Risk management and control and hidden danger investigation method based on multi-factor variance analysis

Abstract

The invention discloses a risk management and control and hidden danger investigation method based on multi-factor variance analysis, and the method comprises the steps: determining an overall target of risk management and control and hidden danger investigation, and recognizing and defining key variables affecting the risk management and control effect according to the overall target; obtaining production data; performing data preprocessing on the production data to obtain preprocessed data; constructing a multi-factor variance analysis model according to the overall target and the key variables; inputting the preprocessed data into a multi-factor variance analysis model, and determining key factors; and determining a risk management and control strategy according to the key factors. And the accuracy and effectiveness of risk management and control and hidden danger investigation are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of risk management and control, and in particular to a method for risk management and hidden danger detection based on multi-factor variance analysis. Background Art

[0002] At present, risk control and hidden danger investigation are important tasks for enterprises to ensure safe production. Based on years of industry experience, a variety of assessment methods have been formed, including risk matrix method, LEC method, hierarchical analysis method, and HAZOP analysis, which can qualitatively or quantitatively analyze the possibility and impact of accident risks, help enterprises better identify risks and eliminate hidden dangers. However, these methods have their own limitations to varying degrees. 1. Risk matrix method: The risk matrix method is an assessment method that quantifies the possibility and impact of risk events, but it has the following limitations: Subjectivity: The assessment of risk probability and impact often depends on personal subjective judgment and may be biased. Simplify the problem: The matrix diagram simplifies risk factors into three levels: high, medium, and low, which may not fully reflect complex risk situations. Dynamic changes: Risk factors and market environment are dynamically changing, and the matrix diagram needs to be updated regularly to maintain its effectiveness. 2. LEC method: The LEC evaluation method (operating condition hazard assessment method) is a semi-quantitative safety assessment method used to assess the danger and harm of operators when working in potentially dangerous environments. Although the LEC evaluation method has a wide range of applications in risk management, it also has some limitations, as follows: Highly subjective: The LEC evaluation method relies heavily on the subjective judgment and experience of the evaluator. This may lead to differences in the hazard assessment results of the same operating conditions by different evaluators, affecting the accuracy and consistency of the assessment. Not covering all hazardous factors: The LEC evaluation method focuses on the possibility of accidents, exposure frequency and consequences, but may not cover all possible hazardous factors. This may lead to incomplete assessment results. Difficult to accurately assess complex and high-risk operations: For some complex and high-risk operating environments, the LEC evaluation method may have difficulty in accurately assessing their hazards. This is because complex and high-risk operations often involve multiple interrelated risk factors, and the interactions and impacts between these factors may be difficult to accurately describe using a simple LEC model. Reliance on historical data and experience: The implementation of the LEC evaluation method needs to rely on historical data and experience. However, in some cases, historical data may be incomplete or unavailable, or experience may not be sufficient to support accurate assessments. This may affect the accuracy and reliability of the assessment results. 3. AHP: AHP is a decision analysis method that decomposes complex problems into multiple levels and factors, and obtains the relative importance of each factor through comparison and judgment. However, it also has the following limitations: It can only reveal explicit relationships: AHP can only reveal the explicit relationship between the structural level and direct components of the syntactic structure, and cannot reveal the implicit semantic structure relationship. Limited scope of application: Any theoretical method has a certain scope of application, and AHP is no exception. It may not be applicable to all types of complex problems.4. HAZOP analysis. HAZOP analysis is a structured analysis method used to identify design defects, potential process hazards and operability problems, but it also has limitations: Limited to process flow assessment: HAZOP analysis mainly evaluates process flows and may not fully consider other risks. Time-consuming: The HAZOP analysis process takes a lot of time and may not be suitable for time-sensitive projects. Lack of quantitative analysis basis: HAZOP analysis is a qualitative risk analysis method that can only analyze whether potential risks exist, but cannot accurately quantify the possibility of risk occurrence and the consequences of accidents. This may lead to distorted analysis results of the design review of the safety and operability of the device, and fail to ensure the safe operation of the chemical plant.

