Plant safety operation risk data processing method based on artificial intelligence

Abstract

The present invention relates to the field of data processing technology, and specifically to a method for processing factory safety operation risk data based on artificial intelligence, including: collecting safety operation risk data of several branches, subsidiaries or production workshops to obtain a safety production warning index model and a safety risk warning model; obtaining data points corresponding to each branch, subsidiary or production workshop; obtaining abnormal data points in the overall data points; analyzing the branches, subsidiaries or production workshops corresponding to the abnormal data points, obtaining comparative risk values ​​for each branch, subsidiary or production workshop, and prioritizing each branch, subsidiary or production workshop, and formulating corresponding risk mitigation measures. The present invention optimizes the problem of data incompatibility when using the LOF algorithm for abnormality detection, and improves the safety performance and efficiency of the factory.

Description

Technical Field

[0001] The present invention relates to the technical field of data processing, and in particular to a method for processing factory safety operation risk data based on artificial intelligence. Background Art

[0002] With the development of science and technology, AI-based technologies can help enterprises more effectively manage and deal with factory safety risk issues. Refining and chemical companies have also begun to gradually apply AI-based factory safety operation risk data processing methods to improve factory safety and risk management efficiency. In factories, with the help of data collection technology, big data and cloud computing, automated control, and intelligent monitoring technologies, more accurate, real-time, and intelligent safety risk management solutions can be provided for enterprises and factories.

[0003] The LOF algorithm is often used for anomaly detection, but the LOF algorithm may have data mismatch problems when processing data. The algorithm itself needs to calculate the local density of sample points based on a set of feature data. If the input data does not match the assumptions and requirements of the LOF algorithm, the results may be inaccurate. Therefore, consider extracting the factory safety operation risk data to obtain a model of both safety and risk, which is consistent with the LOF algorithm. The algorithm is used to screen the high-risk subsidiaries and analyze and quantify the risk level in combination with the major hazards under them. Summary of the invention

[0004] The present invention provides a factory safety operation risk data processing method based on artificial intelligence to solve the existing problems.

[0005] The artificial intelligence-based factory safety operation risk data processing method of the present invention adopts the following technical solutions:

[0006] An embodiment of the present invention provides a method for processing plant safety operation risk data based on artificial intelligence, the method comprising the following steps:

[0007] Collect the safety operation risk data of several branches, subsidiaries or production workshops to obtain the safety production early warning index model and the safety risk early warning model; obtain the parameter data of each branch, subsidiary or production workshop through the safety production early warning index model and the safety risk early warning model;

[0008] Obtain the risk index of each branch, subsidiary or production workshop based on the parameter data of each branch, subsidiary or production workshop; obtain the production safety warning index of each branch, subsidiary or production workshop based on the production safety warning index model; obtain the data point corresponding to each branch, subsidiary or production workshop based on the risk index of each branch, subsidiary or production workshop and the production safety warning index of each branch, subsidiary or production workshop, and combine all the data points into an overall data point;

[0009] Obtain the distance value when using the LOF algorithm; obtain the local anomaly factor of each data point based on the distance value, and obtain all abnormal data points based on the local anomaly factor of each data point;

[0010] All major hazards included in the branches, subsidiaries or production workshops corresponding to all abnormal data points are recorded as overall major hazards, and the comparative risk level of each major hazard source is obtained based on the comparison of the overall major hazards; the comparative risk value of each branch, subsidiary or production workshop is obtained based on the comparative risk level of each major hazard source under each branch, subsidiary or production workshop; each branch, subsidiary or production workshop is prioritized according to its comparative risk value, and corresponding risk mitigation measures are formulated.

[0011] Preferably, the collecting of safety operation risk data of several branches, subsidiaries or production workshops to obtain a safety production early warning index model and a safety risk early warning model includes the following specific steps:

[0012] For any branch, subsidiary or production workshop, different types of data are collected, including Class A "inherent risk", Class B "production operation", Class C "field operation", Class D "double prevention", Class E "LDAR", Class F "equipment operation" and Class G "other factors"; and a production safety early warning index model is obtained from this; the input of the production safety early warning index model is: any branch, subsidiary or production workshop and the date of the day, and the output is: the production safety early warning index of any branch, subsidiary or production workshop on the date of the day;

[0013] For any major hazard source, different types of data are collected, including Class A "inherent risk", Class B "safety inspection", Class C "hidden danger control", Class D "on-site operation", Class E "process parameter control" and Class F "climate risk", a total of 6 different types of data, and a safety risk warning model is obtained from this; the input of the safety risk warning model is: any major hazard source and the date of the day, and the output is: the safety risk value of any major hazard source on the date of the day.

