Security risk assessment and early warning method

Abstract

The invention discloses a safety risk assessment and early warning method, which relates to the technical field of safety assessment, and is characterized in that a temperature and humidity sensor outputs temperature and humidity change data through a vibration sensor dynamic signal; performing feature extraction and preprocessing on the vibration signal based on a feature extraction module, and outputting a feature vector; the risk calculation module calculates an output risk index and generates a preliminary risk level; and the threshold value judgment module outputs a dynamic threshold value, and judges and outputs a final risk level by comparing the risk index with the dynamic threshold value. According to the invention, the vibration characteristics generated by the dangerous chemicals can be detected in advance before the temperature, air pressure or gas concentration of the dangerous chemicals is obviously changed, so that the leakage of the dangerous chemicals is effectively prevented; by generating the dynamic threshold, the method can better fit the actual warehouse condition and environment state, is more accurate in judgment of the risk index, can make the safety risk assessment better fit the reality, and further improves the accuracy of the safety risk assessment.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of safety assessment, more particularly, it relates to a safety risk assessment and early warning method. BACKGROUND

[0002] Dangerous chemicals have the properties of explosion, flammability, toxicity, corrosion and radioactivity, etc., which may cause personnel casualties and property losses in the process of transportation, storage and production. Dangerous chemicals are widely used in many fields such as industry, agriculture, medicine and construction. While promoting economic development, they also bring potential safety risks. The occurrence of accidents poses a serious threat to society and the environment. The number of personnel casualties and economic losses caused by dangerous chemical accidents around the world every year is incalculable. In particular, in some developing countries, the management and safety prevention of dangerous chemicals still face great challenges.

[0003] In order to cope with the potential danger of stored dangerous chemicals, the existing detection method sets multiple sensors to detect dangerous chemicals, among which gas sensor detection is mainly used, which can directly reflect whether dangerous chemicals leak or not. However, this detection has certain limitations. When the gas sensor detects the change of gas concentration, it means that the dangerous chemicals have leaked, which has a certain hysteresis and cannot prevent the leakage of dangerous chemicals in advance. At the same time, the existing threshold judgment is usually a fixed value, which cannot be adjusted according to the actual situation, so there may be a large error.

[0004] Therefore, a new scheme needs to be proposed to solve this problem. SUMMARY

[0005] The purpose of the embodiment of the present application is to provide a safety risk assessment and early warning method to solve the above problems.

[0006] The above technical purpose of the embodiment of the present application is realized by the following technical scheme: a safety risk assessment and early warning method, the safety risk assessment and early warning method comprising the following steps:

[0007] A plurality of vibration sensors are arranged to form a vibration sensor array, and the vibration signals of the warehouse structure are collected through the vibration sensor array. A temperature and humidity sensor is arranged to collect the temperature and humidity changes of the warehouse and output the temperature and humidity change data.

[0008] The feature extraction module extracts and preprocesses the vibration signals and outputs the feature vector.

[0009] The risk calculation module obtains the feature vector, calculates the output risk index and generates a preliminary risk level.

[0010] The threshold judgment module obtains the temperature and humidity change data, outputs a dynamic threshold, compares the risk index with the dynamic threshold, and judges and outputs a final risk level;

[0011] The early warning output module obtains the final risk level, and outputs an early warning signal and a disposal suggestion.

[0012] The application is further provided that the feature extraction module extracts and pre-processes the vibration signal, and outputs a feature vector including:

[0013] The amplitude mean value, the root mean square value and the waveform factor in the vibration signal are extracted;

[0014] The pre-processing includes normalizing the amplitude mean value, the root mean square value and the waveform factor to obtain feature vectors F1, F2 and F3, wherein F1 represents the normalized value of the amplitude mean value, F2 represents the normalized value of the root mean square value, and F3 represents the normalized value of the waveform factor.

