Dangerous goods transportation leakage accident consequence risk grading evaluation method based on accident range area simulation

Abstract

The invention belongs to the field of accident risk grading, and relates to a dangerous goods transportation leakage accident consequence risk grading evaluation method based on accident range area simulation, and the method comprises the steps: obtaining data information after a traffic accident; inputting the data information into the accident range area simulation model to obtain an accident consequence index; performing clustering analysis on the accident consequence indexes by adopting a K-means algorithm to obtain a dangerous goods transportation leakage accident consequence risk grading result; evaluating an accident consequence risk grading result; determining the severity and the influence range of the accident consequence according to the evaluation result; constructing a three-dimensional comprehensive risk evaluation matrix, querying three grading matrixes, obtaining corresponding risk grades, and performing comprehensive risk grade evaluation; carrying out fuzzy clustering analysis to obtain a transportation road section risk grade, and defining a road section risk value as a product of a road section grade and a road section length; calculating the risk value of the transportation route by using the thought of a triangular fuzzy number; screening the transportation routes according to the risk values to obtain an optimal transportation route of the dangerous goods; the risk grading evaluation model can be continuously optimized and perfected by establishing a feedback mechanism, and the accuracy and reliability of an evaluation result are improved.

Description

Technical Field

[0001] This invention belongs to the field of accident risk classification, specifically relating to a method for risk classification and evaluation of the consequences of dangerous goods transportation leakage accidents based on accident range area simulation. Background Technology

[0002] Against the backdrop of economic globalization and the booming development of international trade, the transportation of hazardous materials, as a key component of the modern logistics system, is becoming increasingly important. Nevertheless, the high safety risks and complex legal liabilities associated with hazardous materials transportation necessitate a more comprehensive management system within the industry to ensure the safety and reliability of transportation. This is particularly true for specialized hazardous materials transportation assessment centers, where the effectiveness of process management directly impacts the safety of the entire transportation chain. With rapid economic growth and accelerating industrialization, the volume of hazardous materials transported has increased year after year, accompanied by a surge in hazardous materials road transport accidents, escalating to alarming levels of large and even catastrophic traffic disasters. In current practical applications, outdated manual operation methods cannot meet the standards of high efficiency and safety, easily leading to low operational efficiency and hidden risks in risk control. Data shows that my country experiences over a hundred road hazardous materials transportation accidents annually, causing significant damage and potential threats to the lives and assets of the nation and its people.

[0003] Existing research employs various indicators for accident consequences, primarily categorizing them into three main types: exposed population, environmental damage, and economic losses. However, based on the people-centered development philosophy, human life safety is more crucial than the other two categories. Therefore, how to conduct risk classification and assessment of the consequences of hazardous materials transportation leaks is a pressing issue that needs to be addressed. Summary of the Invention

[0004] To address the problems existing in the prior art, this invention proposes a risk classification and evaluation method for hazardous materials transportation leakage accidents based on accident area simulation. This method includes: acquiring post-traffic accident data; inputting the data into an accident area simulation model to obtain accident consequence indicators; using the K-means algorithm to perform cluster analysis on the accident consequence indicators to obtain risk classification results for hazardous materials transportation leakage accidents; evaluating the risk classification results; determining the severity and scope of the accident consequences based on the evaluation results; constructing a three-dimensional comprehensive risk evaluation matrix based on accident type, transport volume, and population density; querying the three classification matrices to obtain the corresponding risk levels and performing a comprehensive risk level assessment; using fuzzy clustering analysis based on the risk level assessment results to obtain the risk classification of transport segments, introducing a distance factor, and defining the segment risk value as the product of the segment level and segment length; calculating the risk value of the transport route using the concept of triangular fuzzy numbers; and filtering transport routes based on the risk values ​​to obtain the optimal transport route for hazardous materials.

[0005] The process of simulating the accident area using an accident scope simulation model includes: calculating the number of people in the danger zone caused by the hazardous materials leak; calculating the mortality rate of people in the danger zone based on the risk factors of the hazardous materials leak; simulating the accident scope based on the mortality rate; and calculating accident consequence indicators using an accident consequence calculation model.

[0006] The beneficial effects of this invention are:

[0007] 1. Improve the scientific nature and accuracy of risk assessment: This invention, through an accident area simulation model and K-means cluster analysis, can accurately classify the risks of hazardous materials transportation leakage accidents, providing a reliable basis for risk management and emergency decision-making.