[0014] Preferably, the method of obtaining parameter data of each branch, subsidiary or production workshop through the production safety early warning index model and the safety risk early warning model includes the following specific steps:

[0015] The number of level 1 major hazard sources and their weight values, the number of level 2 major hazard sources and their weight values, the number of level 3 major hazard sources and their weight values, and the number of level 4 major hazard sources and their weight values ​​of each branch, subsidiary or production workshop are obtained through the safety production early warning index model; the safety risk value of each level 1 major hazard source on the date, the safety risk value of each level 2 major hazard source on the date, the safety risk value of each level 3 major hazard source on the date, and the safety risk value of each level 4 major hazard source on the date are obtained through the safety risk early warning model; the number of major hazard sources with a safety risk value interval level reaching a preset level under each branch, subsidiary or production workshop is obtained, and its ratio to all major hazard sources under each branch, subsidiary or production workshop is used as the ratio of the existence of major hazard sources with a safety risk value interval level reaching a preset level under each branch, subsidiary or production workshop;

[0016] The parameter data of each branch, subsidiary or production workshop include the number of level 1 major hazard sources and their weight values, the number of level 2 major hazard sources and their weight values, the number of level 3 major hazard sources and their weight values, and the number of level 4 major hazard sources and their weight values ​​of each branch, subsidiary or production workshop, the safety risk value of each level 1 major hazard source on the day, the safety risk value of each level 2 major hazard source on the day, the safety risk value of each level 3 major hazard source on the day, and the safety risk value of each level 4 major hazard source on the day, and the ratio of major hazard sources whose safety risk value interval level reaches the preset level under each branch, subsidiary or production workshop.

[0017] Preferably, the step of obtaining the risk index of each branch, subsidiary or production workshop according to the parameter data of each branch, subsidiary or production workshop comprises the following specific steps:

[0018] For any branch, subsidiary or production workshop, the risk index of the branch, subsidiary or production workshop is calculated by the safety risk values ​​of several major hazardous sources under the branch, subsidiary or production workshop. The calculation expression is:

[0019]

[0020] In the formula, R r Indicates the risk index of a branch, subsidiary or production workshop; Sum w Represents the total risk value of a branch, subsidiary or production workshop; N l1 Indicates the number of Level 1 major hazardous sources in a branch, subsidiary or production workshop; S l1 Indicates the safety risk value of the 11th level 1 major hazard source on the current date; N l2 Indicates the number of Level 2 major hazardous sources in a branch, subsidiary or production workshop; S l2 Indicates the safety risk value of the 12th level 2 major hazard source on the current date; N l3Indicates the number of Level 3 major hazardous sources in a branch, subsidiary or production workshop; S l3 Indicates the safety risk value of the 13th level 3 major hazard source on the current date; N l4 Indicates the number of Level 4 major hazardous sources in a branch, subsidiary or production workshop; S l4 It indicates the safety risk value of the 14th level 4 major hazard source on the current date; W l1 ,W l2 ,W l3 ,W l4 is the preset weight parameter.

[0021] Preferably, the step of obtaining the data points corresponding to each branch, subsidiary or production workshop according to the risk index of each branch, subsidiary or production workshop and the production safety early warning index of each branch, subsidiary or production workshop, and combining all the data points into an overall data point includes the following specific steps:

[0022] For each branch, subsidiary or production workshop, the corresponding safety production early warning index R can be obtained. s and risk index R r , obtain the corresponding rectangular coordinate system, and convert the safety production warning index R s As the horizontal axis, the risk index R r As the vertical coordinate, the data points corresponding to each branch, subsidiary or production workshop are constructed, and the coordinates of each data point are filled into the rectangular coordinate system to form the overall data point.

[0023] Preferably, the step of obtaining the distance value when using the LOF algorithm includes the following specific steps:

[0024] The k distance value of the overall data point when using the LOF algorithm is calculated as follows:

[0025]

[0026] In the formula, n represents the total number of branches, subsidiaries or production workshops; It represents the standard deviation of the production safety early warning index of all branches, subsidiaries or production workshops; It represents the standard deviation of the risk index of a branch, subsidiary or production workshop; β represents the preset parameter; exp() represents an exponential function with a natural constant as the base.