[0003] Risk management and hidden danger detection usually involve multiple factors, such as equipment type, process flow, personnel operation, environmental factors, etc. These factors may affect the difficulty and priority of hidden danger detection individually or together. Accurate risk management and hidden danger detection cannot be performed in the existing technology. Therefore, a method that can comprehensively consider multiple factors and their interactions is needed to determine task items and priorities, thereby improving the accuracy and effectiveness of risk management and hidden danger detection. Summary of the invention

[0004] The present invention aims to solve at least one of the technical problems in the above-mentioned technology to a certain extent. To this end, the purpose of the present invention is to propose a method for risk control and hidden danger detection based on multi-factor variance analysis to improve the accuracy and effectiveness of risk control and hidden danger detection.

[0005] To achieve the above purpose, the embodiment of the present invention proposes a method for risk control and hidden danger investigation based on multi-factor variance analysis, including:

[0006] Determine the overall objectives of risk management and hidden danger investigation, and identify and define the key variables that affect the effectiveness of risk management based on the overall objectives;

[0007] Obtain production data;

[0008] Performing data preprocessing on production data to obtain preprocessed data;

[0009] Construct a multi-factor variance analysis model based on the overall objectives and key variables;

[0010] The preprocessed data were input into the multi-factor ANOVA model to identify the key factors;

[0011] Determine risk management strategies based on key factors.

[0012] According to some embodiments of the present invention, the overall goal is to reduce the incidence of accidents;

[0013] The key variables include task items and other factors that affect the effectiveness of task items; the task items include equipment maintenance, personnel training and implementation of safety regulations; other factors include environmental conditions, working hours and employee experience.

[0014] According to some embodiments of the present invention, data preprocessing is performed on production data to obtain preprocessed data, including:

[0015] Clean production data to remove invalid or abnormal values;

[0016] Fill in missing data in production data; filling methods include using mean, median or interpolation;

[0017] The filled production data is encoded and standardized to obtain preprocessed data.

[0018] According to some embodiments of the present invention, a multi-factor variance analysis model is constructed based on the overall goal and key variables, including:

[0019] In the multi-factor ANOVA model, the dependent variables were set according to the overall objectives;

[0020] In the multi-factor ANOVA model, the independent variables and corresponding level information are set according to the key variables.

[0021] According to some embodiments of the present invention, the preprocessed data is input into a multi-factor variance analysis model to determine key factors, including:

[0022] Perform multi-factor variance analysis on the pre-processed data through the multi-factor variance analysis model based on statistical software, and calculate the F statistic and the accompanying probability P value;

[0023] Based on the F statistic and the accompanying probability P value, determine whether the influence of each variable on the dependent variable is significant; based on the judgment results, determine the key independent variable as the key factor.

[0024] According to some embodiments of the present invention, determining a risk management strategy based on key factors includes:

[0025] Prioritize key factors;

[0026] Formulate specific risk control measures based on priority ranking; risk control measures include strengthening training, optimizing processes, and improving the environment;

[0027] During the implementation of risk control measures, establish a monitoring mechanism and determine the monitoring results;

[0028] Continuously adjust and optimize risk control measures based on monitoring results to obtain the final risk control strategy.

[0029] According to some embodiments of the present invention, before inputting the preprocessed data into the multi-factor variance analysis model, the method further includes:

[0030] Perform data mining on the preprocessed data to obtain several target data;

[0031] Calculate the weight value of each target data, arrange them from large to small according to the weight value, and obtain the input data list corresponding to the preprocessed data;

[0032] The preprocessed data were entered into the multi-factor ANOVA model according to the input data list.

[0033] According to some embodiments of the present invention, data mining is performed on the preprocessed data to obtain a number of target data, including:

[0034] Perform entity recognition on preprocessed data through pre-built knowledge graphs, determine entity data and label them, extract relationships from labeled entity data, and determine the associations between labeled entity data;

[0035] According to the association relationship, the preset association relationship-mining model data table is queried to determine the target mining model;

[0036] Process the preprocessed data according to the target mining model to obtain several mining tasks;

[0037] Acquire the process information of online mining deduction for each mining task; pre-analyze the process information and establish a mining dimension parameter table for online mining deduction for each mining task; the mining dimension parameter table includes mining accuracy, mining speed, mining difficulty of the mining task, and mining tools used;

[0038] The mining result for each mining task is determined according to the mining dimension parameter table, and a number of target data are obtained according to the mining result.