[0015] The application is further provided that the risk calculation module obtains the feature vector, calculates and outputs a risk index, and generates a preliminary risk level including:

[0016] The risk calculation module adopts a weighted fusion algorithm to fuse multiple feature vectors, and the risk index calculation formula is:

[0017] R = w1 · F1 + w2 · F2 + w3 · F3 + α · log (1 + σ 2 )

[0018] wherein R represents the risk index, w1, w2 and w3 represent feature weights, the sizes of the feature weights are determined by an entropy weight method, α represents an adjustment coefficient, and σ 2 represents a feature fluctuation variance;

[0019] The output preliminary risk level includes that when R≤0.3, the preliminary risk level is level one; when 0.3<R≤0.6, the preliminary risk level is level two; when 0.6<R≤0.8, the preliminary risk level is level three; when 0.8<R≤0.9, the preliminary risk level is level four; and when R>0.9, the preliminary risk level is level five.

[0020] The application is further provided that the threshold judgment module obtains the temperature and humidity change data, and outputs a dynamic threshold including:

[0021] The threshold judgment module adopts a dynamic threshold strategy and dynamically adjusts according to environmental factors, and the dynamic threshold includes:

[0022] T = T0 · (1 + k · ΔE)

[0023] Wherein, T represents a dynamic threshold value, T0 represents a basic threshold value, which is set according to historical data; k represents an environmental influence coefficient; and ΔE represents temperature and humidity change data.

[0024] The application further provides that the determining and outputting a final risk level by comparing the risk index with the dynamic threshold value comprises:

[0025] comparing the risk index with the dynamic threshold value,

[0026] when the R is greater than or equal to the T, triggering a higher level of warning, and the final risk level output is one level higher than the preliminary risk level;

[0027] when the R is less than the T, maintaining or reducing the warning level, and the final risk level output is the same level as the preliminary risk level or one level lower than the preliminary risk level.

[0028] The application further provides that the warning output module acquires the final risk level and outputs a warning signal and a disposal suggestion, wherein:

[0029] when the final risk level is level one, the warning signal is blue, and a routine inspection is suggested, and the warning needs to be handled within 24 hours after being issued;

[0030] when the final risk level is level two, the warning signal is green, and a strengthened inspection frequency is suggested, and the warning needs to be handled within 12 hours after being issued;

[0031] when the final risk level is level three, the warning signal is yellow, and an on-site verification and reason analysis are suggested, and the warning needs to be handled within 4 hours after being issued;

[0032] when the final risk level is level four, the warning signal is orange, and an emergency team is suggested to be sent for intervention, and some areas are suggested to be isolated, and the warning needs to be handled within 1 hour after being issued;

[0033] when the final risk level is level five, the warning signal is red, and all personnel are suggested to evacuate, and a comprehensive emergency response is suggested to be started, and the warning needs to be handled immediately after being issued.

[0034] The application further provides that the computer readable storage medium stores a computer program, and the computer program is executed by a computer or a processor to realize the method.

[0035] The application further provides that the computer program product comprises a computer program, and when the computer program is executed by a computer or a processor, the computer or the processor executes the method.

[0036] In summary, the application has the following beneficial effects:

[0037] By adopting the vibration sensor to detect the dangerous chemicals, the development of the structural crack and the tiny leakage can be accurately detected, the vibration characteristics generated by the dangerous chemicals are detected in advance before the temperature, air pressure or gas concentration of the dangerous chemicals changes significantly, and corresponding actions can be made, so that the leakage of the dangerous chemicals is effectively prevented.

[0038] By adopting the dynamic threshold strategy, the dynamic threshold is generated, so that the actual warehouse situation and environment state are more fitted, the judgment of the risk index is more accurate, the safety risk assessment more fitted actual is made, the accuracy of the safety risk assessment is further improved, and the harm and loss caused by the dangerous goods leakage are reduced.

