A method and system for safe operation control based on artificial intelligence identification
Through a multi-level operation monitoring system based on artificial intelligence, the problem that traditional monitoring systems are difficult to identify abnormalities in production processes is solved, real-time abnormal monitoring and early warning are realized, accident risk is reduced, and production management and resource utilization are optimized.
Patent Information
- Application Number
- CN202411247018.1
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-09-06
- Publication Date
- 2024-12-06
- Estimated Expiration
- 2044-09-06
AI Technical Summary
It is difficult for traditional monitoring systems to achieve linkage abnormal identification coverage of the entire production process and individual categories of abnormal warnings, which makes it difficult to detect, locate and deal with abnormal situations in a timely manner, posing huge risks to the production of enterprises.
A multi-level operation monitoring system based on artificial intelligence provides a multi-level operation monitoring system that establishes the upper and lower hierarchical relationship between production link information and timing information, calculates the target abnormality index, virtually connects abnormal points, calculates the abnormal risk rate and hidden danger coefficient, and conducts safety warnings.
Real-time monitoring and early warning are realized, abnormal phenomena in the production process are discovered in a timely manner, the probability of accidents occur, the production management and resource allocation are optimized, the frequency and cost of manual inspections are reduced, and scientific and reasonable production planning and management strategies are provided.
Smart Images

Figure CN118760873B_ABST
Abstract
Description
Technical Field
[0001] The present invention proposes a method and system for safe operation control based on artificial intelligence recognition, and relates to the technical field of safe operation control based on artificial intelligence recognition. Background Art
[0002] With the continuous development of industrial production, the safety and efficiency of the production process have become the focus of enterprises. Traditional monitoring systems can often only monitor a single production link or equipment, and it is difficult to achieve linkage abnormal identification coverage of the entire production process and abnormal warning of a single category. At the same time, due to the lack of an effective abnormal identification and processing mechanism, it is difficult to timely discover, locate and handle abnormal situations, which brings great risks to the production of enterprises. Summary of the invention
[0003] The present invention provides a method and system for safe operation control based on artificial intelligence identification, which is used to solve the problem that traditional monitoring systems can only monitor a single production link or equipment, and it is difficult to achieve linkage abnormal identification coverage of the entire production process and abnormal warning of a single category. At the same time, due to the lack of an effective abnormal identification and processing mechanism, it is difficult to timely discover, locate and handle abnormal situations, which brings great risks to the production of enterprises.
[0004] The present invention proposes a method and system for controlling safe operations based on artificial intelligence identification, the method comprising:
[0005] S1. Establish a multi-level operation monitoring system based on production link information, establish the upper and lower hierarchical relationship of monitoring types based on production environment time series information, and obtain hierarchical monitoring data;
[0006] S2, obtaining hierarchical monitoring data of multi-level identification targets, calculating the target anomaly index of each hierarchical identification target, determining target abnormal points and target normal points according to the target anomaly index, virtually connecting the target abnormal points, and obtaining data anomaly circles;
[0007] S3. Calculate the abnormal hidden danger rate of the data abnormal circle, obtain abnormal category information according to the abnormal hidden danger rate, calculate the hidden danger risk coefficient of each abnormal category information, and issue a safety warning for each abnormal category according to the hidden danger risk coefficient.
[0008] Furthermore, the S1 includes:
[0009] Acquire production link information, classify monitoring type information according to the production link information, and obtain a multi-level operation monitoring system;
[0010] Obtaining production link timing information, and determining the number of levels for each monitoring type based on the production link timing information;
[0011] Establishing a hierarchical relationship between each monitoring category according to the number of levels of each monitoring category to obtain a monitoring hierarchy chain;
[0012] The monitoring data of each monitoring level of the monitoring level chain is acquired, and the monitoring data is preprocessed to obtain the preprocessed level monitoring data.
[0013] Further, the S2 includes:
[0014] Acquire a multi-level recognition target, match the multi-level recognition target with a multi-level operation monitoring system at each level, and obtain a level matching relationship;
[0015] Obtaining hierarchical monitoring data of each hierarchical identification target through the hierarchical matching relationship;
[0016] Calculate the target anomaly index of each level identification target according to the level monitoring data combined with the multi-level identification target information;
[0017] According to the target abnormality index, abnormal point determination is performed on each level identification target to obtain multiple target abnormal points and target normal points;
[0018] Virtually connect multiple target anomaly points, obtain the maximum connection perimeter or side length, and obtain the data anomaly circle.
[0019] Further, the S3 includes:
[0020] Obtaining target abnormal points within the data abnormal circle, and calculating the abnormal potential risk rate of the data abnormal circle according to the target abnormal points within the data abnormal circle combined with target normal point information;
[0021] Obtaining abnormality category information according to the abnormal potential risk rate combined with target abnormal point information;
[0022] A corresponding category warning is triggered according to the abnormal category information, and a warning is issued according to the corresponding category warning.
[0023] Further, triggering a corresponding category warning according to the abnormal category information, and issuing a warning according to the corresponding category warning, includes:
[0024] Calculate the proportion of target anomaly points of each type of anomaly information among all target anomaly points of all anomaly category information;
[0025] Multiply the percentage by the abnormal hidden danger rate to obtain the hidden danger risk coefficient corresponding to each abnormal category information;
[0026] Compare the hidden danger risk coefficient of each abnormal category information with the corresponding preset hidden danger risk threshold to obtain an abnormal comparison result;
[0027] A security warning is issued for the corresponding abnormal category according to the abnormal comparison result.
[0028] Furthermore, the system comprises:
[0029] The multi-level monitoring module is used to establish a multi-level operation monitoring system based on the production link information, establish the upper and lower hierarchical relationship of monitoring types based on the production environment time series information, and obtain hierarchical monitoring data;
[0030] The abnormal range acquisition module is used to obtain the hierarchical monitoring data of the multi-level identification target, calculate the target abnormality index of each hierarchical identification target, determine the target abnormal point and the target normal point according to the target abnormality index, virtually connect the target abnormal point, and obtain the data abnormal circle;
[0031] The classification warning module is used to calculate the abnormal hidden danger rate of the data abnormal circle, obtain the abnormal category information according to the abnormal hidden danger rate, calculate the hidden danger risk coefficient of each abnormal category information, and issue a safety warning for each abnormal category according to the hidden danger risk coefficient.