[0039] According to some embodiments of the present invention, calculating the weight value of each target data includes:

[0040] After comparing several target data in pairs, determine the weight value of each target data;

[0041]

[0042] Among them, Q i is the weight value of the i-th target data; B i / B j is the i-th target data B randomly selected from M target data i With the jth target data B j Important parameters for comparison.

[0043] According to some embodiments of the present invention, the method further includes: inputting the preprocessed data into a multi-factor variance analysis model, and visually displaying the obtained analysis data through a box plot or an interaction plot.

[0044] The present invention proposes a method for risk control and hidden danger investigation based on multi-factor variance analysis. By analyzing the influence of different factors on hidden danger investigation task items and priorities, it can be determined which factors are key factors, thereby guiding the allocation of task items and the setting of priorities. In addition, multi-factor variance analysis can provide the degree of influence of each factor and its interaction on the dependent variable. In risk control hidden danger investigation, the degree of influence of these factors can be used to determine the difficulty and priority of task items. Through the significance test of the analysis results, it can be determined which factors have a significant impact on hidden danger investigation, thereby serving as an important basis for formulating task items and priorities. Using multi-factor variance analysis to determine the task items and priorities of risk control hidden danger investigation has significant advantages. 1. Comprehensiveness and accuracy. Multi-factor variance analysis can simultaneously consider the influence of multiple control variables on observed variables, which makes it comprehensive in determining the task items of risk control hidden danger investigation. By comprehensively analyzing the main effects and interaction effects of different factors, it can be more accurately identified which factors have a significant impact on risk control and hidden danger investigation. This comprehensiveness helps to avoid missing important factors, thereby improving the accuracy and effectiveness of risk control and hidden danger investigation. 2. Scientificity and objectivity. Multivariate variance analysis is based on the method of statistical inference. It determines whether the influence of each factor on the observed variable is significant by calculating the F statistic and conducting the F test. This method is scientific and objective, and can reduce the deviation caused by subjective judgment. When determining the priority of risk control and hidden danger investigation, multivariate variance analysis can provide an objective basis to help decision makers formulate priority ranking more scientifically, so as to ensure that resources are reasonably allocated and utilized. 3. Flexibility and applicability. Multivariate variance analysis is not only applicable to univariate multivariate variance analysis (univariate multivariate variance analysis), but also to multivariate multivariate variance analysis (multivariate multivariate variance analysis). This means that it can flexibly select analysis models according to actual conditions to adapt to different risk control and hidden danger investigation needs. In addition, multivariate variance analysis can also be combined with other statistical methods, such as cluster analysis, regression analysis, etc., to provide a more comprehensive risk analysis and control plan. 4. Efficiency and operability. Using statistical analysis software (such as SPSS) for multivariate variance analysis can greatly improve the efficiency and accuracy of analysis. These software usually have a user-friendly interface and powerful calculation functions. They can automatically calculate key indicators such as F statistics and accompanying probability P values, and generate detailed analysis reports. This makes multi-factor variance analysis efficient and operational in practical applications, and helps to quickly determine the tasks and priorities of risk management and hidden danger investigation. 5. Support continuous improvement and optimization. The results of multi-factor variance analysis can provide strong support for the continuous improvement and optimization of risk management and hidden danger investigation. By analyzing the degree of influence of various factors on risk management and hidden danger investigation, potential improvement points and optimization directions can be identified.For example, if a certain factor is found to have a significant impact on risk control, more specific control measures can be formulated for that factor; if a significant interaction effect is found between multiple factors, the matching and combination of these factors can be further optimized.

[0045] Other features and advantages of the present invention will be described in the following description, and partly become apparent from the description, or understood by practicing the present invention. The purpose and other advantages of the present invention can be realized and obtained by the structures particularly pointed out in the written description and the accompanying drawings.

[0046] The technical solution of the present invention is further described in detail below through the accompanying drawings and embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] The accompanying drawings are used to provide a further understanding of the present invention and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention and do not constitute a limitation of the present invention. In the accompanying drawings:

[0048] Figure 1 is a flow chart of a method for risk control and hidden danger investigation based on multi-factor variance analysis according to an embodiment of the present invention;

[0049] Figure 2 is a flowchart of constructing a multi-factor variance analysis model according to one embodiment of the present invention;

[0050] Figure 3 is a flow chart of determining key factors according to one embodiment of the present invention. DETAILED DESCRIPTION

[0051] 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.

