A fire disaster data verification method based on correlation principle
Through the fire disaster data verification method based on the principle of correlation, the Pearson product moment calculation and weighted correlation coefficient algorithm are used to solve the problem of low accuracy in the authenticity verification of fire statistics, and the authenticity and credibility of data are improved. It reflects the fire safety work situation and promotes the healthy development of the fire safety industry.
Patent Information
- Application Number
- CN202210173494.8
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2022-02-24
- Publication Date
- 2024-11-26
- Estimated Expiration
- 2042-02-24
AI Technical Summary
The existing technology has problems of low accuracy and lack of authenticity in the verification of fire statistics. Especially due to the lack of scientificity of the absolute number control of fire deaths, it is easy to conceal and misreport problems, resulting in deviations in statistical data. .
The fire disaster data verification method based on the principle of correlation is used to calculate the linear correlation coefficient and partial correlation coefficient of the total fire fighting, total rescue, total social assistance and total dispatch through the Pearson product moment, combined with the weighted correlation coefficient algorithm , judge the authenticity of fire data.
It improves the authenticity and credibility of fire data, can objectively reflect the regional fire safety work situation, reduce statistical data deviations, and promote the healthy development of the fire protection industry.
Smart Images

Figure CN114511246B_ABST
Abstract
Description
Technical Field
[0001] The invention relates to a fire disaster data verification method based on the correlation principle, belonging to the technical field of data analysis. Background Art
[0002] With the development of social economy, various new business forms and new economic types are constantly emerging. After a fire occurs in a related place, how to more accurately describe, define and classify the fire attributes under the new situation, and how to use cloud computing and big data technology to more accurately analyze and explore the mechanism and trend behind the fire has become a practical problem that needs to be solved urgently. How to verify the authenticity of fire statistics is the basis of various data analysis. The current fire authenticity verification faces the following problems: firefighting work has a strong social nature. The government, departments, units and citizens are all responsible for firefighting work and bear different fire safety responsibilities; and the occurrence of most fires is not transferred by human will, especially some fire accidents that cause deaths are highly random and sporadic in time and space. However, the absolute number of fire deaths is essentially lacking in scientificity and does not conform to the objective laws of firefighting work. The four fire data are used to evaluate firefighting work, and the query accuracy is not high. Once there is a problem of concealment and underreporting, it will lead to statistical data deviation. Summary of the invention
[0003] In order to solve the above-mentioned technical problems, the present invention provides a fire disaster data verification method based on the correlation principle.
[0004] The object of the present invention is achieved through the following technical solution: a fire disaster data verification method based on the correlation principle, characterized in that it includes the following steps:
[0005] The total fire fighting amount i, the total emergency rescue amount m, the total social assistance amount h, and the total dispatch amount j are used as input sources to perform Pearson product moment calculation. The specific steps are as follows:
[0006] Step 1: Select standard sample data to calculate the linear correlation coefficient. The linear correlation coefficient calculation formula is as follows:
[0007] Calculate the two linear correlation coefficients r between the total fire fighting and the total emergency rescue, and between the total fire fighting and the total dispatch im and r ij The average value V1;
[0008]
[0009] That is r im 、r ij All of them can be substituted into formula (1) to calculate the corresponding correlation coefficient;
[0010] When calculating r imWhen X i is the total amount of fire fighting in a period of time, Y i The total amount of emergency rescue over a period of time;
[0011] When calculating r ij When X i is the total amount of fire fighting in a period of time, Y i Total amount dispatched for a period of time;
[0012] Select multiple time periods to find the correlation coefficient;
[0013] Step 2: Select standard sample data to calculate the partial correlation coefficient. The calculation formula for the partial correlation coefficient is as follows:
[0014]
[0015] V2 is the second-order partial correlation, that is, the total amount of fire fighting is i, the total amount of emergency rescue is m, the total amount of social assistance is h, and the total amount of dispatch is j, where r ij.h It is the second-order partial correlation coefficient of the total fire fighting, total dispatch and total social assistance, and the formula is as follows:
[0016]
[0017] where r ij is the first-order partial correlation coefficient between the total amount of fire fighting and the total amount of emergency rescue;
[0018] r ih is the first-order partial correlation coefficient between the total amount of fire fighting and the total amount of social assistance;
[0019] r jh is the first-order partial correlation coefficient between the total number of dispatches and the total amount of social assistance;
[0020] Step 3: Calculate the weighted correlation coefficient for fire data quality. The weighted correlation coefficient calculation formula is as follows:
[0021] V3=V1×k1+V2×k2 (k1=0.7, k2=0.3)
[0022] Step 4: According to the basic principles of mathematical statistics, a correlation coefficient of 0.8-1.0 indicates a very strong correlation, 0.6-0.8 indicates a strong correlation, 0.4-0.6 indicates a moderate correlation, 0.2-0.4 indicates a weak correlation, and 0.0-0.2 indicates a very weak correlation or no correlation.