[0027] Preferably, the method of obtaining the local anomaly factor of each data point according to the distance value and obtaining all abnormal data points according to the local anomaly factor of each data point comprises the following specific steps:

[0028] After obtaining the k distance value, the LOF algorithm is used to calculate each data point to obtain a local anomaly factor corresponding to each data point; according to the properties of the local anomaly factor, the data point corresponding to the local anomaly factor greater than the preset parameter t in the local anomaly factor is recorded as an abnormal data point.

[0029] Preferably, the specific formula for obtaining the comparative risk level of each major hazard source based on the comparison of the overall major hazard sources is as follows:

[0030]

[0031] Where D c,r Indicates the comparative risk level of the cth level r major hazard source; S c,r Indicates the safety risk value of the cth level r major hazard source; It represents the weighted average of the safety risk values ​​of all major hazardous sources except the cth level r major hazardous source in the overall major hazardous sources; A represents the weighted standard deviation of the safety risk value of the overall major hazard source; r It represents the average safety risk value of all r-level major hazard sources except the c-th r-level major hazard source among the overall major hazard sources; σ l,r It represents the standard deviation of the safety risk values ​​of all r-level major hazardous sources in the overall major hazardous sources; β represents the preset parameter; exp() represents an exponential function with a natural constant as the base.

[0032] Preferably, the specific formula for obtaining the comparative risk value of each branch, subsidiary or production workshop according to the comparative risk level of each major hazard source under each branch, subsidiary or production workshop is as follows:

[0033]

[0034] In the formula, H n Represents the comparative risk value of the nth branch, subsidiary or production workshop; N n,r Indicates the number of r-level major hazardous sources under the n-th branch, subsidiary or production workshop; D c,r Indicates the comparative risk level of the cth level r major hazard source; Sum wn It represents the total risk value of the nth branch, subsidiary or production workshop; It indicates the ratio of the major hazard sources with orange and red color in the safety risk value interval of the nth branch, subsidiary or production workshop; R n It represents the production safety warning index of the nth branch, subsidiary or production workshop; exp() represents an exponential function with a natural constant as the base.

[0035] Preferably, the step of prioritizing each branch, subsidiary or production workshop according to the comparative risk value of each branch, subsidiary or production workshop and formulating corresponding risk mitigation measures includes the following specific steps:

[0036] The preset parameter τ is used. According to the comparative risk value obtained from the analysis and evaluation of each branch, subsidiary or production workshop, the branches, subsidiaries or production workshops are prioritized from large to small according to their comparative risk values, and the sorted sequence is recorded as the first sequence. For the branches, subsidiaries or production workshops that are ranked in the top τ% in the first sequence, corresponding risk mitigation measures are formulated, including optimizing existing safety measures, conducting relevant safety training for employees, and strengthening monitoring and inspection measures.

[0037] The beneficial effects of the technical solution of the present invention are: to address the problem of data mismatch when the LOF algorithm processes data, resulting in inaccurate results; the present invention collects the factory safety operation risk data and obtains a safety and risk correlation model, screens out branches, subsidiaries or production workshops with abnormalities in their own safety and major risk sources, analyzes and quantifies the risk level, and uses this as a reference for the priority ranking of branches, subsidiaries or production workshops. Subsequently, corresponding risk mitigation measures can be formulated based on the reference results; thereby optimizing the problem of data mismatch when using the LOF algorithm for anomaly detection, thereby improving the safety performance and efficiency of the factory. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0039] Figure 1 The present invention is a flowchart of the steps of the method for processing factory safety operation risk data based on artificial intelligence. DETAILED DESCRIPTION

[0040] In order to further explain the technical means and effects adopted by the present invention to achieve the predetermined invention purpose, the following is a detailed description of the specific implementation method, structure, features and effects of the factory safety operation risk data processing method based on artificial intelligence proposed by the present invention in combination with the accompanying drawings and preferred embodiments. In the following description, different "one embodiment" or "another embodiment" does not necessarily refer to the same embodiment. In addition, specific features, structures or characteristics in one or more embodiments may be combined in any suitable form.

[0041] Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.