[0008] 2. Comprehensive consideration of multiple factors: In the evaluation process, this invention comprehensively considers multiple factors such as accident type, exposed population, mortality rate, injury rate and environmental impact, making the risk classification results more comprehensive and accurate.

[0009] 3. Provide decision support: The risk classification results of this invention can be applied to route planning, emergency response plan formulation, and risk management decision support, helping to select transportation routes with lower risks, formulate corresponding emergency response measures, and improve the efficiency and effectiveness of emergency response.

[0010] 4. Possibility of continuous improvement: By establishing a feedback mechanism and introducing advanced technologies, this invention can continuously optimize and improve the risk grading and evaluation model, thereby enhancing the accuracy and reliability of the evaluation results.

[0011] 5. Reduce accident risks and potential losses: This invention helps dangerous goods transportation companies develop reasonable risk management strategies, such as adjusting transportation plans, strengthening safety precautions, and improving personnel training, thereby reducing accident risks and potential losses and protecting the lives and property of the people and social stability. Attached Figure Description

[0012] Figure 1 Flowchart for risk classification and assessment of hazardous materials transportation leakage accidents;

[0013] Figure 2 Architecture diagram of a risk classification and assessment system for hazardous materials transportation leakage accidents;

[0014] Figure 3 Design diagram for a risk classification and assessment module for hazardous materials transportation leakage accidents;

[0015] Figure 4 Map showing the extent of a toxic gas spread accident involving a 10t liquid ammonia leak;

[0016] Figure 5 A graph showing the relationship between toxic gas concentration and time at various locations downwind of the liquid ammonia leak.

[0017] Figure 6 Diagram showing the extent of a 1-ton liquid ammonia leak and jet fire accident;

[0018] Figure 7 This refers to the BLEVE accident involving a 10t liquid ammonia leak.

[0019] Figure 8 This refers to the extent of a toxic gas diffusion accident caused by a 1-ton liquid ammonia leak.

[0020] Figure 9 A graph showing the concentration of toxic gases at various locations downwind;

[0021] Figure 10 This refers to the area affected by a 10t liquid ammonia leak and subsequent jet fire accident.

[0022] Figure 11 This is a graph showing the heat flux variation at different radii.

[0023] Figure 12 Diagram showing the extent of a 1-ton liquid ammonia leak at BLEVE.

[0024] Figure 13 Accident range diagrams for different transport volumes;

[0025] Figure 14 This is a flowchart of the K-means clustering algorithm. Detailed Implementation

[0026] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0027] A risk classification and assessment method for the consequences of hazardous materials transportation leaks based on accident area simulation, such as... Figures 1-3The process includes: acquiring data information after a traffic accident; inputting the data information into an accident area simulation model to obtain accident consequence indicators; using the K-means algorithm to perform cluster analysis on the accident consequence indicators to obtain the risk classification results of hazardous materials transportation leakage accidents; evaluating the risk classification results of accident consequences; determining the severity and scope of impact of the accident consequences based on the evaluation results; constructing a three-dimensional comprehensive risk assessment matrix based on accident type, transport volume, and population density; querying the three classification matrices to obtain the corresponding risk levels and conducting a comprehensive risk level assessment; using fuzzy cluster analysis based on the risk level assessment results to obtain the risk classification of transport segments, introducing a distance factor, and defining the segment risk value as the product of the segment level and the segment length; calculating the risk value of the transport route using the concept of triangular fuzzy numbers; and filtering the transport routes based on the risk values ​​to obtain the optimal transport route for hazardous materials.

[0028] The process of simulating the accident area using an accident scope simulation model includes: calculating the number of people in the danger zone caused by the hazardous materials leak; calculating the mortality rate of people in the danger zone based on the risk factors of the hazardous materials leak; simulating the accident scope based on the mortality rate; and calculating accident consequence indicators using an accident consequence calculation model.

[0029] The number of fatalities in a hazardous materials road transport accident is a crucial indicator for determining the consequences of a leak, primarily related to the exposed population and the probability of death within the accident area. The exposed population mainly includes two categories: those on the road and those off the road. The on-road population is mainly related to vehicular traffic, non-vehicle traffic, pedestrians, and the length of the affected road section; they are the direct victims of the accident. The off-road population is related to population density within the accident area and the extent of the accident's impact, generally exhibiting a banded distribution, and can be determined by the population density of the affected areas on both sides of the road.