[0039] The early warning method compares the risk index with the dynamic threshold, so that the each stage of the dangerous chemical leakage can be accurately judged and corresponding measures are made, and the harm caused by the dangerous chemical leakage is effectively reduced. BRIEF DESCRIPTION OF DRAWINGS

[0040] Figure 1 It is a flowchart of the safety risk assessment and early warning method of the present application;

[0041] Figure 2 It is a flowchart of the comparison between the risk index and the dynamic threshold in the safety risk assessment and early warning method of the present application. DETAILED DESCRIPTION

[0042] The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor are within the scope of protection of the present application.

[0043] In a feasible embodiment, referring to Figure 1 The safety risk assessment and early warning method includes the following steps:

[0044] Step 101: A plurality of vibration sensors are arranged to form a vibration sensor array, vibration signals of the warehouse structure are collected through the vibration sensor array, a temperature and humidity sensor is arranged to collect temperature and humidity changes of the warehouse, and temperature and humidity change data is output;

[0045] Step 102: The vibration signal is characterized and preprocessed based on a feature extraction module, and a feature vector is output;

[0046] Step 103: The risk calculation module acquires the feature vector, calculates and outputs a risk index, and generates a preliminary risk level;

[0047] Step 104: The threshold judgment module obtains the temperature and humidity change data, outputs the dynamic threshold, compares the risk index with the dynamic threshold, and judges the output final risk level;

[0048] Step 105: The early warning output module obtains the final risk level and outputs the early warning signal and disposal suggestion.

[0049] Specifically, by using a vibration sensor to detect dangerous chemicals, the development of structural cracks and small leaks can be accurately detected. The vibration characteristics generated by dangerous chemicals can be detected in advance before the temperature, pressure or gas concentration of dangerous chemicals changes significantly, so that appropriate actions can be taken to effectively prevent dangerous chemical leaks.

[0050] By using a dynamic threshold strategy, a dynamic threshold is generated, which can better fit the actual warehouse situation and environmental state, and the risk index judgment is more accurate, which can make a more realistic safety risk assessment, further improve the accuracy of safety risk assessment, and reduce the harm and loss caused by dangerous goods leakage;

[0051] The early warning method compares the risk index with the dynamic threshold to make accurate judgments and corresponding measures at each stage of dangerous chemical leakage, effectively reducing the harm caused by dangerous chemical leakage.

[0052] Specifically, please refer to Figure 1 As shown in FIG. 1, in step 101, the vibration sensor adopts a uniform distribution strategy to ensure complete coverage of the entire warehouse area. High-frequency vibration sensors are installed on the warehouse support columns to detect structural vibrations. Medium-frequency vibration sensors are arranged at the connection of the shelves to detect changes in the state of the goods. Low-frequency vibration sensors are set in the container storage area to capture abnormal signals such as leaks. The sampling frequency is set to 10kHz-20kHz, so that the full-band characteristics of the vibration signal can be completely captured.

[0053] Specifically, please refer to Figure 1 As shown in FIG. 2, in step 102, the feature extraction module extracts and preprocesses the vibration signal to output a feature vector, which includes:

[0054] The amplitude mean, root mean square value and waveform factor in the vibration signal are extracted. The amplitude mean reflects the average energy level of the vibration signal, the root mean square value reflects the effective amplitude of the vibration signal, and the waveform factor reflects the waveform characteristics of the vibration signal.

[0055] The preprocessing includes normalizing the amplitude mean value, the root mean square value and the waveform factor, so as to eliminate the dimensional difference of different characteristics, make the normalized data more accurate, and obtain characteristic vectors F1, F2 and F3, wherein F1 represents the normalized value of the amplitude mean value; F2 represents the normalized value of the root mean square value; and F3 represents the normalized value of the waveform factor.