[0032] Furthermore, the multi-level monitoring module includes:
[0033] A hierarchical classification module is used to obtain production link information, classify monitoring type information according to the production link information, and obtain a multi-level operation monitoring system;
[0034] The upper and lower level determination module is used to obtain the production link timing information and determine the number of levels of each monitoring type according to the production link timing information;
[0035] Establishing a hierarchical relationship between each monitoring category according to the number of levels of each monitoring category to obtain a monitoring hierarchy chain;
[0036] The data acquisition module is used to obtain monitoring data of each monitoring level of the monitoring level chain, pre-process the monitoring data, and obtain pre-processed level monitoring data.
[0037] Furthermore, the abnormal range acquisition module includes:
[0038] A target data matching module is used to obtain a multi-level recognition target, match the multi-level recognition target with a multi-level operation monitoring system at each level, and obtain a level matching relationship;
[0039] Obtaining hierarchical monitoring data of each hierarchical identification target through the hierarchical matching relationship;
[0040] An abnormality index calculation module, used to calculate the target abnormality index of each level identification target according to the level monitoring data combined with the multi-level identification target information;
[0041] An abnormality determination module, used for determining abnormal points of each level identification target according to the target abnormality index, and obtaining a plurality of target abnormal points and target normal points;
[0042] The abnormal circle acquisition module is used to virtually connect multiple target abnormal points, obtain the maximum connection perimeter or side length, and obtain the data abnormal circle.
[0043] Furthermore, the classification warning module includes:
[0044] A hidden danger rate calculation module is used to obtain a target abnormal point in a data abnormal circle, and calculate the abnormal hidden danger rate of the data abnormal circle according to the target abnormal point in the data abnormal circle combined with the target normal point information;
[0045] A category information acquisition module, used to obtain abnormal category information according to the abnormal hidden danger rate combined with target abnormal point information;
[0046] The category warning module is used to trigger a corresponding category warning according to the abnormal category information and issue a warning according to the corresponding category warning.
[0047] Furthermore, the category warning module includes:
[0048] A category risk coefficient acquisition module is used to calculate the proportion of target abnormal points of each abnormal category information among the target abnormal points of all abnormal category information;
[0049] Multiply the percentage by the abnormal hidden danger rate to obtain the hidden danger risk coefficient corresponding to each abnormal category information;
[0050] A comparison warning module is used to compare the hidden danger risk coefficient of each abnormal category information with the corresponding preset hidden danger risk threshold to obtain an abnormal comparison result;
[0051] A security warning is issued for the corresponding abnormal category according to the abnormal comparison result.
[0052] Beneficial effects of the present invention: Through real-time monitoring and early warning, abnormal phenomena in the production process can be discovered in time, reducing the probability of accidents. Through hierarchical monitoring of the production process, the operation of each production link can be more clearly understood, which is helpful to optimize production management and resource allocation. Once an abnormal phenomenon is detected, the system can quickly generate early warning information and notify relevant personnel, thereby improving the response speed to abnormal situations. Through automated and intelligent monitoring methods, the frequency and cost of manual inspections are reduced, and the operation and maintenance efficiency is improved. The monitoring data and analysis results collected by the system can provide strong support for the company's decision-making, helping the company to formulate more scientific and reasonable production plans and management strategies. BRIEF DESCRIPTION OF THE DRAWINGS
[0053] Figure 1 A schematic diagram of a safe operation control method based on artificial intelligence identification;
[0054] Figure 2 Schematic diagram of data anomaly circle. DETAILED DESCRIPTION
[0055] The preferred embodiments of the present invention are described below in conjunction with the accompanying drawings. It should be understood that the preferred embodiments described herein are only used to illustrate and explain the present invention, and are not used to limit the present invention.
[0056] In one embodiment of the present invention, a method and system for controlling safe operation based on artificial intelligence identification is proposed, and the method comprises:
[0057] S1. Establish a multi-level operation monitoring system based on production link information, establish the upper and lower hierarchical relationship of monitoring types based on production environment time series information, and obtain hierarchical monitoring data;
[0058] S2, obtaining hierarchical monitoring data of multi-level identification targets, calculating the target anomaly index of each hierarchical identification target, determining target abnormal points and target normal points according to the target anomaly index, virtually connecting the target abnormal points, and obtaining data anomaly circles;
[0059] S3. Calculate the abnormal hidden danger rate of the data abnormal circle, obtain abnormal category information according to the abnormal hidden danger rate, calculate the hidden danger risk coefficient of each abnormal category information, and issue a safety warning for each abnormal category according to the hidden danger risk coefficient.
[0060] The working principle of the above technical solution is as follows: according to the information of the production link, the production process is divided into multiple levels, and each level represents a different production stage or operation link. Set corresponding monitoring points at each level, collect the operation data of the level, and form hierarchical monitoring data. Analyze the time series information of the production environment and understand the dependencies and impact paths between different monitoring types. Based on the time series information and the production process, establish the upper and lower hierarchical relationship of the monitoring types, and clarify which monitoring types have a direct impact on other types. Obtain hierarchical monitoring data of multi-level identification targets in real time through sensors, data interfaces or other data acquisition means. Sort and analyze these data according to the hierarchical structure. For each level of identification target, calculate its target anomaly index according to the preset anomaly judgment rules or algorithms. Compare the calculated target anomaly index with the preset threshold. If the anomaly index exceeds the threshold, the identification target is determined to be an abnormal point; otherwise, it is determined to be a normal point. Virtually connect the determined target abnormal points to form a data anomaly circle. The data anomaly circle represents the distribution and diffusion of abnormal phenomena in the production process. Calculate the abnormal hidden danger rate. The abnormal hidden danger rate reflects the severity and potential risk of abnormal phenomena in the abnormal circle. According to the abnormal hidden danger rate and the specific characteristics of the abnormal point, the abnormal category information is determined. For each abnormal category, the hidden danger risk coefficient is calculated. According to the level of the hidden danger risk coefficient, a safety warning is issued for each abnormal category. The warning information can be sent to relevant personnel through the alarm system, notification, etc., so that timely response measures can be taken.