[0030] Obtaining pedestrian flow data at the time of the accident presents a challenge. To simplify the model, this paper assumes that the on-street population is only related to motor vehicles. Therefore, this paper constructs an on-street population density calculation model based on the average daily traffic volume of the accident section. The speed-density relationship model is shown in Table 1.

[0031] Table 1 Velocity-Density Model

[0032]

[0033] Where u is the speed of the motor vehicle, u f Let k be the maximum speed of the motor vehicle and k be the traffic density. j For the blocking density, u m For the optimal vehicle speed, k m The optimal density.

[0034] The relationship between vehicle speed and traffic density is fundamental to the study of traffic flow volume-speed-density. Models for the speed-density relationship are mainly divided into two categories: single-piece functions and multi-piece functions. Typical single-piece linear models include the speed-density linear model, a logarithmic model suitable for congested traffic, and another logarithmic model for low traffic density conditions, as well as family of curve models. Representative multi-piece function models include the Edie model, the Underwood modified model, and the Dick model.

[0035] The on-street population is defined as the product of on-street population density and road segment length. On-street population density is linearly related to vehicle density, thus the on-street population and vehicle density are closely correlated. This paper assumes that the traffic flow is in a steady state, where the higher the traffic density, the more significant the interaction between vehicles, and the greater the decrease in vehicle speed; traffic density and vehicle speed are linearly related. Based on the Greenshields linear model, the model is transformed into a relationship between traffic volume and vehicle density, expressed as:

[0036] q = u f (kk 2 / k j )

[0037] Where q represents the traffic flow at the accident site.

[0038] After function transformation, the annual average daily traffic volume is used instead of traffic flow, and the annual average daily traffic volume is converted into the hourly average traffic volume, resulting in the following relationship between traffic density and traffic volume:

[0039]

[0040] Due to the high degree of randomness in traffic flow arrival, this paper assumes uniform vehicle arrival to describe the relationship between on-street population and traffic flow. That is, given a fixed traffic volume on a road segment within a certain time period, the vehicle density on that segment remains constant during that time period. Based on this assumption, on-street population is defined as the product of on-street population density and road segment length. By introducing the average vehicle occupancy factor, directional non-uniformity coefficient, and accident segment length, a functional relationship between on-street population and traffic volume is obtained.

[0041] P on =δL on k / K D

[0042] Among them, P on Let L be the roadside population, δ be the average number of people per vehicle, and L be the average number of people per vehicle. on Let K be the length of the road segment affected by the accident, and k be the relationship between traffic density and traffic volume. D k is the coefficient of directional non-uniformity. jFor congestion density, AADT is the annual average daily traffic volume, u f This refers to the maximum speed of a motor vehicle.

[0043] In this embodiment, the underground population is mainly represented by the product of the population density within the accident area and the accident range. This paper uses the average population density as the population density within the accident-affected area. Due to varying degrees of injury, the underground population can be divided into indoor and outdoor populations. At the same distance, the degree of injury to indoor populations is significantly reduced due to the obstruction of buildings. The impact on outdoor populations is taken as 100%, and the impact on indoor populations as 20%. The underground population is:

[0044] P off =p off S(0.8λ+0.2)

[0045] Among them, P off For the population below the road, p off Let S be the underground population density, S be the accident impact range, and λ be the proportion of the population outdoors.

[0046] In this embodiment, the mortality rate of the exposed population is directly related to the type of injury and the type of accident area. This paper considers three types of injury: toxic gas diffusion, BLEVE, and flash fire, and uses the probability function method to calculate the probability of mortality. The probability function method is a method that describes the consequences of a leak by describing the impact of a person's exposure to a certain type of injury over a certain period of time.

[0047] When gaseous chemicals leak during transportation, they gradually diffuse in the environment due to wind and atmospheric turbulence, which is a major cause of secondary accidents. Although the Gaussian model's simulation accuracy is not high, it is widely used in simulating hazardous material accidents because it was developed early on, has abundant reference data, and is applicable to both continuous and transient leaks. This model is easy to understand, computationally inexpensive, and its results closely approximate experimental values. Depending on the type of leak source and the duration of the leak, the Gaussian model can be divided into two cases: plume diffusion and smoke plume diffusion.