[0042] The specific scheme of the factory safety operation risk data processing method based on artificial intelligence provided by the present invention is described in detail below with reference to the accompanying drawings.

[0043] See also Figure 1 , which shows a flowchart of a method for processing factory safety operation risk data based on artificial intelligence provided by an embodiment of the present invention, the method comprising the following steps:

[0044] Step S001: Collect safety operation risk data of several branches, subsidiaries or production workshops to obtain a safety production early warning index model and a safety risk early warning model.

[0045] Specifically, for any branch, subsidiary or production workshop, a total of 7 different types of data are collected, including Class A "inherent risk", Class B "production operation", Class C "on-site operation", Class D "double prevention", Class E "LDAR", Class F "equipment operation" and Class G "other factors", and the production safety early warning index model is obtained from this; the input of the production safety early warning index model is: any branch, subsidiary or production workshop and the date of the day, and the output is: the production safety early warning index of any branch, subsidiary or production workshop on the date of the day.

[0046] For any major hazard source, six different types of data are collected, including Class A "inherent risk", Class B "safety inspection", Class C "hidden danger control", Class D "on-site operation", Class E "process parameter control" and Class F "climate risk", and a safety risk warning model is obtained from it; the input of the safety risk warning model is: any major hazard source and the date of the day, and the output is: the safety risk value of any major hazard source on the date of the day.

[0047] It should be noted that each branch, subsidiary or production workshop has level 1 major hazard source, level 2 major hazard source, level 3 major hazard source and level 4 major hazard source; in the safety production early warning index model, the size of the safety production early warning index and the corresponding classification are as follows:

[0048] 100-95 is blue / low risk, 95-87.5 is yellow / general risk, 87.5-80 is orange / high risk, and 80 and below is red / high risk.

[0049] In the security risk warning model, the security risk value and corresponding classification are as follows:

[0050] 130 and below is blue / low risk, 130-160 is yellow / general risk, 160-190 is orange / higher risk, and 190 and above is red / high risk.

[0051] Specifically, the number of level 1 major hazard sources and their weight values, the number of level 2 major hazard sources and their weight values, the number of level 3 major hazard sources and their weight values, and the number of level 4 major hazard sources and their weight values ​​of each branch, subsidiary or production workshop are obtained through the production safety early warning index model; the safety risk value of each level 1 major hazard source on the same date, the safety risk value of each level 2 major hazard source on the same date, the safety risk value of each level 3 major hazard source on the same date, and the safety risk value of each level 4 major hazard source on the same date are obtained through the safety risk early warning model; the number of major hazard sources whose safety risk value interval level colors reach orange and red under each branch, subsidiary or production workshop is obtained, and the ratio of the number of major hazard sources to all major hazard sources under each branch, subsidiary or production workshop is taken as the ratio of the existence of major hazard sources whose safety risk value interval level colors reach orange and red under each branch, subsidiary or production workshop.

[0052] At this point, the production safety early warning index model and safety risk early warning model are obtained.

[0053] Step S002: Obtain the data points corresponding to each branch, subsidiary or production workshop.

[0054] It should be noted that for any branch, subsidiary or production workshop, the production safety warning index corresponding to the branch, subsidiary or production workshop on that date is obtained through the production safety warning index model. The larger the index is, the safer the branch, subsidiary or production workshop on that date is.

[0055] For any major hazard source of the branch, subsidiary or production workshop, the safety risk value corresponding to the major hazard source of the branch, subsidiary or production workshop on the current date is obtained through the safety risk warning model; the larger the risk value, the greater the risk of the major hazard source of the branch, subsidiary or production workshop on the current date.

[0056] It should be further explained that the above two-dimensional model can obtain the safety production warning index of each branch, subsidiary or production workshop and the safety risk value of several major hazardous sources under each branch, subsidiary or production workshop on the same day. To judge the safe operation of each branch, subsidiary or production workshop, on the one hand, it is reflected by its own safety level, and on the other hand, it is reflected by the risk level of several major hazardous sources under each branch, subsidiary or production workshop; because each branch, subsidiary or production workshop has several major hazardous sources (there may be only one or more), it is necessary to extract indicators that can represent several major risk sources through the safety risk values ​​obtained by the major hazardous sources under the safety risk warning model.