[0056] Specifically, as shown in Figure 1 and Figure 2 , in step 103, the risk calculation module obtains the characteristic vectors, calculates the output risk index and generates the preliminary risk level, including:

[0057] The risk calculation module adopts a weighted fusion algorithm to fuse multiple characteristic vectors, and the risk index calculation formula is:

[0058] R = w1·F1 + w2·F2 + w3·F3 + a·log(1 + s 2 )

[0059] wherein R represents the risk index, ranging from 0 to 1; w1, w2 and w3 represent characteristic weights, the sizes of which are determined by the entropy weight method; a represents an adjustment coefficient for controlling the influence degree of volatility; s 2 represents the characteristic volatility variance, reflecting the system stability, and by adding the characteristic volatility variance, a more accurate risk index value can be obtained;

[0060] The output preliminary risk level includes: when R≤0.3, the preliminary risk level is level one; when 0.3

[0061] Specifically, as shown in Figure 1 and Figure 2 , in a feasible embodiment, the warehouse storing dangerous chemicals is sampled for one period, the vibration sensor measures that the amplitude mean value is 4.2 mm / s, the root mean square value is 5.8 mm / s, and the waveform factor is 0.35; by normalizing the amplitude mean value, the root mean square value and the waveform factor, F1=0.513, F2=0.470 and F3=0.545 are obtained; at the same time, the volatility variance is calculated according to the normalized characteristic values of the last 10 periods:

[0062] Period 1: F1=0.45, F2=0.40 and F3=0.50;

[0063] Period 2: F1=0.48, F2=0.42 and F3=0.52;

[0064] Cycle 3, F1=0.50, F2=0.45, F3=0.53;

[0065] Cycle 4, F1=0.52, F2=0.470, F3=0.54;

[0066] Cycle 5, F1=0.49, F2=0.43, F3=0.51;

[0067] Cycle 6, F1=0.51, F2=0.46, F3=0.55;

[0068] Cycle 7, F1=0.53, F2=0.48, F3=0.56;

[0069] Cycle 8, F1=0.50, F2=0.44, F3=0.52;

[0070] Cycle 9, F1=0.54, F2=0.49, F3=0.57;

[0071] Cycle 10, F1=0.513, F2=0.470, F3=0.545;

[0072] The calculated F1 variance is σ 2 =0.00076; F2 variance is σ 2 =0.00089; F3 variance is σ 2 =0.00054; the total fluctuation variance is σ 2 =0.00073; the weight of w1, w2, w3 is analyzed by entropy weight method, the weight proportion of amplitude mean value, root mean square value and waveform factor in risk index is calculated through historical data, the proportion P1 of each feature is 0.1019; P2 is 0.1042; P1 is 0.545 / (0.50+0.52+0.53+0.54+0.51+0.55+0.56+0.52+0.57+0.545)=0.1020; according to the formula E i =-k×∑(P i ×lnP i), wherein k = 1 / ln(n), n is the number of samples, i.e. 10, k = 1 / ln(10) = 0.434, then E1 = -0.434 x (0.1019 x ln(0.1019)) = 0.100; E2 = -0.434 x (0.1042 x ln(0.1042)) = 0.102; E3 = -0.434 x (0.1020 x ln(0.1020)) = 0.101;d i = 1 - E i , then according to the weight w i = d i / ∑d i , w1 = 0.100 / (0.100+0.102+0.101) = 0.33, w2 = 0.102 / (0.100+0.102+0.101) = 0.34, w3 = 0.101 / (0.100+0.102+0.101) = 0.33, set α = 1, substitute the formula can get R = 0.33 x 0.513 + 0.34 x 0.470 + 0.33 x 0.545 + 0.1 x log(1+0.00073) = 0.5089, then the preliminary risk level is judged to be two levels.

[0073] Specifically, please refer to Figure 1 and Figure 2 , in step 104, the threshold judgment module obtains the temperature and humidity change data, and outputs the dynamic threshold value, including:

[0074] The threshold judgment module adopts a dynamic threshold strategy and dynamically adjusts according to environmental factors. The dynamic threshold value includes:

[0075] T = T0·(1+k·ΔE)

[0076] Wherein, T represents the dynamic threshold value, T0 represents the basic threshold value, which is set according to historical data, i.e. the historical threshold value of this type of warehouse; k represents the environmental influence coefficient; ΔE represents the temperature and humidity change data.