[0048] In instantaneous leaks and partially continuous source leaks, or in light winds (<1 m·s) -1 Under certain conditions, the Gaussian plume model is often used for calculations. The formula for calculating point source concentration is:

[0049]

[0050] When the leak is a continuous source, or the leak time is greater than or equal to the diffusion time, a Gaussian plume model is often used for calculation. The formula for calculating point source concentration is:

[0051]

[0052] Where c(x,y,z,T) is the concentration of toxic gas at a point in space, and Q c Let σ be the leakage source intensity, u be the average wind speed at the height of the leakage point, T be the time it takes for the plume to travel from the leakage point to that location, and σ be the average wind speed at the leakage source. x , σy, σ z , respectively, are the gas diffusion coefficients in the downwind, crosswind, and vertical directions, and H is the effective height of the leak point.

[0053] This paper uses the Gaussian plume model for accident simulation. Regarding the consequences of toxic gas leaks, a probability function method for calculating the consequences of toxic gas leaks and the conversion relationship between the probability of death and the mortality rate are presented. The expression for the probability function is:

[0054] Pr=A+Bln(C n t)

[0055] Where A, B, and n represent the properties of the toxic substance, C represents the concentration of the toxic substance, and t represents the duration of human exposure to the toxic substance.

[0056] In this embodiment, the fire primarily harms people and objects within its affected area through heat and smoke, causing environmental pollution. The heat generated by the fire is mainly transferred in three ways: thermal radiation, thermal conduction, and convection, with thermal radiation being the most harmful.

[0057] Injuries caused by jet fire can be classified into three different types: first-degree burns, burns, and death. This article only considers the probability of death:

[0058] Pr = –36.38 + 2.56ln D

[0059] Where Pr is the probability of death caused by a fire or explosion, and D is the amount of heat radiation.

[0060] In this embodiment, the explosion and leakage of substances is an accident that involves drastic physical and chemical changes. It releases a large amount of energy in a short period of time, causing great harm to people and buildings. The main hazards are thermal radiation and shock wave overpressure. The thermal radiation damage is calculated in the same way as the jet fire, so this section introduces the shock wave damage type.

[0061] Deaths caused by BLEVE shockwaves include three types: death from lung injury, death from head injury, and death from whole-body impact. Current research indicates that, assuming the same probability of death, calculations show that the radius of death from lung injury is smaller than that from whole-body impact, which in turn is smaller than the radius of death from head impact, with the head having the largest radius of death. Therefore, the probability of death from head injury is chosen as a representation of the probability of death from a vapor cloud explosion, and its expression is:

[0062] Pr = 5.0–8.49ln S s

[0063] Among them, S S This represents the shock value caused by the explosion.

[0064] In this embodiment, the simulation of the consequences of a hazardous materials road transport accident aims to predict the potential impact range and threat level after the accident occurs. Common methods include formula calculation and software simulation. Compared to formula calculation, software simulation has advantages such as speed, convenience, flexibility, and the ability to directly display threat area diagrams, and is therefore widely used in various operating conditions. Currently, commonly used engineering software for hazardous materials accident simulation includes AFTOX, DEGIDAS, SLAB, and ALOHA.

[0065] Using ALOHA software for emergency accident simulation can effectively improve the speed and efficiency of emergency response. It can also be used to assess potential risks in chemical enterprises, providing a reference for the development of environmental risk management plans and emergency response plans for sudden environmental pollution accidents. In the event of an accident, software simulation can quickly and effectively determine the approximate leakage and diffusion range, and display relevant information such as diffusion concentration and impact area graphically. Rescue personnel can refer to the hazardous material leakage and diffusion range to organize the evacuation of trapped personnel to a safe area.

[0066] The software displays data including indicators such as the properties of hazardous materials and tank conditions. Specific operational steps include setting the geographical location, selecting the building type, choosing the date and time, setting external atmospheric conditions, setting the leakage situation and type, selecting calculation options, and display options. After the parameters are entered, the ALOHA software will summarize the input parameters in a document called "Text Summary," which can be used to check the accuracy of the input information and make modifications as needed. After completing the data input, you can select to display the diffusion area, which can display different levels of distribution areas and their shapes and sizes. You can also specify a specific location to observe the leakage concentration, and by selecting "Threat At point" under "Display" and choosing a relative coordinate system, you can see the leakage concentration curve for that point.