[0052] like Figure 1 As shown, the embodiment of the present invention proposes a method for risk control and hidden danger investigation based on multi-factor variance analysis, including steps S1-S6:

[0053] S1. Determine the overall goal of risk management and hidden danger investigation, and identify and define the key variables that affect the effectiveness of risk management based on the overall goal;

[0054] S2, obtain production data;

[0055] S3, preprocessing the production data to obtain preprocessed data;

[0056] S4. Construct a multi-factor variance analysis model based on the overall objectives and key variables;

[0057] S5, input the preprocessed data into the multi-factor ANOVA model to determine the key factors;

[0058] S6. Determine risk management strategies based on key factors.

[0059] Working principle of the above technical solution: Clarifying the overall goal of risk control and hidden danger investigation will help focus and guide the subsequent steps. The overall goal may be to reduce the safety accident rate at the production site. Based on the overall goal, identify the key factors that may affect the effect of risk control. These variables may include personnel behavior, equipment status, environmental factors, management strategies, etc. Collect production data related to risk control and hidden danger investigation. These data may come from multiple channels such as on-site monitoring, equipment records, and personnel reports. Preprocess the collected construction data, and the preprocessed data will be used for subsequent multi-factor variance analysis. According to the overall goal and the identified key variables, a multi-factor analysis of variance (ANOVA) model is constructed. This model will be used to evaluate the degree of influence of different variables on the risk control effect and identify which variables are key factors. Input the preprocessed data into the multi-factor variance analysis model, run the analysis and output the results. Based on the results, determine which variables have a significant impact on the risk control effect, that is, the key factors. According to the identified key factors, formulate targeted risk control strategies. These strategies may include improving personnel training, optimizing equipment maintenance processes, and improving the working environment. Ensure that the implementation of strategies can effectively reduce risks and improve overall security levels.

[0060] Beneficial effects of the above technical solution: By analyzing the impact of different factors on hidden danger inspection task items and priorities, it is possible to determine which factors are key factors, thereby guiding the allocation of task items and the setting of priorities. In addition, multi-factor variance analysis can provide the degree of influence of each factor and its interaction on the dependent variable. In risk control hidden danger inspection, the degree of influence of these factors can be used to determine the difficulty and priority of task items. Through the significance test of the analysis results, it can be determined which factors have a significant impact on hidden danger inspection, which serves as an important basis for formulating task items and priorities. Using multi-factor variance analysis to determine the task items and priorities of risk control hidden danger inspection has significant advantages. 1. Comprehensiveness and accuracy. Multi-factor variance analysis can simultaneously consider the impact of multiple control variables on observed variables, which makes it comprehensive in determining the task items of risk control hidden danger inspection. By comprehensively analyzing the main effects and interaction effects of different factors, it is possible to more accurately identify which factors have a significant impact on risk control and hidden danger inspection. This comprehensiveness helps to avoid missing important factors, thereby improving the accuracy and effectiveness of risk control and hidden danger inspection. 2. Scientificity and objectivity. Multivariate variance analysis is based on the method of statistical inference. It determines whether the influence of each factor on the observed variable is significant by calculating the F statistic and conducting the F test. This method is scientific and objective, and can reduce the deviation caused by subjective judgment. When determining the priority of risk control and hidden danger investigation, multivariate variance analysis can provide an objective basis to help decision makers formulate priority ranking more scientifically, so as to ensure that resources are reasonably allocated and utilized. 3. Flexibility and applicability. Multivariate variance analysis is not only applicable to univariate multivariate variance analysis (univariate multivariate variance analysis), but also to multivariate multivariate variance analysis (multivariate multivariate variance analysis). This means that it can flexibly select analysis models according to actual conditions to adapt to different risk control and hidden danger investigation needs. In addition, multivariate variance analysis can also be combined with other statistical methods, such as cluster analysis, regression analysis, etc., to provide a more comprehensive risk analysis and control plan. 4. Efficiency and operability. Using statistical analysis software (such as SPSS) for multivariate variance analysis can greatly improve the efficiency and accuracy of analysis. These software usually have a user-friendly interface and powerful calculation functions. They can automatically calculate key indicators such as F statistics and accompanying probability P values, and generate detailed analysis reports. This makes multi-factor variance analysis efficient and operational in practical applications, and helps to quickly determine the tasks and priorities of risk management and hidden danger investigation. 5. Support continuous improvement and optimization. The results of multi-factor variance analysis can provide strong support for the continuous improvement and optimization of risk management and hidden danger investigation. By analyzing the degree of influence of various factors on risk management and hidden danger investigation, potential improvement points and optimization directions can be identified.For example, if a certain factor is found to have a significant impact on risk control, more specific control measures can be formulated for that factor; if a significant interaction effect is found between multiple factors, the matching and combination of these factors can be further optimized. By clarifying goals, identifying key variables, collecting and analyzing data, and formulating targeted strategies, risks can be effectively reduced and overall safety performance can be improved.