[0057] According to the safety production early warning index model, four weight parameters W are preset. l1 ,W l2 ,W l3 ,W l4 , wherein this embodiment uses W l1 =0.4,W l2 =0.3,W l3 =0.2,W l4 =0.1 is described in detail in this example, and this embodiment is not specifically limited, wherein W l1 >W l2 >W l3 >W l4 And it depends on the specific implementation situation.

[0058] Specifically, for any branch, subsidiary or production workshop, the risk index of the branch, subsidiary or production workshop is calculated based on the safety risk values ​​of several major hazards under the branch, subsidiary or production workshop. The calculation expression is:

[0059]

[0060] In the formula, R r Indicates the risk index of a branch, subsidiary or production workshop; Sum w Represents the total risk value of a branch, subsidiary or production workshop; N l1 Indicates the number of Level 1 major hazardous sources in a branch, subsidiary or production workshop; S l1 Indicates the safety risk value of the 11th level 1 major hazard source on the current date; N l2 Indicates the number of Level 2 major hazardous sources in a branch, subsidiary or production workshop; S l2 Indicates the safety risk value of the 12th level 2 major hazard source on the current date; N l3 Indicates the number of Level 3 major hazardous sources in a branch, subsidiary or production workshop; S l3 Indicates the safety risk value of the 13th level 3 major hazard source on the current date; N l4Indicates the number of Level 4 major hazardous sources in a branch, subsidiary or production workshop; S l4 It indicates the safety risk value of the 14th level 4 major hazard source on the current date; W l1 ,W l2 ,W l3 ,W l4 is the preset weight parameter.

[0061] Therefore, for each branch, subsidiary or production workshop, the corresponding safety production warning index R can be obtained. s and risk index R r , obtain the corresponding rectangular coordinate system, and convert the safety production warning index R s As the horizontal axis, the risk index R r As the vertical coordinate, the data points corresponding to each branch, subsidiary or production workshop are constructed, and the coordinates of each data point are filled into the rectangular coordinate system to form the overall data point.

[0062] At this point, the data points corresponding to each branch, subsidiary or production workshop are obtained.

[0063] Step S003: Obtain abnormal data points in the overall data points.

[0064] It should be noted that, since the overall data points are relatively close to each other when distributed, it is impossible to directly screen outliers. The LOF algorithm is used to calculate the local anomaly factor of each data point, and the data points whose local anomaly factor does not meet the expectations are screened as outliers. The LOF algorithm is an algorithm for outlier detection based on local density distribution. When calculating the LOF algorithm, one of the important parameters is the k distance value. The size of this value is related to the distribution of the overall data points and the scale of the overall data points.

[0065] A parameter β is preset, wherein this embodiment is described as β=0.1, and this embodiment is not specifically limited, wherein β is determined according to specific implementation conditions.

[0066] Specifically, the k distance value of the overall data point when using the LOF algorithm is calculated as follows:

[0067]

[0068] In the formula, n represents the total number of branches, subsidiaries or production workshops, and in the coordinate axis it is the total number of all data points; It represents the standard deviation of the production safety early warning index of all branches, subsidiaries or production workshops; It represents the standard deviation of the risk index of a branch, subsidiary or production workshop; β represents the preset parameter; exp() represents an exponential function with a natural constant as the base.

[0069] It should be noted that the choice of k distance value is related to the distribution of the overall data points and the size of the overall data points. The specific relationship is that the greater the overall density of the data points, that is, the smaller the standard deviation of the horizontal and vertical coordinates, the corresponding The larger the value of , the larger the k distance value is. In addition, the larger the number of overall data points, the larger the k distance value is.

[0070] Specifically, after obtaining the k-distance value, the LOF algorithm is used to calculate each data point, and a local anomaly factor corresponding to each data point is calculated. According to the properties of the local anomaly factor, the data point corresponding to the local anomaly factor greater than 1 in the local anomaly factor is recorded as an abnormal data point.

[0071] It should be further explained that when the local anomaly factor is greater than 1, it means that the density of the data point is lower than that of its neighboring data points, that is, the data point has a relatively isolated feature. This may indicate that the data point is significantly different from other data points, or that it is in an abnormal subgroup. Therefore, data points with a local anomaly factor greater than 1 are usually considered abnormal data points.

[0072] At this point, the abnormal data points among the overall data points are obtained.

[0073] Step S004: Analyze the branches, subsidiaries or production workshops corresponding to the abnormal data points, obtain the comparative risk value of each branch, subsidiary or production workshop, prioritize each branch, subsidiary or production workshop, and formulate corresponding risk mitigation measures.