[0067] Example 1

[0068] The accident background is set as follows: a chemical plant, L, needs to deliver a batch of liquid ammonia, totaling 10t, in a tanker truck with a length of 8.963m and a diameter of 2.372m. Other conditions are shown in Table 3.

[0069] Table 3. Accident Simulation Assumptions

[0070]

[0071]

[0072] By importing the above parameters into ALOHA to perform accident simulation calculations on a 10t liquid ammonia leak, the accident range area for different accident types can be obtained.

[0073] Liquid ammonia has a strong, pungent odor, is highly volatile, and toxic. A leak can easily cause poisoning, asphyxiation, or even death. Acute Exposure Guidance Levels (AEGLs) are guidance levels developed by the National Advisory Committee on Acute Exposure Guidance Levels for Hazardous Substances in the United States. This paper selects AEGLs as the standard for determining the hazardous area extent of a leak. Table 4 shows the AEGLs standards.

[0074]

[0075] Importing the above data into ALOHA for accident simulation yields the extent of toxic gas diffusion, such as... Figure 4 As shown. By Figure 5 As can be seen, the diffusion distances of liquid ammonia AEGL-1, AEGL-2, and AEGL-3 are shown in Table 5.

[0076] Table 5. Distance of spread in different accident areas

[0077]

[0078] Figure 5 The duration of toxic gas concentration at different points within AEGL-1 is shown. The duration of toxic gas concentrations greater than 1100 ppm is generally maintained at 15 minutes. Therefore, the duration t in the calculation of the probability of death due to toxic gas diffusion is set to 15 minutes.

[0079] Calculations show that the mortality rate is 0 at a = 0.245 km and b = 0.021 km, where the toxic gas concentration is 1775 ppm. This means that when the diffusion range exceeds these distances, deaths will not occur if the toxic gas concentration is less than 1715 ppm. Therefore, the death zone at the toxic gas concentration can be determined to be 0.008 km². 2 .analyze Figure 5 It can be seen that the indoor concentration of toxic gases is less than 1715 ppm, indicating that no one will die in the event of a toxic gas accident.

[0080] The main type of injury from jet fire is thermal radiation, and different thermal radiation intensity thresholds affect human life safety. Figure 4 The diagram shows the accident extent of a jet fire caused by a 10t liquid ammonia leak. As can be seen from the figure, the heat flux of the jet fire is >10kW·m³. -2 5~10kW·m -2 2~5kW·m -2The diffusion radii were 10m, 17m, and 30m, respectively. Simulations of the heat flux at each point showed that when the diffusion distance was 6m and the heat flux was 13kW·m³, the diffusion radii were optimal. -2 The mortality rate reached its minimum, with a dead area of ​​113m². 2 .

[0081] The hazards of BLEVE include blast shockwaves and thermal radiation. Due to limitations of the ALOHA software, this paper uses thermal radiation as the primary hazard to calculate the mortality rate. Figure 4 The figure shows the area affected by a BLEVE accident caused by a 10t liquid ammonia leak. The graph indicates that the heat flux of the BLEVE is greater than 10kW·m³. -2 5~10kW·m -2 2~5kW·m -2 The diffusion radii were 142 m, 206 m, and 326 m, respectively. Simulations of the heat flux at each point showed that when the diffusion distance was 225 m and the heat flux was 4.2 kW·m³, the diffusion was most effective. -2 The mortality rate reached its minimum at that time, with a dead zone of 0.159 km². 2 .

[0082] In this embodiment, the accident evolution probabilities of three types of accidents—BLEVE, jet fire, and toxic gas diffusion—are introduced, and the cumulative number of deaths from these three accident types is used as an indicator of accident consequences. The accident consequences indicator is as follows:

[0083]

[0084] Among them, C i–j f represents the death rate index for road segment ij. K The mortality rate for Class K accidents, AADT i–j Let be the average daily traffic volume of road segment ij.