[0061] According to some embodiments of the present invention, the overall goal is to reduce the incidence of accidents;

[0062] The key variables include task items and other factors that affect the effectiveness of task items; the task items include equipment maintenance, personnel training and implementation of safety regulations; other factors include environmental conditions, working hours and employee experience.

[0063] The working principle and beneficial effects of the above technical solutions: Task items: These are the main activities or processes that directly affect the accident rate. Equipment maintenance: Including regular inspection, repair and maintenance of equipment to ensure that the equipment is in good operating condition. Personnel training: Provide necessary safety education and skills training to improve employees' safety awareness and operating skills. Implementation of safety regulations: Ensure that all employees strictly abide by safety regulations and operating procedures to reduce the risks caused by illegal operations. Other factors: These are external conditions that may affect the effectiveness of task items and thus indirectly affect the accident rate. Environmental conditions: Such as lighting, ventilation, temperature and humidity in the workplace, these factors may affect employees' work efficiency and attention. Working hours: Including working hours, shift arrangements, etc. Long working hours or fatigue may lead to decreased attention and increased errors. Employee experience: Employees' work experience and skill level may affect their ability to respond to emergencies and handle complex tasks.

[0064] According to some embodiments of the present invention, data preprocessing is performed on production data to obtain preprocessed data, including:

[0065] Clean production data to remove invalid or abnormal values;

[0066] Fill in missing data in production data; filling methods include using mean, median or interpolation;

[0067] The filled production data is encoded and standardized to obtain preprocessed data.

[0068] Working principle and beneficial effects of the above technical solution: Remove invalid or abnormal values ​​to ensure the accuracy and consistency of data. Check invalid values: Identify and delete or replace data that obviously does not conform to logic or business rules. For example, negative numbers or values ​​outside the reasonable range appear in the age field. Handle abnormal values: Analyze the source of abnormal values. If they are caused by data entry errors, they can be deleted or replaced with reasonable values; if abnormal values ​​represent real situations (such as data under extreme weather conditions), they may need to be retained and specially processed. Fill missing data to ensure the integrity of data and avoid affecting subsequent analysis due to missing values. Mean filling: Use the mean of the entire variable to fill missing values. This method is simple but may introduce bias, especially when there are many missing values ​​or the distribution is uneven. Median filling: Use the median of the variable to fill missing values. This method is more robust than mean filling because it is less sensitive to extreme values. Interpolation: Estimate missing values ​​based on the values ​​of adjacent data points. This method is suitable for time series data or data with obvious trends. Encoding processing converts non-numeric data (such as text data) into numerical data for mathematical and statistical analysis. Categorical variables are converted to numerical vectors by using One-Hot Encoding, or mapped to integers by using Label Encoding. Standardization is a process that makes different variables have the same scale so that they are comparable during comparison and analysis. Methods include: Z-score standardization: Subtract the mean of each variable from its value and then divide it by the standard deviation. This method makes the processed data have zero mean and unit standard deviation. Min-Max standardization: Scale the value of each variable to a specified range (such as 0 to 1). This method retains the original distribution of the data but changes the scale of the data.

[0069] like Figure 2 As shown, according to some embodiments of the present invention, a multi-factor variance analysis model is constructed according to the overall goal and key variables, including steps S41-S42:

[0070] S41. Set dependent variables according to overall objectives in the multi-factor ANOVA model;

[0071] S42. In the multi-factor ANOVA model, set the independent variables and corresponding level information based on the key variables.