[0061] The technical effect of the above technical solution is: through real-time monitoring and early warning, abnormal phenomena in the production process can be discovered in time, reducing the probability of accidents. Through hierarchical monitoring of the production process, the operation of each production link can be more clearly understood, which is helpful to optimize production management and resource allocation. Once an abnormal phenomenon is detected, the system can quickly generate early warning information and notify relevant personnel, thereby improving the response speed to abnormal situations. Through automated and intelligent monitoring methods, the frequency and cost of manual inspections are reduced, and the operation and maintenance efficiency is improved. The monitoring data and analysis results collected by the system can provide strong support for the company's decision-making and help the company formulate more scientific and reasonable production plans and management strategies.
[0062] In one embodiment of the present invention, the S1 includes:
[0063] Acquire production link information, classify monitoring type information according to the production link information, and obtain a multi-level operation monitoring system; the monitoring types include monitoring of various operations in the workshop, monitoring of different rooms of production equipment, etc.
[0064] Obtain the timing information of the production links, and determine the number of levels for each monitoring type based on the timing information of the production links; the timing information of the production links is different production links at different times. For example, samples are produced at 5 o'clock, and the samples produced at 5 o'clock are tested at 6 o'clock. Then the samples produced at 5 o'clock and the tests at 6 o'clock are the upper and lower hierarchical relationship of the monitoring types.
[0065] Establishing a hierarchical relationship between each monitoring category according to the number of levels of each monitoring category to obtain a monitoring hierarchy chain;
[0066] The monitoring data of each monitoring level of the monitoring level chain is acquired, and the monitoring data is preprocessed to obtain the preprocessed level monitoring data.
[0067] The working principle of the above technical solution is as follows: collect specific information about the production process, including production process, operation tasks, equipment layout, etc. According to the production process information, classify the types that need to be monitored, such as monitoring of different operations in the workshop (such as assembly operations, packaging operations, etc.), monitoring of production equipment in different rooms (such as injection molding machines, stamping machines, etc.), etc. Obtain the timing information of the production process, that is, the activity sequence of different production processes at different time points. According to these timing information, determine the number of levels of each monitoring type. For example, in a process where a production sample is followed by a test, the production sample and the test can be regarded as two different levels of monitoring types. Based on the production process timing information, clarify the upper and lower hierarchical relationships between different monitoring types. These relationships reflect the dependencies and sequences between different links in the production process. Connect these relationships in series to form a monitoring hierarchy chain, each level corresponding to one or more monitoring types. For each monitoring level in the monitoring hierarchy chain, collect the corresponding monitoring data. These data may come from sensors, log records, operator input, etc. Preprocess the monitoring data, including data cleaning (removing noise, error values, etc.), data standardization (unifying measurement units, formats, etc.), data conversion (converting raw data into a format that can be used for analysis), etc. The preprocessed hierarchical monitoring data will be used for subsequent analysis and early warning.
[0068] The technical effect of the above technical solution is: through the multi-level operation monitoring system, the enterprise can clearly understand the status and changes of each link in the production process, so as to better grasp the production progress and efficiency. Based on the analysis and early warning of monitoring data, the enterprise can timely discover potential risks in the production process, such as equipment failure, abnormal operation, etc., and take corresponding measures to intervene and reduce losses. By monitoring the use of different production links and equipment, the enterprise can more reasonably allocate resources, such as manpower, material resources, financial resources, etc., so as to improve resource utilization efficiency. The data and analysis results provided by the multi-level operation monitoring system can provide strong support for the decision-making of the enterprise, and help the enterprise to formulate more scientific and reasonable production plans, management strategies, etc. Through real-time monitoring and early warning of the production links, the enterprise can timely discover and deal with possible safety hazards, so as to ensure the safety and stability of the production process.
[0069] In one embodiment of the present invention, S2 includes:
[0070] Acquire a multi-level recognition target, match the multi-level recognition target with the multi-level operation monitoring system at each level, and obtain a hierarchical matching relationship; the multi-level recognition target includes objects that need to be monitored at each level, such as laboratory operating procedures and production samples or production equipment in a production workshop, etc., specifically things that need to be monitored.
[0071] Obtaining hierarchical monitoring data of each hierarchical identification target through the hierarchical matching relationship;
[0072] Calculate the target anomaly index of each level identification target according to the level monitoring data combined with the multi-level identification target information;
[0073] The calculation formula of the target abnormality index is:
[0074]
[0075] in, The target anomaly index for the hierarchical identification target, The maximum value of the hierarchical monitoring data for the hierarchical identification target, The minimum value of the hierarchical monitoring data for the hierarchical identification target, Preset monitoring data for tier identification targets.
[0076] According to the target abnormality index, abnormal point determination is performed on each level identification target to obtain multiple target abnormal points and target normal points; the target abnormality index is compared with a preset index threshold, and when the target abnormality index is greater than the preset index threshold, the identification target is determined to be a target abnormal point, otherwise it is a target normal point.
[0077] Virtually connect multiple target anomaly points, obtain the maximum connection perimeter or side length, and obtain the data anomaly circle.
[0078] The working principle of the above technical solution is: determine multi-level recognition targets, which represent specific objects or matters that need to be monitored at each level, such as laboratory operation procedures, production samples or production equipment in the production workshop, etc. Match these multi-level recognition targets with the previously established multi-level operation monitoring system to ensure that each recognition target can correspond to the corresponding monitoring level to form a hierarchical matching relationship. Through the hierarchical matching relationship, the system can accurately obtain the real-time monitoring data of each level recognition target. According to the acquired hierarchical monitoring data, the target anomaly index of each level recognition target is calculated. This index reflects the degree of deviation of the current monitoring data from the normal state. The system compares the calculated target anomaly index with the preset index threshold. If the target anomaly index exceeds the threshold, the system determines that the recognition target is an abnormal point; otherwise, it is determined to be a normal point. In this way, the system can identify potential problems or abnormal situations in the production process.
[0079] For multiple identified target anomalies, the system uses virtual connection technology to connect them. This connection can be based on wireless network connections or logical connections based on production processes. The system calculates the maximum perimeter or side length of these connections to form a data anomaly circle. This anomaly circle represents the distribution and diffusion of anomalies in the production process, which helps companies quickly locate problem areas and assess potential risks.