[0074] 1. Obtain the comparative risk level of each major hazard source.

[0075] It should be noted that since each abnormal data point essentially corresponds to a branch, subsidiary or production workshop, that is, the branch, subsidiary or production workshop has a significant degree of outlier in terms of production safety warning index and risk index compared with the other data points, it is necessary to analyze these abnormal branches, subsidiaries or production workshops next; in terms of the risks embodied by major hazardous sources, one is to compare the major hazardous source with other major hazardous sources in the branch, subsidiary or production workshop to which it belongs, and the other is to compare it with other major hazardous sources under other branches, subsidiaries or production workshops.

[0076] Specifically, all major hazard sources included in the branches, subsidiaries or production workshops corresponding to all abnormal data points are recorded as the overall major hazard sources. The comparative risk level of each major hazard source is obtained based on the comparison of the overall major hazard sources. The calculation formula is:

[0077]

[0078] Where Dc,r Indicates the comparative risk level of the cth level r major hazard source; S c,r Indicates the safety risk value of the cth level r major hazard source; It represents the weighted average of the safety risk values ​​of all major hazardous sources except the cth level r major hazardous source in the overall major hazardous sources; A represents the weighted standard deviation of the safety risk value of the overall major hazard source; r It represents the average safety risk value of all r-level major hazard sources except the c-th r-level major hazard source among the overall major hazard sources; σ l,r It represents the standard deviation of the safety risk values ​​of all r-level major hazardous sources in the overall major hazardous sources; β represents the preset parameter; exp() represents an exponential function with a natural constant as the base.

[0079] It should be further explained that for all major hazardous sources included in the branch, subsidiary or production workshop corresponding to each abnormal data point, each hazardous source has a corresponding safety risk value. If the safety risk values ​​of these major hazardous sources are smaller, the corresponding A or The smaller it is, the more stable the risk value distribution is, corresponding to σ wr The smaller it is, the more prominent one of the major hazard sources with a relatively large safety risk value will appear, that is, the risk level of this major hazard source is higher than that of other major hazard sources in the branch, subsidiary or production workshop to which it belongs. According to this idea, the comparison object is expanded to all major hazard sources, but the comparison is with the safety risk values ​​of other major hazard sources of the same risk level in the whole, which also includes other major hazard sources of the same risk level in the branch, subsidiary or production workshop to which the major hazard source belongs except for this major hazard source.

[0080] At this point, the comparative risk level of each major hazard source is obtained.

[0081] 2. Obtain comparative risk values ​​for each branch, subsidiary or production workshop.

[0082] Specifically, according to the comparative risk level of each major hazard source under the nth branch, subsidiary or production workshop, the comparative risk value of the nth branch, subsidiary or production workshop is obtained, and its calculation expression is:

[0083]

[0084] In the formula, H n Represents the comparative risk value of the nth branch, subsidiary or production workshop; N n,r Indicates the number of r-level major hazardous sources under the n-th branch, subsidiary or production workshop; D c,r Indicates the comparative risk level of the cth level r major hazard source; Sum wIt represents the total risk value of the nth branch, subsidiary or production workshop; It indicates the ratio of the major hazard sources with orange and red color in the safety risk value interval of the nth branch, subsidiary or production workshop; R n It represents the production safety warning index of the nth branch, subsidiary or production workshop; exp() represents an exponential function with a natural constant as the base.

[0085] At this point, the comparative risk value of each branch, subsidiary or production workshop is obtained.

[0086] It should be noted that, based on the comparative risk values ​​obtained from the analysis and evaluation of each branch, subsidiary or production workshop, the branches, subsidiaries or production workshops are prioritized according to their comparative risk values ​​to determine which branches, subsidiaries or production workshops face the highest risks; resources are reasonably allocated to deal with the safety risks of each branch, subsidiary or production workshop to meet the needs of high risks. In addition, if only major hazardous sources are targeted, the priority can be ranked according to the comparative risk level of major hazardous sources.

[0087] A parameter τ is preset, wherein this embodiment is described as an example of τ=20, and this embodiment is not specifically limited, wherein τ is determined according to specific implementation conditions.