[0077] Specifically, set the basic threshold value T0 = 0.6, the environmental influence coefficient K = 0.05, and the temperature and humidity change data ΔE = 0.02. Then the current dynamic threshold value is T = 0.6 x (1 + 0.05 x 0.02) = 0.6006, and the current dynamic threshold value is 0.6006.

[0078] Specifically, please refer to Figure 1 and Figure 2 , in step 104, by comparing the risk index with the dynamic threshold value, the final risk level is judged and output, including:

[0079] The risk index is compared with the dynamic threshold value to judge,

[0080] When R≥T, a higher level of early warning is triggered, and the final risk level output is one level higher than the preliminary risk level;

[0081] When R<T, the early warning level is maintained or reduced, and the final risk level output is the same level as the preliminary risk level or one level lower;

[0082] At this time, R and T are compared to determine that R=0.5089<T=0.6006, so the original early warning level is maintained, and the final risk level output is level two.

[0083] Specifically, please refer to Figure 1 As shown in FIG. 1, in step 105, the early warning output module obtains the final risk level and outputs the early warning signal and the disposal suggestion, including:

[0084] When the final risk level is level one, the early warning signal is blue, and it is suggested to perform routine inspection, and the early warning needs to be handled within 24 hours after being issued;

[0085] When the final risk level is level two, the early warning signal is green, and it is suggested to strengthen the inspection frequency, and the early warning needs to be handled within 12 hours after being issued;

[0086] When the final risk level is level three, the early warning signal is yellow, and it is suggested to perform on-site verification and cause analysis, and the early warning needs to be handled within 4 hours after being issued;

[0087] When the final risk level is level four, the early warning signal is orange, and it is suggested to send an emergency team to intervene and isolate some areas, and the early warning needs to be handled within 1 hour after being issued;

[0088] When the final risk level is level five, the early warning signal is red, and it is suggested to evacuate all personnel and start a comprehensive emergency response, and the early warning needs to be handled immediately after being issued;

[0089] By setting different final risk levels, different measures can be taken for different dangerous chemical accidents, and the waste of resources can be avoided while solving the risk of dangerous chemicals.

[0090] The embodiment of the application also provides a computer readable storage medium, which can be any available medium that can be stored by a computing device or a data storage device such as a data center containing one or more available media, which can be a magnetic medium (for example, a floppy disk, a hard disk, a magnetic tape), an optical medium (for example, a DVD), or a semiconductor medium (for example, a solid state disk), etc., the computer readable storage medium contains instructions for instructing the computing device to execute the foregoing time synchronization method.

[0091] The embodiment of the present application further provides a computer program product containing instructions, which can be software or program product containing instructions, capable of running on a computing device or being stored in any available medium, and when the computer program product runs on the computing device, the computing device is caused to execute the foregoing time synchronization method.

[0092] The preferred embodiments of the application disclosed above are only used to help explain the application. The preferred embodiments do not describe all the details and limit the application to the specific embodiments described. Obviously, many modifications and variations can be made in light of the contents of the specification. The specification selects and specifically describes these embodiments in order to better explain the principles and practical applications of the application, so that those skilled in the art can well understand and utilize the application.

Claims

1. A method for security risk assessment and early warning, characterized in that, The safety risk assessment and early warning method comprises the following steps: A plurality of vibration sensors are arranged to form a vibration sensor array, vibration signals of the warehouse structure are collected through the vibration sensor array, a temperature and humidity sensor is arranged to collect temperature and humidity changes of the warehouse, and temperature and humidity change data is output; The feature extraction module extracts and preprocesses the vibration signals to output a feature vector; The risk calculation module obtains the feature vector, calculates a risk index, and generates a preliminary risk level; The threshold judgment module obtains the temperature and humidity change data, outputs a dynamic threshold, compares the risk index with the dynamic threshold, and judges a final risk level; The early warning output module obtains the final risk level, outputs an early warning signal and a disposal suggestion.

