improve the flexibility and accuracy of alarm management by statistically analyzing and filtering alarm information from multiple dimensions. Users can freely configure the system based on various dimensions such as alarm zones, flowcharts, tag numbers, and alarm types to adapt to complex industrial control scenarios. It effectively categorizes similar or related alarm information, reduces data clutter, and improves readability. By combining time windows with a similarity calculation model, it avoids repeated triggering of the same alarm event, reducing false alarms and missed alarms, and further enhancing the intelligence level of alarm management. A priority mechanism ensures the priority presentation of key alarm information, focusing on important content and temporarily suppressing useless information, improving processing efficiency and simplicity. The modular, multi-dimensional configuration process significantly simplifies configuration work, greatly reduces manpower requirements, and supports rapid deployment and template-based management, making it suitable for complex industrial environments. The system adopts a dynamic visualization design, adjusting the light bar display status in real time based on alarm statistics, helping users quickly grasp the operating status of production equipment and improve management efficiency. Its highly adaptable design makes it widely applicable to distributed control systems (DCS) in industrial fields such as thermal power, petrochemical, and chemical industries, providing users with flexible, efficient, and secure alarm processing solutions, and comprehensively optimizing alarm configuration efficiency and information management capabilities.

[0098] like Figure 2 In one embodiment shown, Figure 2 This invention presents a flowchart illustrating the same alarm event judgment process of a multi-dimensional alarm light display generation method and system. The multi-dimensional alarm light display generation method designed in this invention includes the entire process of configuring, reading, filtering, and processing alarm information. Specifically, the alarm configuration is first configured based on the tag number and alarm information. In this step, the relevant tag numbers in the DCS system are associated with the alarm information, and the priorities of different alarm configurations are set to ensure that the system can process and display alarm data according to the set logic. The priority setting allows the system to prioritize the presentation of key alarm information in complex multi-dimensional alarm scenarios, while secondary information is reserved as supplementary or backup data.

[0099] Next, the system can read the configured alarm settings and obtain all alarm information. This step aims to provide complete data support for subsequent alarm filtering and processing. By reading the system's real-time alarm data, it ensures that all possible abnormal states can be captured and sent to the filtering and analysis modules for further processing.

[0100] After reading the alarm information, the system filters alarms according to the configured multi-dimensional alarm configuration and the alarm dimension type. Alarm filtering dimensions can include tag number, tag number alarm type, alarm zone, flowchart, alarm level, and alarm type. Taking tag number filtering as an example, the system filters alarm data based on the configured tag number list, retaining only alarm records related to the target tag number. For the flowchart dimension, the system parses the flowchart configuration file, extracts relevant tag number information, and then filters alarm records accordingly. If the dimension is alarm level or alarm type, the system directly extracts alarm records of the corresponding alarm level and alarm type, and removes other irrelevant information.

[0101] After filtering out the target alarm information, the system further processes it, including time window judgment and similarity calculation to identify whether they are the same alarm event. The core of time window judgment is to sort all filtered alarm information according to timestamps and calculate the time difference between each alarm record and subsequent records. If the time difference is within a preset time window, it is considered likely to be the same alarm event, and the system proceeds to the next step of similarity calculation; otherwise, it continues to the next record. Similarity calculation involves analyzing the content of the alarm text, specifically including the following steps: The first step is to preprocess the alarm text, including removing meaningless stop words and punctuation marks, and stemming words to convert the alarm text into a standardized word set. Then, the system calculates the size of the intersection and union of the word sets of two alarm texts, and uses the size of the intersection divided by the size of the union to calculate the first similarity. If the first similarity is lower than a set threshold, the system continues to calculate the second similarity; otherwise, it is directly determined to be the same alarm event.

[0102] The second step is to convert the alarm text into a vector representation and calculate the second similarity using the dot product and modulus of the vectors. When the value of the second similarity is higher than the threshold, the alarm texts are identified as the same alarm event; otherwise, the system treats them as different alarm events and returns to the time window judgment step to start calculating the next set of alarm records.

[0103] After filtering alarm events, the system processes the same alarm event according to its priority. The highest priority alarm information will be displayed and recorded as the primary alarm, while lower priority alarm information will be analyzed to see if it contains any additional information. If additional information is found, it will be retained as supplementary information; otherwise, the lower priority alarm information will be directly deleted, thereby simplifying alarm data and improving the efficiency of operator attention.