[0080] The technical effect of the above technical solution is: by matching the multi-level identification target with the multi-level operation monitoring system, the system can more accurately obtain the real-time monitoring data of each level identification target, and calculate the target anomaly index based on these data, thereby improving the accuracy of anomaly identification. The system can calculate the target anomaly index in real time and make anomaly point judgments. Once an anomaly point is identified, it can immediately trigger the corresponding early warning mechanism to help enterprises achieve rapid response and processing. By constructing a data anomaly circle, enterprises can clearly see the distribution and diffusion of anomalies in the production process, so as to allocate resources in a targeted manner, optimize production processes, and improve production efficiency. The data anomaly circle not only shows the distribution of current anomalies, but also reflects the diffusion trend and scope of potential risks, providing enterprises with more comprehensive and accurate risk warning information. Through real-time monitoring and anomaly identification, the system can promptly discover and deal with safety hazards in the production process, reduce the probability of accidents, and improve production safety.
[0081] In one embodiment of the present invention, S3 includes:
[0082] Obtaining target abnormal points within the data abnormal circle, and calculating the abnormal potential risk rate of the data abnormal circle according to the target abnormal points within the data abnormal circle combined with target normal point information;
[0083] The calculation formula of the abnormal hidden danger rate is:
[0084]
[0085] in, is the abnormal potential risk rate, c is the number of levels contained in the data abnormal circle, is the number of target anomaly points at the i-th level, is the number of target normal points at the i-th level;
[0086] The larger the number of levels, the more important the monitoring types are, because the later monitoring types rely on the data of the earlier monitoring types.
[0087] The abnormality category information is obtained according to the abnormality hidden danger rate combined with the target abnormal point information; the abnormality hidden danger rate is compared with the preset hidden danger threshold, and when the abnormality hidden danger rate is greater than the preset hidden danger threshold, the monitoring category of each target abnormal point in the data abnormal circle is obtained, otherwise the monitoring category is not obtained. The monitoring category at this time is the abnormality category information.
[0088] A corresponding category warning is triggered according to the abnormal category information, and a warning is issued according to the corresponding category warning.
[0089] The working principle of the above technical solution is as follows: the system identifies and determines all target abnormal points within the data abnormal circle. These abnormal points are determined by the previous analysis steps and represent potential problems or abnormal situations in the production process. The system calculates the abnormal hidden danger rate of the data abnormal circle based on the number and severity of the target abnormal points within the data abnormal circle and the information of the target normal points. This hidden danger rate is a quantitative indicator that reflects the potential risk of the abnormal situation and the potential impact on the production process. According to the abnormal hidden danger rate and the specific information of the target abnormal point, the system analyzes and determines the abnormal category information. These abnormal category information may include equipment failure, production process abnormality, operation error, etc., depending on the production environment and monitoring requirements. The system compares the abnormal hidden danger rate with the preset hidden danger threshold. If the abnormal hidden danger rate exceeds the threshold, the system triggers the corresponding warning level. The warning level is usually divided according to the severity of the hidden danger and the potential impact on the production process, such as low-level warning, intermediate warning, high-level warning, etc. Once the warning level is triggered, the system will perform warning operations according to the preset warning mechanism. This may include sending alarm information to relevant personnel, starting emergency response procedures, adjusting production plans, etc. The purpose of early warning is to notify relevant personnel in a timely manner and take appropriate measures to deal with abnormal situations to reduce potential losses. When it is a negative number, it is considered that the abnormal hidden danger rate is less than the preset hidden danger threshold.
[0090] The technical effect of the above technical solution is: by calculating the abnormal hidden danger rate and determining the abnormal category information, the system can more accurately assess the potential risks in the production process and trigger the corresponding warning level. This helps enterprises to discover and deal with abnormal situations in a timely manner and reduce potential losses. The system can allocate resources such as manpower, material resources, and financial resources in a targeted manner according to the warning level and abnormal category information. This helps enterprises to deal with abnormal situations more efficiently and improve production efficiency and safety. Through real-time monitoring and early warning mechanisms, the system can timely discover safety hazards in the production process and take corresponding measures to deal with them. This helps to reduce the probability of accidents and improve production safety. The data and analysis results provided by the early warning mechanism can provide strong support for the decision-making of enterprises. Enterprises can formulate more scientific and reasonable production plans, management strategies, etc. based on early warning information to improve production efficiency and competitiveness. The system can calculate the abnormal hidden danger rate and trigger the early warning level in real time to achieve rapid response. This helps enterprises to take timely measures to deal with abnormal situations and reduce potential losses and downtime.
[0091] In one embodiment of the present invention, triggering a corresponding category warning according to the abnormal category information, and issuing a warning according to the corresponding category warning, includes:
[0092] Calculate the proportion of target anomaly points of each type of anomaly information among all target anomaly points of all anomaly category information;
[0093] Multiply the percentage by the abnormal hidden danger rate to obtain the hidden danger risk coefficient corresponding to each abnormal category information;
[0094] The hidden danger risk coefficient of each abnormal category information is compared with the corresponding preset hidden danger risk threshold to obtain the abnormal comparison result; when the hidden danger risk coefficient is greater than the corresponding preset hidden danger risk threshold, a safety warning is issued for the corresponding abnormal category, otherwise no warning is issued.
[0095] A security warning is issued for the corresponding abnormal category according to the abnormal comparison result.
[0096] The working principle of the above technical solution is as follows: the system counts the number of target abnormal points of each abnormal category information, and calculates the proportion of target abnormal points of each abnormal category information in the target abnormal points of all abnormal category information. This proportion reflects the relative importance or frequency of each abnormal category in the current production process. The system multiplies the proportion of each abnormal category information with the abnormal hidden danger rate calculated previously to obtain the hidden danger risk coefficient corresponding to each abnormal category information. The hidden danger risk coefficient is a comprehensive indicator that takes into account both the proportion of abnormal categories (i.e., its importance or frequency) and the abnormal hidden danger rate of the entire production process. The system presets the hidden danger risk threshold corresponding to each abnormal category. Then, the system compares the hidden danger risk coefficient of each abnormal category information with the corresponding preset hidden danger risk threshold. If the hidden danger risk coefficient exceeds the threshold, the system triggers a safety warning for the corresponding abnormal category; otherwise, no warning is issued. Once the safety warning is triggered, the system will perform corresponding operations according to the preset warning mechanism and level. This may include sending alarm information to relevant personnel, starting emergency response procedures, adjusting production plans, etc. The purpose of the warning is to notify relevant personnel in a timely manner and take corresponding safety measures to deal with or avoid potential risks.