[0072] Working principle and beneficial effects of the above technical solution: The accident rate (expressed in percentage, frequency or number of times, etc.) is set as the dependent variable. In the multi-factor variance analysis model, the independent variables and corresponding level information are set according to the key variables. Task items: Equipment maintenance: Different maintenance levels can be set, such as "frequent maintenance", "regular maintenance" and "rare maintenance". Personnel training: Different training levels can be set, such as "regular training", "occasional training" and "no training". Safety regulations implementation: Different implementation levels can be set, such as "strict implementation", "partial implementation" and "no implementation". Other factors: Environmental conditions: Different environmental conditions can be set, such as "good", "general" and "bad", or more specifically describe the environmental conditions (such as temperature, humidity, etc.). Working hours: Different time periods or shifts can be set, such as "daytime", "night shift" and "overtime". Employee experience: Different experience levels can be set, such as "novice", "experienced" and "senior". Level information: For each independent variable, the specific meaning and classification criteria of each level are clarified. These levels represent the different values ​​or states that the independent variable can take in the experiment.

[0073] like Figure 3 As shown, according to some embodiments of the present invention, the preprocessed data is input into a multi-factor variance analysis model to determine key factors, including steps S51-S52:

[0074] S51, performing multi-factor variance analysis on the pre-processed data through a multi-factor variance analysis model based on statistical software, and calculating the F statistic and the accompanying probability P value;

[0075] S52. Determine whether the influence of each variable on the dependent variable is significant based on the F statistic and the accompanying probability P value; determine the key independent variable as the key factor based on the judgment result.

[0076] The working principle and beneficial effects of the above technical solution: Input the preprocessed data into statistical software, such as SPSS, SAS, R, etc. Generally, the data should be presented in a table, in which each column represents a variable (including dependent variables and independent variables) and each row represents an observation. Set the multi-factor ANOVA model in the statistical software, specify the dependent variable and independent variables, and any possible interaction terms. Run the multi-factor ANOVA in the statistical software, and the software will automatically calculate the F statistic and the accompanying probability P value. The F statistic is used to measure the degree of influence of the independent variable on the dependent variable. It is the ratio of the between-group variation to the within-group variation, which is used to test whether the null hypothesis (that is, all independent variables have no significant effect on the dependent variable) is valid. The P value is used to evaluate the significance of the F statistic. If the P value is less than the predetermined significance level (such as 0.05), the null hypothesis is rejected, and it is believed that at least one independent variable has a significant effect on the dependent variable. Single factor test: First, the effect of each independent variable on the dependent variable can be tested separately. If the P value of an independent variable is less than the significance level, it is believed that the independent variable has a significant effect on the dependent variable. Interaction effect test: Next, the interaction effect between the independent variables should be tested. If the P value of the interaction term of two or more independent variables is less than the significance level, it is considered that there is a significant interaction effect between them. Judgment based on significance: According to the results of the single-factor test and the interaction effect test, determine which independent variables have a significant effect on the dependent variable. These independent variables will be considered as key factors. Consider business significance: When determining key factors, in addition to statistical significance, the significance and importance of these factors in actual business should also be considered. Sometimes, even if the P value of an independent variable is slightly higher than the significance level, it may be considered as a key factor if it has an important impact on the business. When performing multi-factor ANOVA, you should ensure that the data meets the assumptions of the model, such as normality, homogeneity of variance, and independence. Interpret the results: Interpret the results of the multi-factor ANOVA to explain which independent variables are key factors and how they affect the dependent variable.

[0077] According to some embodiments of the present invention, determining a risk management strategy based on key factors includes:

[0078] Prioritize key factors;

[0079] Formulate specific risk control measures based on priority ranking; risk control measures include strengthening training, optimizing processes, and improving the environment;

[0080] During the implementation of risk control measures, establish a monitoring mechanism and determine the monitoring results;

[0081] Continuously adjust and optimize risk control measures based on monitoring results to obtain the final risk control strategy.

[0082] Working principle and beneficial effects of the above technical solution: Set priority ranking according to the impact, urgency and feasibility of the task items. Task items with large impact, urgency and easy implementation will be given higher priority. For each key task item, formulate specific risk control measures, such as strengthening training, optimizing processes, improving the environment, etc. Put the measures into practice, establish a monitoring mechanism, and regularly evaluate the results to ensure the effectiveness of the measures. According to the monitoring results, continuously adjust and optimize the risk control strategy to form a closed loop of continuous improvement. Integrate the optimized and adjusted risk control measures to form the final risk control strategy.