[0023] According to the correlation coefficient values of V1, V2, and V3, the correlation of fire data is obtained:
[0024] If V2>0.2 and V1≥0.6, the authenticity of the fire data is highly reliable;
[0025] If V2>0.2, V1<0.6, and V3≥0.6, the authenticity of the fire data is highly reliable;
[0026] If V2>0.2, V1<0.6, 0.4≤V3<0.6, the authenticity and credibility of the fire data are average;
[0027] If V2>0.2, V1<0.6, V3<0.2, or V2≤0.2, the authenticity of the fire data is less reliable.
[0028] The method of the present invention is based on the basic principle of data statistics, and utilizes the principle of correlation and basic objective assumptions to focus on solving the problem of low accuracy and lack of authenticity of fire disaster data. Through the fire disaster data verification method based on the principle of correlation, the effective verification of disaster data is achieved, which can objectively and truly reflect the status of regional fire safety work and also has many beneficial effects on the healthy development of the fire protection industry. BRIEF DESCRIPTION OF THE DRAWINGS
[0029] Figure 1 It is a flow chart of credibility determination of authenticity of fire data of the present invention. DETAILED DESCRIPTION
[0030] The basic design principles and technical solutions of the present invention are as follows:
[0031] (1) Basic assumptions about the correlation of disaster data
[0032] The number of fires is an important disaster in the fire rescue responsibilities and can objectively and truly reflect the status of regional fire safety work. During the many exchanges and surveys with fire rescue personnel from various places, we have obtained the following basic assumptions:
[0033] 1. Fire data is often correlated with the total amount of disasters and the number of dispatches.
[0034] 2. Due to the influence of interference factors such as emergency rescue, social assistance, etc., the fire data has a partial correlation with the total number of dispatches.
[0035] (2) Fire disaster data verification technology based on correlation analysis
[0036] The present invention uses Pearson product moment to quantify the linear correlation coefficient between fire data and the number of dispatches.
[0037] 1. Pearson correlation coefficient, also known as Pearson product-moment correlation coefficient, is a linear correlation coefficient and the most commonly used correlation coefficient. We calculate the total amount of fire fighting, total amount of emergency rescue, total amount of social assistance, total amount of dispatched teams, total amount of dispatched vehicles, total amount of dispatched personnel, etc. The correlation coefficient r is used to quantify the linear correlation between fire data and the number of dispatches.
[0038] 2. When other variables are fixed, that is, they are controlled, the correlation coefficient between any two given variables is called the partial correlation coefficient, which is the statistic that truly reflects the correlation between the two variables. Here, we control the impact of the total amount of emergency rescue and the total amount of social assistance, and use the second-order partial correlation coefficient to quantify the partial correlation between the fire data and the number of dispatches.
[0039] 3. Weighted correlation algorithm for fire data quality.
[0040] The second-order partial correlation coefficient can only reflect the data quality of the total number of fires and the number of dispatches in the existing data, but cannot reflect the situation of underreporting and concealment. This method designs a correlation weighting algorithm to verify the authenticity of fire data by combining linear correlation and partial correlation.
[0041] 4. Credibility determination process of the authenticity of fire data:
[0042] The total amount of fire fighting i, the total amount of emergency rescue m, the total amount of social assistance h, the total amount of dispatched personnel j and other data are used as input sources for Pearson product moment calculation. Based on the basic assumption about the correlation of disaster data, the authenticity of the fire data is verified by the correlation coefficient of factors such as the total amount of fire fighting and the total number of dispatched personnel.