[0097] The technical effect of the above technical solution is: by calculating the hidden danger coefficient of each abnormal category information and comparing it with the preset hidden danger threshold, the system can achieve accurate early warning and targeted processing of different abnormal categories. This helps to improve the accuracy and effectiveness of the early warning and reduce unnecessary false alarms and missed reports. Since the system can identify the importance and degree of danger of each abnormal category, enterprises can allocate resources such as manpower, material resources, and financial resources more targetedly according to the early warning information. This helps enterprises to deal with abnormal situations more efficiently and improve production efficiency and safety. Through real-time monitoring and early warning mechanisms, the system can timely discover potential risks in the production process and trigger corresponding safety warnings. This helps enterprises to take timely measures to deal with or avoid potential risks, reduce the probability of accidents, and improve production safety. The data and analysis results provided by the early warning mechanism can provide strong support for the decision-making of enterprises. Enterprises can formulate more scientific and reasonable production plans, management strategies, etc. based on the early warning information to improve production efficiency and competitiveness. The system achieves rapid response and early warning by automatically calculating and comparing the hidden danger coefficient with the preset threshold. This helps enterprises to take timely measures to deal with abnormal situations and reduce potential losses and downtime.
[0098] In one embodiment of the present invention, the system comprises:
[0099] The multi-level monitoring module is used to establish a multi-level operation monitoring system based on the production link information, establish the upper and lower hierarchical relationship of monitoring types based on the production environment time series information, and obtain hierarchical monitoring data;
[0100] The abnormal range acquisition module is used to obtain the hierarchical monitoring data of the multi-level identification target, calculate the target abnormality index of each hierarchical identification target, determine the target abnormal point and the target normal point according to the target abnormality index, virtually connect the target abnormal point, and obtain the data abnormal circle;
[0101] The classification warning module is used to calculate the abnormal hidden danger rate of the data abnormal circle, obtain the abnormal category information according to the abnormal hidden danger rate, calculate the hidden danger risk coefficient of each abnormal category information, and issue a safety warning for each abnormal category according to the hidden danger risk coefficient.
[0102] The working principle of the above technical solution is as follows: according to the information of the production link, the production process is divided into multiple levels, and each level represents a different production stage or operation link. Set corresponding monitoring points at each level, collect the operation data of the level, and form hierarchical monitoring data. Analyze the time series information of the production environment and understand the dependencies and impact paths between different monitoring types. Based on the time series information and the production process, establish the upper and lower hierarchical relationship of the monitoring types, and clarify which monitoring types have a direct impact on other types. Obtain hierarchical monitoring data of multi-level identification targets in real time through sensors, data interfaces or other data acquisition means. Sort and analyze these data according to the hierarchical structure. For each level of identification target, calculate its target anomaly index according to the preset anomaly judgment rules or algorithms. Compare the calculated target anomaly index with the preset threshold. If the anomaly index exceeds the threshold, the identification target is determined to be an abnormal point; otherwise, it is determined to be a normal point. Virtually connect the determined target abnormal points to form a data anomaly circle. The data anomaly circle represents the distribution and diffusion of abnormal phenomena in the production process. Calculate the abnormal hidden danger rate. The abnormal hidden danger rate reflects the severity and potential risk of abnormal phenomena in the abnormal circle. According to the abnormal hidden danger rate and the specific characteristics of the abnormal point, the abnormal category information is determined. For each abnormal category, the hidden danger risk coefficient is calculated. According to the level of the hidden danger risk coefficient, a safety warning is issued for each abnormal category. The warning information can be sent to relevant personnel through the alarm system, notification, etc., so that timely response measures can be taken.
[0103] The technical effect of the above technical solution is: through real-time monitoring and early warning, abnormal phenomena in the production process can be discovered in time, reducing the probability of accidents. Through hierarchical monitoring of the production process, the operation of each production link can be more clearly understood, which is helpful to optimize production management and resource allocation. Once an abnormal phenomenon is detected, the system can quickly generate early warning information and notify relevant personnel, thereby improving the response speed to abnormal situations. Through automated and intelligent monitoring methods, the frequency and cost of manual inspections are reduced, and the operation and maintenance efficiency is improved. The monitoring data and analysis results collected by the system can provide strong support for the company's decision-making and help the company formulate more scientific and reasonable production plans and management strategies.
[0104] In one embodiment of the present invention, the multi-level monitoring module includes:
[0105] The hierarchical classification module is used to obtain production link information, classify monitoring type information according to the production link information, and obtain a multi-level operation monitoring system; the monitoring types include monitoring of various operations in the workshop, monitoring of different rooms of production equipment, etc.
[0106] The upper and lower level determination module is used to obtain the timing information of the production links, and determine the number of levels of each monitoring type according to the timing information of the production links; the timing information of the production links is different production links at different times. For example, samples are produced at 5 o'clock, and the samples produced at 5 o'clock are tested at 6 o'clock. Then the samples produced at 5 o'clock and the tests at 6 o'clock are the upper and lower hierarchical relationship of the monitoring types.
[0107] Establishing a hierarchical relationship between each monitoring category according to the number of levels of each monitoring category to obtain a monitoring hierarchy chain;
[0108] The data acquisition module is used to obtain monitoring data of each monitoring level of the monitoring level chain, pre-process the monitoring data, and obtain pre-processed level monitoring data.
[0109] The working principle of the above technical solution is as follows: collect specific information about the production process, including production process, operation tasks, equipment layout, etc. According to the production process information, classify the types that need to be monitored, such as monitoring of different operations in the workshop (such as assembly operations, packaging operations, etc.), monitoring of production equipment in different rooms (such as injection molding machines, stamping machines, etc.), etc. Obtain the timing information of the production process, that is, the activity sequence of different production processes at different time points. According to these timing information, determine the number of levels of each monitoring type. For example, in a process where a production sample is followed by a test, the production sample and the test can be regarded as two different levels of monitoring types. Based on the production process timing information, clarify the upper and lower hierarchical relationships between different monitoring types. These relationships reflect the dependencies and sequences between different links in the production process. Connect these relationships in series to form a monitoring hierarchy chain, each level corresponding to one or more monitoring types. For each monitoring level in the monitoring hierarchy chain, collect the corresponding monitoring data. These data may come from sensors, log records, operator input, etc. Preprocess the monitoring data, including data cleaning (removing noise, error values, etc.), data standardization (unifying measurement units, formats, etc.), data conversion (converting raw data into a format that can be used for analysis), etc. The preprocessed hierarchical monitoring data will be used for subsequent analysis and early warning.