[0104] In one embodiment, the size of the time window is first determined, and a reasonable time window is determined based on system requirements and actual conditions.

[0105] In this embodiment, the time interval is set to 5 minutes. Next, alarm information is acquired. Alarm information is received in real time within the system, and each alarm message should include a timestamp. The received alarm messages are sorted according to their timestamps to ensure easy comparison within the time range. For each alarm record, the difference between its timestamp and those of other alarm records is checked. Alarm records within the time window are marked for subsequent merging processing.

[0106] Then, the alarm content undergoes text preprocessing, removing stop words and punctuation marks, and performing stemming. Each alarm content is converted into a word set, resulting in A and B. The size of the intersection and union of the two sets is calculated, and the similarity between the two alarm contents is calculated using the following formula:

[0107]

[0108] The similarity score is compared to a threshold. If the events are determined not to be the same, a second similarity comparison is performed.

[0109] Convert the alarm content into a vector representation to obtain the feature vectors A and B for each alarm, and calculate the dot product of vectors A and B: A·B.

[0110] Calculate the magnitudes of vectors A and B:

[0111]

[0112] The second similarity is calculated as follows:

[0113] This invention offers flexible, multi-dimensional alarm management capabilities, supporting configurations across alarm zones, flowcharts, tag numbers, alarm levels, and other dimensions. It adapts to the diverse needs of complex industrial scenarios, helping operators quickly locate target alarm information and improving management efficiency. Simultaneously, by utilizing time window judgment and similarity calculation models, it effectively avoids repeated triggering of the same alarm event, accurately analyzes alarm content, reduces false alarms and missed alarms, and significantly enhances the intelligence level of alarm management. A priority mechanism further optimizes the information processing flow, making key alarm information intuitively presented, while secondary information is filtered to retain useful content, resulting in simplicity and efficiency. Furthermore, this invention supports modular and template-based batch configuration functions, significantly reducing configuration time and labor costs, and achieving standardized management, making it particularly suitable for large-scale industrial control systems. Through real-time data reading and dynamic analysis, the system possesses strong real-time performance and broad compatibility, applicable to various DCS control scenarios such as thermal power, petrochemical, and chemical industries. In addition, the alarm indicator lights, combined with dynamic adjustment and intuitive display, significantly improve the user experience, helping to quickly grasp the system's operating status and avoid confusion caused by information overload. Intelligent analysis and efficient configuration further reduce the workload and error rate of operators, providing an efficient, accurate and safe alarm management solution for complex industrial environments.

[0114] This invention is not limited to the above-described embodiments. Any changes made to its shape or material composition, or any structural design using the methods provided by this invention, are considered variations of this invention and should be considered within the scope of protection of this invention.

Claims

1. A method for generating multi-dimensional alarm light signs, characterized in that, Includes the following steps: S1: Configure alarm configurations based on tag numbers and alarm information, and set the priority of each alarm configuration. Multiple dimensions include alarm zones, flowcharts, tag numbers, tag number alarm types, alarm levels, and alarm types. S2: Read alarm configuration and obtain all system alarms; S3: Determine the dimension and perform filtering, calculate the similarity of alarms within the time window, process the filtered alarms, and determine whether there are the same alarm events by comparing the time window with the similarity model. S4: For the same alarm event, the alarm will be triggered according to the information configured for the high-priority alarm, and the information configured for the low-priority alarm will be used as a supplement.

2. The method for generating a multi-dimensional alarm light sign according to claim 1, characterized in that, Step S3 further includes the following steps: S3.1: Perform corresponding filtering based on the filtering type, such as tag number, tag number alarm type, alarm level, or alarm type; S3.2: Set the required time window and obtain the timestamp of each alarm message; S3.3: Sort the received alarm information according to the timestamp, and check the difference between the timestamp of the alarm record and the timestamp of the next alarm record in sequence; S3.4: Determine whether the timestamp difference is within the time window range. If yes, proceed to step S3.5; otherwise, return to step S3.3 and calculate the next timestamp difference. S3.5: Process the alarm text and calculate its similarity.