2. The method of claim 1, wherein: The feature extraction module extracts and preprocesses the vibration signals to output a feature vector, which comprises: The amplitude mean value, root mean square value and waveform factor in the vibration signal are extracted; The preprocessing comprises normalizing the amplitude mean value, root mean square value and waveform factor to obtain feature vectors F1, F2 and F3, wherein F1 represents the normalized value of the amplitude mean value, F2 represents the normalized value of the root mean square value, and F3 represents the normalized value of the waveform factor.

3. The method of claim 2, wherein: The risk calculation module obtains the feature vector, calculates a risk index, and generates a preliminary risk level, which comprises: The risk calculation module adopts a weighted fusion algorithm to fuse a plurality of feature vectors, and the risk index calculation formula is: R = w1 · F1 + w2 · F2 + w3 · F3 + a · log(l + s 2 ) Wherein, R represents the risk index, w1, w2, w3 represent the characteristic weight, the size of the characteristic weight is determined by the entropy weight method; α represents the adjustment coefficient; σ 2 represents the characteristic fluctuation variance; The preliminary risk level comprises the following: when R is less than or equal to 0.3, the preliminary risk level is level one; when 0.3 is less than R and is less than or equal to 0.6, the preliminary risk level is level two; when 0.6 is less than R and is less than or equal to 0.8, the preliminary risk level is level three; when 0.8 is less than R and is less than or equal to 0.9, the preliminary risk level is level four; and when R is greater than 0.9, the preliminary risk level is level five.

4. The method of claim 3, wherein: The threshold judgment module obtains the temperature and humidity change data and outputs a dynamic threshold, which comprises: The threshold judgment module adopts a dynamic threshold strategy and dynamically adjusts according to environmental factors, and the dynamic threshold comprises: T = T0·(1+k·ΔE) Wherein, T represents the dynamic threshold, T0 represents the basic threshold value, which is set according to historical data; k represents the environmental influence coefficient; and ΔE represents the temperature and humidity change data.

5. The method of claim 4, wherein: The threshold judgment module obtains the temperature and humidity change data and outputs a dynamic threshold, which comprises: The risk index is compared with the dynamic threshold to judge, When R is greater than or equal to T, a higher level of early warning is triggered, and the final risk level is one level higher than the preliminary risk level; When R is less than T, the early warning level is maintained or reduced, and the final risk level is the same as or one level lower than the preliminary risk level.

6. The method of claim 5, wherein: The early warning output module obtains the final risk level, outputs an early warning signal and a disposal suggestion, which comprises: When the final risk level is level one, the early warning signal is blue, and the suggestion is to perform routine inspection, and the early warning needs to be processed within 24 hours after being issued; When the final risk level is level two, the early warning signal is green, and the suggestion is to strengthen the inspection frequency, and the early warning needs to be processed within 12 hours after being issued; When the final risk level is three, the warning signal is yellow, suggesting on-site verification, cause analysis, and the warning needs to be handled within 4 hours after being issued; When the final risk level is four, the warning signal is orange, suggesting that an emergency team should be sent to intervene, and some areas should be isolated, and the warning needs to be handled within 1 hour after being issued; When the final risk level is five, the warning signal is red, suggesting that all personnel should be evacuated, and a comprehensive emergency response should be started, and the warning needs to be handled immediately after being issued.

7. A computer-readable storage medium, characterized in that: The computer readable storage medium stores a computer program, and the computer program is executed by a computer or a processor to implement the method in any one of claims 1-6.

8. A computer program product, characterised in that: The computer program product comprises a computer program, and when the computer program is executed by a computer or a processor, the computer or the processor executes the method in any one of claims 1-6.

Images