[0110] The technical effect of the above technical solution is: through the multi-level operation monitoring system, the enterprise can clearly understand the status and changes of each link in the production process, so as to better grasp the production progress and efficiency. Based on the analysis and early warning of monitoring data, the enterprise can timely discover potential risks in the production process, such as equipment failure, abnormal operation, etc., and take corresponding measures to intervene and reduce losses. By monitoring the use of different production links and equipment, the enterprise can more reasonably allocate resources, such as manpower, material resources, financial resources, etc., so as to improve resource utilization efficiency. The data and analysis results provided by the multi-level operation monitoring system can provide strong support for the decision-making of the enterprise, and help the enterprise to formulate more scientific and reasonable production plans, management strategies, etc. Through real-time monitoring and early warning of the production links, the enterprise can timely discover and deal with possible safety hazards, so as to ensure the safety and stability of the production process.
[0111] In one embodiment of the present invention, the abnormal range acquisition module includes:
[0112] The target data matching module is used to obtain multi-level recognition targets, match the multi-level recognition targets with the multi-level operation monitoring system at each level, and obtain a hierarchical matching relationship; the multi-level recognition targets include objects that need to be monitored at each level, such as laboratory operating procedures and production samples or production equipment in a production workshop, etc., specifically things that need to be monitored.
[0113] Obtaining hierarchical monitoring data of each hierarchical identification target through the hierarchical matching relationship;
[0114] An abnormality index calculation module, used to calculate the target abnormality index of each level identification target according to the level monitoring data combined with the multi-level identification target information;
[0115] The calculation formula of the target abnormality index is:
[0116]
[0117] in, The target anomaly index for the hierarchical identification target, The maximum value of the hierarchical monitoring data for the hierarchical identification target, The minimum value of the hierarchical monitoring data for the hierarchical identification target, Preset monitoring data for tier identification targets.
[0118] The abnormality determination module is used to determine the abnormal points of each level identification target according to the target abnormality index to obtain multiple target abnormal points and target normal points; compare the target abnormality index with the preset index threshold, and when the target abnormality index is greater than the preset index threshold, determine the identification target as a target abnormal point, otherwise it is a target normal point.
[0119] The abnormal circle acquisition module is used to virtually connect multiple target abnormal points, obtain the maximum connection perimeter or side length, and obtain the data abnormal circle.
[0120] The working principle of the above technical solution is: determine multi-level recognition targets, which represent specific objects or matters that need to be monitored at each level, such as laboratory operation procedures, production samples or production equipment in the production workshop, etc. Match these multi-level recognition targets with the previously established multi-level operation monitoring system to ensure that each recognition target can correspond to the corresponding monitoring level to form a hierarchical matching relationship. Through the hierarchical matching relationship, the system can accurately obtain the real-time monitoring data of each level recognition target. According to the acquired hierarchical monitoring data, the target anomaly index of each level recognition target is calculated. This index reflects the degree of deviation of the current monitoring data from the normal state. The system compares the calculated target anomaly index with the preset index threshold. If the target anomaly index exceeds the threshold, the system determines that the recognition target is an abnormal point; otherwise, it is determined to be a normal point. In this way, the system can identify potential problems or abnormal situations in the production process.
[0121] For multiple identified target anomalies, the system uses virtual connection technology to connect them. This connection can be based on wireless network connections or logical connections based on production processes. The system calculates the maximum perimeter or side length of these connections to form a data anomaly circle. This anomaly circle represents the distribution and diffusion of anomalies in the production process, which helps companies quickly locate problem areas and assess potential risks.
[0122] The technical effect of the above technical solution is: by matching the multi-level identification target with the multi-level operation monitoring system, the system can more accurately obtain the real-time monitoring data of each level identification target, and calculate the target anomaly index based on these data, thereby improving the accuracy of anomaly identification. The system can calculate the target anomaly index in real time and make anomaly point judgments. Once an anomaly point is identified, it can immediately trigger the corresponding early warning mechanism to help enterprises achieve rapid response and processing. By constructing a data anomaly circle, enterprises can clearly see the distribution and diffusion of anomalies in the production process, so as to allocate resources in a targeted manner, optimize production processes, and improve production efficiency. The data anomaly circle not only shows the distribution of current anomalies, but also reflects the diffusion trend and scope of potential risks, providing enterprises with more comprehensive and accurate risk warning information. Through real-time monitoring and anomaly identification, the system can promptly discover and deal with safety hazards in the production process, reduce the probability of accidents, and improve production safety.
[0123] In one embodiment of the present invention, the classification warning module includes:
[0124] A hidden danger rate calculation module is used to obtain a target abnormal point in a data abnormal circle, and calculate the abnormal hidden danger rate of the data abnormal circle according to the target abnormal point in the data abnormal circle combined with the target normal point information;
[0125] The calculation formula of the abnormal hidden danger rate is:
[0126]
[0127] in, is the abnormal potential risk rate, c is the number of levels contained in the data abnormal circle, is the number of target anomaly points at the i-th level, is the number of target normal points at the i-th level;
[0128] The category information acquisition module is used to obtain the abnormal category information according to the abnormal hidden danger rate combined with the target abnormal point information; compare the abnormal hidden danger rate with the preset hidden danger threshold, and when the abnormal hidden danger rate is greater than the preset hidden danger threshold, obtain the monitoring category of each target abnormal point in the data abnormal circle, otherwise do not obtain the monitoring category. The monitoring category at this time is the abnormal category information.
[0129] The category warning module is used to trigger a corresponding category warning according to the abnormal category information and issue a warning according to the corresponding category warning.