[0083] According to some embodiments of the present invention, before inputting the preprocessed data into the multi-factor variance analysis model, the method further includes:

[0084] Perform data mining on the preprocessed data to obtain several target data;

[0085] Calculate the weight value of each target data, arrange them from large to small according to the weight value, and obtain the input data list corresponding to the preprocessed data;

[0086] The preprocessed data were entered into the multi-factor ANOVA model according to the input data list.

[0087] The working principle and beneficial effects of the above technical solution are as follows: Before the pre-processed data is input into the multi-factor variance analysis model, data mining is performed on the pre-processed data to obtain a number of target data, which are key information or features related to the overall goal. The weight value of each target data is calculated, and the data is arranged from large to small according to the weight value to obtain the input data list corresponding to the pre-processed data; the input data list is based on the data priority judgment made in advance based on data mining, and the input order of the data is determined, which is convenient for improving the accuracy of data analysis.

[0088] According to some embodiments of the present invention, data mining is performed on the preprocessed data to obtain a number of target data, including:

[0089] Perform entity recognition on preprocessed data through pre-built knowledge graphs, determine entity data and label them, extract relationships from labeled entity data, and determine the associations between labeled entity data;

[0090] According to the association relationship, the preset association relationship-mining model data table is queried to determine the target mining model;

[0091] Process the preprocessed data according to the target mining model to obtain several mining tasks;

[0092] Acquire the process information of online mining deduction for each mining task; pre-analyze the process information and establish a mining dimension parameter table for online mining deduction for each mining task; the mining dimension parameter table includes mining accuracy, mining speed, mining difficulty of the mining task, and mining tools used;

[0093] The mining result for each mining task is determined according to the mining dimension parameter table, and a number of target data are obtained according to the mining result.

[0094] The working principle and beneficial effects of the above technical solution are as follows: Entity recognition is performed on preprocessed data using a pre-built knowledge graph. Knowledge graph is a structured knowledge representation method that can describe entities and their relationships. The identified entity data will be labeled, and these labels will help with subsequent relationship extraction and data analysis. Relationship extraction is performed on the labeled entity data to determine the associations between them. These relationships can be direct (such as parent-child relationships, inclusion relationships) or indirect (such as relationships connected through other entities). According to the extracted associations, the preset association-mining model data table is queried. This data table contains the mapping relationship between various associations and corresponding mining models. Based on the query results, the mining model suitable for the current data is determined. This model will be used to guide subsequent mining tasks. The process information of each mining task for online mining deduction is obtained, including task status, execution time, resource consumption, etc. The process information is pre-analyzed to establish a mining dimension parameter table. This parameter table includes key indicators such as mining accuracy, mining speed, mining difficulty of the mining task, and the mining tools used. According to the mining dimension parameter table, the mining results of each mining task are evaluated. This includes judging the accuracy, completeness and practicality of the mining results. The mining results for each mining task are determined according to the mining dimension parameter table, and a number of target data are obtained according to the mining results. The target data is valid data. It is convenient to accurately determine the valid data and facilitate data analysis.

[0095] According to some embodiments of the present invention, calculating the weight value of each target data includes:

[0096] After comparing several target data in pairs, determine the weight value of each target data;

[0097]

[0098] Among them, Q i is the weight value of the i-th target data; B i / B j is the i-th target data B randomly selected from M target data i With the jth target data B j Important parameters for comparison.

[0099] The beneficial effect of the above technical solution is that after comparing several target data in pairs, it is convenient to accurately determine the weight value of each target data, thereby improving the accuracy of sorting.

[0100] According to some embodiments of the present invention, the method further includes: inputting the preprocessed data into a multi-factor variance analysis model, and visually displaying the obtained analysis data through a box plot or an interaction plot.

[0101] The beneficial effect of the above technical solution is that the analysis data obtained by the multi-factor variance analysis model is visualized through a box plot or an interaction plot, so as to more intuitively understand the data characteristics and the interaction between factors.