[0043] The specific steps are as follows:
[0044] Step 1: Select standard sample data to calculate the linear correlation coefficient. The linear correlation coefficient calculation formula is as follows:
[0045] Calculate the two linear correlation coefficients r between the total fire fighting and the total emergency rescue, and between the total fire fighting and the total dispatch im and r ij The average value V1;
[0046]
[0047] That is r im 、r ij All of them can be substituted into formula (1) to calculate the corresponding correlation coefficient;
[0048] When calculating r im When X i is the total amount of fire fighting in a period of time, Y i The total amount of emergency rescue over a period of time;
[0049] When calculating r ij When X i is the total amount of fire fighting in a period of time, Y i Total amount dispatched for a period of time;
[0050] Select multiple time periods to find the correlation coefficient;
[0051] Step 2: Select standard sample data to calculate the partial correlation coefficient. The calculation formula for the partial correlation coefficient is as follows:
[0052]
[0053] V2 is the second-order partial correlation, that is, the total amount of fire fighting is i, the total amount of emergency rescue is m, the total amount of social assistance is h, and the total amount of dispatch is j, where r ij.h It is the second-order partial correlation coefficient of the total fire fighting, total dispatch and total social assistance, and the formula is as follows:
[0054]
[0055] where r ij is the first-order partial correlation coefficient between the total amount of fire fighting and the total amount of emergency rescue;
[0056] ri h is the first-order partial correlation coefficient between the total amount of fire fighting and the total amount of social assistance;
[0057] rjh is the first-order partial correlation coefficient of the total number of dispatches and the total amount of social assistance;
[0058] Step 3: Calculate the weighted correlation coefficient for fire data quality. The weighted correlation coefficient calculation formula is as follows:
[0059] V3=V1×k1+V2Xk2 (k1=0.7, k2=0.3)
[0060] Step 4: According to the basic principles of mathematical statistics, a correlation coefficient of 0.8-1.0 indicates a very strong correlation, 0.6-0.8 indicates a strong correlation, 0.4-0.6 indicates a moderate correlation, 0.2-0.4 indicates a weak correlation, and 0.0-0.2 indicates a very weak correlation or no correlation.
Claims
1. A fire disaster data verification method based on the correlation principle, It is characterized in that The steps include: The total fire fighting amount i, the total emergency rescue amount m, the total social assistance amount h, and the total dispatch amount j are used as input sources to perform Pearson product moment calculation. The specific steps are as follows: Step 1: Select standard sample data to calculate the linear correlation coefficient. The linear correlation coefficient calculation formula is as follows: Calculate the two linear correlation coefficients r between the total fire fighting and the total emergency rescue, and between the total fire fighting and the total dispatch im and r ij The average value V1; That is r im 、r ij All of them can be substituted into formula (1) to calculate the corresponding correlation coefficient; When calculating r im When X i is the total amount of fire fighting in a period of time, Y i The total amount of emergency rescue over a period of time; When calculating r ij When X i is the total amount of fire fighting in a period of time, Y i Total amount dispatched for a period of time; Select multiple time periods to find the correlation coefficient; Step 2: Select standard sample data to calculate the partial correlation coefficient. The calculation formula for the partial correlation coefficient is as follows: V2 is the second-order partial correlation, that is, the total amount of fire fighting is i, the total amount of emergency rescue is m, the total amount of social assistance is h, and the total amount of dispatch is j, where r ij.h It is the second-order partial correlation coefficient of the total fire fighting, total dispatch and total social assistance, and the formula is as follows: where r ij is the first-order partial correlation coefficient between the total amount of fire fighting and the total amount of emergency rescue; r ih is the first-order partial correlation coefficient between the total amount of fire fighting and the total amount of social assistance; r jh is the first-order partial correlation coefficient between the total number of dispatches and the total amount of social assistance; Step 3: Calculate the weighted correlation coefficient for fire data quality. The weighted correlation coefficient calculation formula is as follows: V3=V1×k1+V2×k2 (k1=0.7, k2=0.3) Step 4: According to the basic principles of mathematical statistics, a correlation coefficient of 0.8-1.0 indicates a very strong correlation, 0.6-0.8 indicates a strong correlation, 0.4-0.6 indicates a moderate correlation, 0.2-0.4 indicates a weak correlation, and 0.0-0.2 indicates a very weak correlation or no correlation.
2. According to claim 1, a fire disaster data verification method based on the correlation principle, It is characterized in that According to the correlation coefficient values of V1, V2, and V3, the correlation of fire data is obtained: If V2>0.2 and V1≥0.6, the authenticity of the fire data is highly reliable; If V2>0.2, V1<0.6, and V3≥0.6, the authenticity of the fire data is highly reliable; If V2>0.2, V1<0.6, 0.4≤V3<0.6, the authenticity and credibility of the fire data are average; If V2>0.2, V1<0.6, V3<0.2, or V2≤0.2, the authenticity of the fire data is less reliable.