[0130] The working principle of the above technical solution is as follows: the system identifies and determines all target abnormal points within the data abnormal circle. These abnormal points are determined by the previous analysis steps and represent potential problems or abnormal situations in the production process. The system calculates the abnormal hidden danger rate of the data abnormal circle based on the number and severity of the target abnormal points within the data abnormal circle and the information of the target normal points. This hidden danger rate is a quantitative indicator that reflects the potential risk of the abnormal situation and the potential impact on the production process. According to the abnormal hidden danger rate and the specific information of the target abnormal point, the system analyzes and determines the abnormal category information. These abnormal category information may include equipment failure, production process abnormality, operation error, etc., depending on the production environment and monitoring requirements. The system compares the abnormal hidden danger rate with the preset hidden danger threshold. If the abnormal hidden danger rate exceeds the threshold, the system triggers the corresponding warning level. The warning level is usually divided according to the severity of the hidden danger and the potential impact on the production process, such as low-level warning, intermediate warning, high-level warning, etc. Once the warning level is triggered, the system will perform warning operations according to the preset warning mechanism. This may include sending alarm information to relevant personnel, starting emergency response procedures, adjusting production plans, etc. The purpose of early warning is to notify relevant personnel in a timely manner and take appropriate measures to deal with abnormal situations in order to reduce potential losses.
[0131] The technical effect of the above technical solution is: by calculating the abnormal hidden danger rate and determining the abnormal category information, the system can more accurately assess the potential risks in the production process and trigger the corresponding warning level. This helps enterprises to discover and deal with abnormal situations in a timely manner and reduce potential losses. The system can allocate resources such as manpower, material resources, and financial resources in a targeted manner according to the warning level and abnormal category information. This helps enterprises to deal with abnormal situations more efficiently and improve production efficiency and safety. Through real-time monitoring and early warning mechanisms, the system can timely discover safety hazards in the production process and take corresponding measures to deal with them. This helps to reduce the probability of accidents and improve production safety. The data and analysis results provided by the early warning mechanism can provide strong support for the decision-making of enterprises. Enterprises can formulate more scientific and reasonable production plans, management strategies, etc. based on early warning information to improve production efficiency and competitiveness. The system can calculate the abnormal hidden danger rate and trigger the early warning level in real time to achieve rapid response. This helps enterprises to take timely measures to deal with abnormal situations and reduce potential losses and downtime.
[0132] In one embodiment of the present invention, the category warning module includes:
[0133] A category risk coefficient acquisition module is used to calculate the proportion of target abnormal points of each abnormal category information among the target abnormal points of all abnormal category information;
[0134] Multiply the percentage by the abnormal hidden danger rate to obtain the hidden danger risk coefficient corresponding to each abnormal category information;
[0135] The comparison and warning module is used to compare the hidden danger risk coefficient of each abnormal category information with the corresponding preset hidden danger risk threshold to obtain the abnormal comparison result; when the hidden danger risk coefficient is greater than the corresponding preset hidden danger risk threshold, a safety warning is issued for the corresponding abnormal category, otherwise no warning is issued.
[0136] A security warning is issued for the corresponding abnormal category according to the abnormal comparison result.
[0137] The working principle of the above technical solution is: the system counts the number of target abnormal points of each abnormal category information, and calculates the proportion of target abnormal points of each abnormal category information in the target abnormal points of all abnormal category information. This proportion reflects the relative importance or frequency of each abnormal category in the current production process. The system multiplies the proportion of each abnormal category information with the abnormal hidden danger rate calculated previously to obtain the hidden danger risk coefficient corresponding to each abnormal category information. The hidden danger risk coefficient is a comprehensive indicator that takes into account both the proportion of abnormal categories (i.e., their importance or frequency) and the abnormal hidden danger rate of the entire production process. The system presets the hidden danger risk threshold corresponding to each abnormal category. Then, the system compares the hidden danger risk coefficient of each abnormal category information with the corresponding preset hidden danger risk threshold. If the hidden danger risk coefficient exceeds the threshold, the system triggers a safety warning for the corresponding abnormal category; otherwise, no warning is issued.
[0138] Once a safety warning is triggered, the system will perform corresponding operations according to the preset warning mechanism and level. This may include sending alarm information to relevant personnel, initiating emergency response procedures, adjusting production plans, etc. The purpose of the warning is to notify relevant personnel in a timely manner and take corresponding safety measures to deal with or avoid potential risks.
[0139] The technical effect of the above technical solution is: by calculating the hidden danger coefficient of each abnormal category information and comparing it with the preset hidden danger threshold, the system can achieve accurate early warning and targeted processing of different abnormal categories. This helps to improve the accuracy and effectiveness of the early warning and reduce unnecessary false alarms and missed reports. Since the system can identify the importance and degree of danger of each abnormal category, enterprises can allocate resources such as manpower, material resources, and financial resources more targetedly according to the early warning information. This helps enterprises to deal with abnormal situations more efficiently and improve production efficiency and safety. Through real-time monitoring and early warning mechanisms, the system can timely discover potential risks in the production process and trigger corresponding safety warnings. This helps enterprises to take timely measures to deal with or avoid potential risks, reduce the probability of accidents, and improve production safety. The data and analysis results provided by the early warning mechanism can provide strong support for the decision-making of enterprises. Enterprises can formulate more scientific and reasonable production plans, management strategies, etc. based on the early warning information to improve production efficiency and competitiveness. The system achieves rapid response and early warning by automatically calculating and comparing the hidden danger coefficient with the preset threshold. This helps enterprises to take timely measures to deal with abnormal situations and reduce potential losses and downtime.
[0140] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to include these modifications and variations.