[0102] 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 risk control and hidden danger investigation based on multi-factor variance analysis, characterized in that: include: Determine the overall objectives of risk management and hidden danger investigation, and identify and define the key variables that affect the effectiveness of risk management based on the overall objectives; Obtain production data; Performing data preprocessing on production data to obtain preprocessed data; Construct a multi-factor variance analysis model based on the overall objectives and key variables; The preprocessed data were input into the multi-factor ANOVA model to identify the key factors; Determine risk management strategies based on key factors.

2. The method for risk control and hidden danger investigation based on multi-factor variance analysis as claimed in claim 1, characterized in that: The overall goal stated is to reduce the accident rate; The key variables include task items and other factors that affect the effectiveness of task items; the task items include equipment maintenance, personnel training and implementation of safety regulations; other factors include environmental conditions, working hours and employee experience.

3. The method for risk control and hidden danger investigation based on multi-factor variance analysis as claimed in claim 1, characterized in that: Perform data preprocessing on production data to obtain preprocessed data, including: Clean production data to remove invalid or abnormal values; Fill in missing data in production data; filling methods include using mean, median or interpolation; The filled production data is encoded and standardized to obtain preprocessed data.

4. The method for risk control and hidden danger investigation based on multi-factor variance analysis according to claim 1, characterized in that: Construct a multi-factor variance analysis model based on the overall objectives and key variables, including: In the multi-factor ANOVA model, the dependent variables were set according to the overall objectives; In the multi-factor ANOVA model, the independent variables and corresponding level information are set according to the key variables.

5. The method for risk control and hidden danger investigation based on multi-factor variance analysis as claimed in claim 1, characterized in that: The pre-processed data were input into a multi-factor ANOVA model to identify key factors, including: Perform multi-factor variance analysis on the pre-processed data through the multi-factor variance analysis model based on statistical software, and calculate the F statistic and the accompanying probability P value; Based on the F statistic and the accompanying probability P value, determine whether the influence of each variable on the dependent variable is significant; based on the judgment results, determine the key independent variable as the key factor.

6. The method for risk control and hidden danger investigation based on multi-factor variance analysis as claimed in claim 5, characterized in that: Determine risk management strategies based on key factors, including: Prioritize key factors; Formulate specific risk control measures based on priority ranking; risk control measures include strengthening training, optimizing processes, and improving the environment; During the implementation of risk control measures, establish a monitoring mechanism and determine the monitoring results; Continuously adjust and optimize risk control measures based on monitoring results to obtain the final risk control strategy.

7. The method for risk control and hidden danger investigation based on multi-factor variance analysis according to claim 1, characterized in that: Before inputting the preprocessed data into the multi-factor ANOVA model, it also includes: Perform data mining on the preprocessed data to obtain several target data; Calculate the weight value of each target data, arrange them from large to small according to the weight value, and obtain the input data list corresponding to the preprocessed data; The preprocessed data were entered into the multi-factor ANOVA model according to the input data list.

8. The method for risk control and hidden danger investigation based on multi-factor variance analysis as claimed in claim 7, characterized in that: Data mining is performed on the preprocessed data to obtain several target data, including: Perform entity recognition on preprocessed data through pre-built knowledge graphs, determine entity data and label them, extract relationships from labeled entity data, and determine the associations between labeled entity data; According to the association relationship, the preset association relationship-mining model data table is queried to determine the target mining model; Process the preprocessed data according to the target mining model to obtain several mining tasks; Acquire the process information of online mining deduction for each mining task; pre-analyze the process information and establish a mining dimension parameter table for online mining deduction for each mining task; the mining dimension parameter table includes mining accuracy, mining speed, mining difficulty of the mining task, and mining tools used; The mining result for each mining task is determined according to the mining dimension parameter table, and a number of target data are obtained according to the mining result.

9. The method for risk control and hidden danger investigation based on multi-factor variance analysis according to claim 7, characterized in that: Calculate the weight value of each target data, including: After comparing several target data in pairs, determine the weight value of each target data; Among them, Q i is the weight value of the i-th target data; B i / B j is the i-th target data B randomly selected from M target data i With the jth target data B j Important parameters for comparison.

10. The method for risk control and hidden danger investigation based on multi-factor variance analysis according to claim 1, characterized in that: Also includes: The preprocessed data were input into the multi-factor ANOVA model, and the obtained analysis data were visualized through box plots or interaction plots.

Images