[0088] Specifically, based on the comparative risk values ​​obtained from the analysis and evaluation of each branch, subsidiary or production workshop, the branches, subsidiaries or production workshops are prioritized from large to small according to their comparative risk values, and the sorted sequence is recorded as the first sequence; for the branches, subsidiaries or production workshops that are ranked in the top τ% in the first sequence, corresponding risk mitigation measures are formulated, including optimizing existing safety measures, conducting relevant safety training for employees, and strengthening monitoring and inspection measures.

[0089] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A method for processing factory safety operation risk data based on artificial intelligence, characterized in that: The method comprises the following steps: Collect the safety operation risk data of several branches, subsidiaries or production workshops to obtain the safety production early warning index model and the safety risk early warning model; obtain the parameter data of each branch, subsidiary or production workshop through the safety production early warning index model and the safety risk early warning model; Obtain the risk index of each branch, subsidiary or production workshop based on the parameter data of each branch, subsidiary or production workshop; obtain the production safety warning index of each branch, subsidiary or production workshop based on the production safety warning index model; obtain the data point corresponding to each branch, subsidiary or production workshop based on the risk index of each branch, subsidiary or production workshop and the production safety warning index of each branch, subsidiary or production workshop, and combine all the data points into an overall data point; Get Application The distance value during the algorithm; obtain the local anomaly factor of each data point according to the distance value, and obtain all abnormal data points according to the local anomaly factor of each data point; All major hazard sources included in the branches, subsidiaries or production workshops corresponding to all abnormal data points are recorded as overall major hazard sources, and the comparative risk level of each major hazard source is obtained based on the comparison of the overall major hazard sources; the comparative risk value of each branch, subsidiary or production workshop is obtained based on the comparative risk level of each major hazard source under each branch, subsidiary or production workshop; each branch, subsidiary or production workshop is prioritized according to its comparative risk value, and corresponding risk mitigation measures are formulated; The acquisition application The distance value of the algorithm includes the following specific steps: Overall data points in use When the algorithm The distance value is calculated as follows: In the formula, Indicates the total number of branches, subsidiaries or production workshops; It represents the standard deviation of the production safety early warning index of all branches, subsidiaries or production workshops; It represents the standard deviation of the risk index of a branch, subsidiary or production workshop; Indicates preset parameters; It represents the exponential function with the natural constant e as the base; The specific formula for obtaining the comparative risk value of each branch, subsidiary or production workshop according to the comparative risk level of each major hazard source under each branch, subsidiary or production workshop is as follows: In the formula, Indicates Comparative risk values ​​of individual branches, subsidiaries or production plants; Indicates A branch, subsidiary or production workshop The number of major hazard sources of level 1; Indicates indivual The comparative risk levels of major hazards of level 1; Indicates The total risk value of each branch, subsidiary or production workshop; Indicates The ratio of major hazards with orange and red levels in the safety risk value interval of each branch, subsidiary or production workshop; Indicates Safety production early warning index for individual branches, subsidiaries or production workshops; No. indivual The specific calculation formula for the comparative risk level of the major hazard source of level 1 is: In the formula, Indicates indivual The comparative risk levels of major hazards of level 1; Indicates indivual Safety risk value of major hazardous sources of level 1; Indicates that among the major hazards, except indivual The weighted average of the safety risk values ​​of the major hazardous sources of level 1 and other major hazardous sources; Represents the weighted standard deviation of the safety risk value of the overall major hazard source; Indicates that among the major hazards, except indivual Other major hazard sources The average safety risk value of major hazardous sources of level 1; Indicates all major hazards in the overall Standard deviation of safety risk values ​​for major hazardous sources of level 1; Indicates preset parameters.

2. The method for processing factory safety operation risk data based on artificial intelligence according to claim 1 is characterized in that: The specific steps of collecting the safety operation risk data of several branches, subsidiaries or production workshops and obtaining the safety production early warning index model and the safety risk early warning model are as follows: For any branch, subsidiary or production workshop, different types of data are collected, including Type "inherent risk", Class "Production Operation", Class "field work", Type "double prevention", Class "LDAR", Class "Device Operation" and Category "Other factors"; And thereby obtain a production safety early warning index model; the input of the production safety early warning index model is: any branch company and different types of data collected on the same date or any production workshop and different types of data collected on the same date, and the output is: the production safety early warning index of any branch company on the same date or the production safety early warning index of any production workshop on the same date; For any major hazard source, different types of data are collected, including Type "inherent risk", Class "Safety Check", Category "Hazard Management" Class "field work", Category "Process Parameter Control" and Category "climate risk" Different types of data are collected from different sources of hazard and a safety risk warning model is obtained from them; the input of the safety risk warning model is: different types of data collected from any major hazard source and its current date, and the output is: the safety risk value of any major hazard source on that date.