Claims
1. A method for controlling safe operation based on artificial intelligence identification, characterized in that: The method comprises: S1. Establish a multi-level operation monitoring system based on production link information, establish the upper and lower hierarchical relationship of monitoring types based on production environment time series information, and obtain hierarchical monitoring data; S2, obtaining hierarchical monitoring data of multi-level identification targets, calculating the target anomaly index of each hierarchical identification target, determining target abnormal points and target normal points according to the target anomaly index, virtually connecting the target abnormal points, and obtaining data anomaly circles; Wherein, the S2 includes: Acquire a multi-level recognition target, match the multi-level recognition target with a multi-level operation monitoring system at each level, and obtain a level matching relationship; Obtaining hierarchical monitoring data of each hierarchical identification target through the hierarchical matching relationship; Calculate the target anomaly index of each level identification target according to the level monitoring data combined with the multi-level identification target information; According to the target abnormality index, abnormal point determination is performed on each level identification target to obtain multiple target abnormal points and target normal points; Virtually connect multiple target anomaly points, obtain the maximum connection perimeter or side length, and obtain the data anomaly circle; S3, calculating the abnormal potential risk rate of the data abnormal circle, obtaining abnormal category information according to the abnormal potential risk rate, calculating the potential risk coefficient of each abnormal category information, and performing a safety warning for each abnormal category according to the potential risk coefficient; Wherein, the S3 includes: Obtaining target abnormal points within the data abnormal circle, and calculating the abnormal potential risk rate of the data abnormal circle according to the target abnormal points within the data abnormal circle combined with target normal point information; The calculation formula of the abnormal hidden danger rate is: in, is the abnormal potential risk rate, c is the number of levels contained in the data abnormal circle, is the number of target anomaly points at the i-th level, is the number of target normal points at the i-th level; Obtaining abnormality category information according to the abnormal potential risk rate combined with target abnormal point information; A corresponding category warning is triggered according to the abnormal category information, and a warning is issued according to the corresponding category warning.
2. According to claim 1, a method for controlling safe operation based on artificial intelligence identification is characterized in that: The S1 includes: Acquire production link information, classify monitoring category information according to the production link information, and obtain a multi-level operation monitoring system; Obtaining production link timing information, and determining the number of levels for each monitoring type based on the production link timing information; Establishing a hierarchical relationship of each monitoring type according to the number of levels of each monitoring type to obtain a monitoring hierarchy chain; The monitoring data of each monitoring level of the monitoring level chain is acquired, and the monitoring data is preprocessed to obtain the preprocessed level monitoring data.
3. According to claim 1, a method for controlling safe operation based on artificial intelligence identification is characterized in that: The triggering of a corresponding category warning according to the abnormal category information, and issuing a warning according to the corresponding category warning, includes: Calculate the proportion of target anomaly points of each type of anomaly information among all target anomaly points of all anomaly category information; Multiply the percentage by the abnormal hidden danger rate to obtain the hidden danger risk coefficient corresponding to each abnormal category information; Compare the hidden danger risk coefficient of each abnormal category information with the corresponding preset hidden danger risk threshold to obtain an abnormal comparison result; A security warning is issued for the corresponding abnormal category according to the abnormal comparison result.
4. A safety operation control system based on artificial intelligence identification, characterized in that: The system comprises: The multi-level monitoring module is used to establish a multi-level operation monitoring system based on the production link information, establish the upper and lower hierarchical relationship of monitoring types based on the production environment time series information, and obtain hierarchical monitoring data; The abnormal range acquisition module is used to obtain the hierarchical monitoring data of the multi-level identification target, calculate the target abnormality index of each hierarchical identification target, determine the target abnormal point and the target normal point according to the target abnormality index, virtually connect the target abnormal point, and obtain the data abnormal circle; Wherein, the abnormal range acquisition module includes: A target data matching module is used to obtain a multi-level recognition target, match the multi-level recognition target with a multi-level operation monitoring system at each level, and obtain a level matching relationship; Obtaining hierarchical monitoring data of each hierarchical identification target through the hierarchical matching relationship; An abnormality index calculation module, used to calculate the target abnormality index of each level identification target according to the level monitoring data combined with the multi-level identification target information; An abnormality determination module, used for determining abnormal points of each level identification target according to the target abnormality index, and obtaining a plurality of target abnormal points and target normal points; Anomaly circle acquisition module, used to virtually connect multiple target anomaly points, obtain the maximum connection perimeter or side length, and obtain the data anomaly circle; A classification warning module is used to calculate the abnormal potential risk rate of the data abnormal circle, obtain abnormal category information according to the abnormal potential risk rate, calculate the potential risk coefficient of each abnormal category information, and issue a safety warning for each abnormal category according to the potential risk coefficient; Wherein, the classification warning module includes: A hidden danger rate calculation module is used to obtain a target abnormal point in a data abnormal circle, and calculate the abnormal hidden danger rate of the data abnormal circle according to the target abnormal point in the data abnormal circle combined with the target normal point information; The calculation formula of the abnormal hidden danger rate is: in, is the abnormal potential risk rate, c is the number of levels contained in the data abnormal circle, is the number of target anomaly points at the i-th level, is the number of target normal points at the i-th level; A category information acquisition module, used to obtain abnormal category information according to the abnormal hidden danger rate combined with target abnormal point information; The category warning module is used to trigger a corresponding category warning according to the abnormal category information and issue a warning according to the corresponding category warning.
5. According to claim 4, a safe operation control system based on artificial intelligence identification is characterized in that: The multi-level monitoring module includes: A hierarchical classification module is used to obtain production link information, classify monitoring type information according to the production link information, and obtain a multi-level operation monitoring system; The upper and lower level determination module is used to obtain the production link timing information and determine the number of levels of each monitoring type according to the production link timing information; Establishing a hierarchical relationship of each monitoring type according to the number of levels of each monitoring type to obtain a monitoring hierarchy chain; The data acquisition module is used to obtain monitoring data of each monitoring level of the monitoring level chain, pre-process the monitoring data, and obtain pre-processed level monitoring data.
6. According to claim 4, a safe operation control system based on artificial intelligence identification is characterized in that: The category warning module includes: A category risk coefficient acquisition module is used to calculate the proportion of target abnormal points of each abnormal category information among the target abnormal points of all abnormal category information; Multiply the percentage by the abnormal hidden danger rate to obtain the hidden danger risk coefficient corresponding to each abnormal category information; A comparison warning module is used to compare the hidden danger risk coefficient of each abnormal category information with the corresponding preset hidden danger risk threshold to obtain an abnormal comparison result; A security warning is issued for the corresponding abnormal category according to the abnormal comparison result.
Citation Information
Patent Citations
Hardware fault monitoring method and system based on IT operation and maintenance
CN115225463A
Fault detection method and device, electronic equipment and readable storage medium
CN117560264A
Autonomous real-time fault isolation method based on event log
CN117640350A
IT operation method and system based on artificial intelligence
CN117692345A
Non-supervision abnormal behavior detection method and system based on complex network
CN118449718A