3. The method for generating a multi-dimensional alarm light sign according to claim 2, characterized in that, Step S3.5 specifically includes the following steps: S3.51: Perform text preprocessing on the alarm content, removing stop words and punctuation, and extracting stems; S3.52: Convert each alarm message into a word set, and calculate the size of the intersection and union of the two sets; S3.53: Calculate the value of the first similarity and determine whether the first similarity is less than the first similarity threshold. If yes, proceed to step S3.54; otherwise, determine that they are the same alarm event. S3.54: Convert the alarm content into a vector representation to obtain the feature vector of each alarm; S3.55: Calculate the dot product and magnitude of the two vectors, and calculate the second similarity value; S3.56: Determine whether the second similarity is less than the second similarity threshold. If yes, then it is not the same alarm event and return to step S3.

3. If no, then it is determined to be the same alarm event.

4. A method for generating multi-dimensional alarm light signs according to claim 1, 2, or 3, characterized in that, In step S4, if the alarm events are determined to be the same, the alarm information with the highest priority is selected and an alarm is triggered. It is then determined whether the alarm information with lower priority has any additional information compared to the alarm information with the highest priority. If it does, it is retained as supplementary information; if it does not, the alarm information with lower priority is deleted.

5. The method for generating a multi-dimensional alarm light sign according to claim 4, characterized in that, In step S1, configuring the alarm configuration according to the tag number includes configuring according to the alarm partition, flowchart and direct tag number. The alarm partition is a tag number partition or a custom alarm partition. The custom alarm partition classifies the tag number into various custom partition names according to actual needs. One or more alarm partitions can be selected. The flowchart includes several graphical objects, and the graphical objects are associated with several tag numbers.

6. The method for generating a multi-dimensional alarm light sign according to claim 5, characterized in that, The direct tag number is selected by the tag number selector of the system to select one or more tag numbers. If the alarm type corresponding to the tag number is selected, it is the tag number alarm type dimension. The alarm configuration according to the tag number also includes alarm tag numbers. The alarm tag numbers group the related tag number alarm types together, and the alarm tag numbers handle the alarm status of the entire group.

7. The method for generating a multi-dimensional alarm light sign according to claim 6, characterized in that, In step S1, configuring the alarm configuration according to the alarm information includes configuring according to the alarm level and alarm category. Each alarm type has a corresponding alarm level. A tag number has several alarm types. Tag numbers of the same type have the same alarm type, and tag numbers of different types can also have the same alarm type. One or more of the alarm level and alarm type can be selected.

8. A method for generating a multi-dimensional alarm light sign according to claim 5, 6, or 7, characterized in that, In step S3, if the dimension is an alarm partition, the alarm partition is obtained, and all the relevant tag numbers are obtained through the alarm partition; if the dimension is a flowchart, the flowchart name is extracted and the flowchart configuration file is parsed, and then the tag numbers are extracted. If multiple tag numbers are extracted from the flowchart, all the extracted tag numbers are integrated into a list, duplicate tag numbers are removed, and all the relevant tag numbers are obtained. If the dimension is a tag number, the selected tag number is obtained directly; if the dimension is a tag number alarm type, the tag number alarm type is obtained directly; if the dimension is an alarm tag number, all tag number alarm types are obtained through alarm statistics tag numbers; if the dimension is an alarm level, the selected alarm level is obtained; if the dimension is an alarm type, the selected alarm type is obtained.

9. The method for generating a multi-dimensional alarm light sign according to claim 3, characterized in that, In step S3.53, the specific method for calculating the value of the first similarity is as follows: the first similarity is the quotient of the size of the intersection of the two sets and the size of the size of the union of the two sets; the specific method for calculating the value of the second similarity is as follows: the second similarity is the quotient of the dot product of the two vectors and the product of the moduli of the two vectors.

10. A multi-dimensional alarm light sign system, employing the multi-dimensional alarm light sign generation method according to any one of claims 1-9, characterized in that, include: The system includes a data acquisition module, a dimension judgment module, and an alarm module. The data acquisition module collects alarm status and tag number information, as well as other real-time data, and transmits them to the dimension judgment module. The dimension judgment module receives configuration information transmitted by the configuration module and transmits the filtered alarm data to the alarm module. The alarm module filters and analyzes the alarm data and generates the final result to display the alarm light sign display status, which is then sent to the display module.

Images