3. The method for processing factory safety operation risk data based on artificial intelligence according to claim 1 is characterized in that: The specific steps of obtaining parameter data of each branch, subsidiary or production workshop through the production safety early warning index model and the safety risk early warning model are as follows: Obtain parameter data of each branch, subsidiary or production workshop through the safety production early warning index model; The parameter data includes The number of major hazard sources and their weights, The number of major hazard sources and their weights, The number of major hazard sources and their weights and The number of major hazard sources and their weights, and the number of The safety risk value of the major hazardous source of level 1, the date of each The safety risk value of the major hazardous source of level 1, the date of each The safety risk value of the major hazardous source of level 1, the date of each The ratio of the safety risk value of the major hazardous sources of the level 1 to the safety risk value interval level of each branch, subsidiary or production workshop that reaches the preset level of major hazardous sources.

4. The method for processing factory safety operation risk data based on artificial intelligence according to claim 3 is characterized in that: The specific steps of obtaining the risk index of each branch, subsidiary or production workshop according to the parameter data of each branch, subsidiary or production workshop are as follows: For any branch, subsidiary or production workshop, the risk index of the branch, subsidiary or production workshop is calculated by the safety risk values ​​of several major hazardous sources under the branch, subsidiary or production workshop. The calculation expression is: In the formula, Indicates the risk index of a branch, subsidiary or production workshop; Indicates the total risk value of a branch, subsidiary or production workshop; Indicates the branch, subsidiary or production workshop Number of major hazard sources of level 1; Indicates the date of the day indivual Safety risk value of major hazardous sources of level 1; Indicates the branch, subsidiary or production workshop Number of major hazard sources of level 1; Indicates the date of the day indivual Safety risk value of major hazardous sources of level 1; Indicates the branch, subsidiary or production workshop Number of major hazard sources of level 1; Indicates the date of the day indivual Safety risk value of major hazardous sources of level 1; Indicates the branch, subsidiary or production workshop Number of major hazard sources of level 1; Indicates the date of the day indivual Safety risk value of major hazardous sources of level 1; is the preset weight parameter.

5. The method for processing factory safety operation risk data based on artificial intelligence according to claim 1 is characterized in that: The specific steps of obtaining the data points corresponding to each branch, subsidiary or production workshop according to the risk index of each branch, subsidiary or production workshop and the safety production warning index of each branch, subsidiary or production workshop, and combining all the data points into an overall data point are as follows: For each branch, subsidiary or production workshop, a corresponding safety production early warning index can be obtained and risk index , obtain the corresponding rectangular coordinate system, and convert the safety production warning index As the horizontal axis, the risk index As the vertical coordinate, the data points corresponding to each branch, subsidiary or production workshop are constructed, and the coordinates of each data point are filled into the rectangular coordinate system to form the overall data point.

6. The method for processing factory safety operation risk data based on artificial intelligence according to claim 1 is characterized in that: The specific steps of obtaining the local anomaly factor of each data point according to the distance value and obtaining all abnormal data points according to the local anomaly factor of each data point are as follows: get After the distance value, use The algorithm calculates and obtains a local anomaly factor corresponding to each data point; according to the properties of the local anomaly factor, the local anomaly factor greater than the preset parameter The data points corresponding to the local abnormal factors are recorded as abnormal data points.

7. The method for processing factory safety operation risk data based on artificial intelligence according to claim 1 is characterized in that: The method of prioritizing each branch, subsidiary or production workshop according to the comparative risk value of each branch, subsidiary or production workshop and formulating corresponding risk mitigation measures includes the following specific steps: Preset Parameters According to the comparative risk values ​​obtained by the analysis and evaluation of each branch, subsidiary or production workshop, the branches, subsidiaries or production workshops are prioritized from large to small according to their comparative risk values, and the sequence obtained after sorting is recorded as the first sequence; For each branch, subsidiary or production workshop, formulate corresponding risk mitigation measures, including optimizing existing safety measures, conducting relevant safety training for employees, and strengthening monitoring and inspection measures.

Images