[0085] Example 2

[0086] Taking liquid ammonia transportation as an example, if a leak occurs during transportation, event tree analysis suggests potential consequences such as toxic gas spread, fire, and explosion, threatening life and property in the surrounding area. Key indicators include the number of exposed individuals, environmentally sensitive areas, and economic losses. Among these three, the number of deaths is relatively easy to calculate and ranks first in importance in the accident's consequences, directly reflecting the severity of the incident. Therefore, this chapter constructs a grading model of accident consequences using the number of deaths as an indicator.

[0087] Accident severity is classified into four levels, with the classification criteria determined based on the number of fatalities. Specific classification principles are shown in Table 6. Table 6: Accident Consequence Level Classification

[0088]

[0089] To facilitate model construction, this paper defines the number of deaths as the product of population density, mortality rate, and accident area, with the mortality rate and accident area closely related to the amount of hazardous material leakage. Therefore, this paper simulates toxic gas diffusion, jet fire, and BLEVE events in liquid chlorine transport leaks of different volumes, converting the accident level into population density and transport volume levels.

[0090]

[0091] Among them, C P Let P be the number of deaths, ρ be the average population density of the accident site, and P be the number of deaths. K Let f be the probability of occurrence of accident type K. K Let S be the mortality rate for accident type K. K Let K be the accident type and the accident range area.

[0092] Example 3

[0093] The maximum transport weight of liquid ammonia is 20t. Therefore, this paper simulates a leakage accident involving 1-20t liquid ammonia transport as a sample, and uses ALOHA simulation to calculate the mortality rate and accident area. Taking a toxic gas diffusion accident caused by a 1t liquid ammonia leak as an example, the simulation experiment was conducted using ALOHA software, and the results are as follows: Figure 5 The extent of the toxic gas diffusion is shown.

[0094] Depend on Figure 5 It can be seen that the maximum toxic gas diffusion distances along the windward direction of 1 ton of liquid ammonia leaking into AEGL-1, AEGL-2, and AEGL-3 are 3.200 km, 1.600 km, and 0.761 km, respectively, and the maximum longitudinal diffusion distances are 0.110 km, 0.055 km, and 0.026 km, respectively. Therefore, the areas of AEGL-1, AEGL-2, and AEGL-3 can be calculated to be 0.414 km². 2 0.107km 2 0.031km 2 .

[0095] Figure 6 The curves showing the relationship between toxic gas concentration and time at downwind locations of 0.001 km, 0.400 km, 0.600 km, and 0.761 km within the AEGL-1 area are displayed. The comparison shows that the farther the point is from the accident site, the lower the peak toxic gas concentration and the longer the exposure time for the exposed population. Considering that the mortality rate is related to both toxic gas concentration and exposure time, this chapter calculates the maximum mortality rate at the downwind location for different areas.

[0096] Example 4

[0097] Taking a 1-ton liquid ammonia leak and subsequent jet fire accident as an example, the simulation experiment was conducted using ALOHA software, and the results were as follows: Figure 8 The range of the jet fire is shown.

[0098] Depend on Figure 7 It can be seen that the heat flux of a jet fire accident caused by a 1t liquid ammonia leak is greater than 10.0 kW·m. -2 Greater than 5.0 kW·m -2 Greater than 2.0 kW·m -2 The maximum fire radii are 10m, 17m, and 29m respectively, from which the areas of the three damage zones can be calculated to be 314.00m². 2 593.46m 2 1733.28m 2 .

[0099] Figure 11 The graph shows the heat flux variations at radii of 0.01m, 3.3m, 6.6m, and 10m. As shown in the graph, a jet fire caused by a 1t liquid ammonia leak lasted 1.5 minutes. The maximum heat flux initially increased and then gradually decreased with increasing radius. Simulations of the accident area for different volumes of liquid ammonia revealed that the accident area and heat flux remained unchanged, with only the accident duration differing; the accident area remained consistently 113m². 2 The probability of death is 0.5.

[0100] Example 5

[0101] Using thermal radiation to calculate mortality rates, taking a 1-ton liquid ammonia leak and subsequent BLEVE event as an example, and importing the data into ALOHA software for simulation experiments, the following results can be obtained: Figure 9 The range of the jet fire is shown.

[0102] Depend on Figure 12 It can be seen that the heat flux of a 1t liquid ammonia leak causing a BLEVE accident is greater than 10.0 kW·m³. -2 Greater than 5.0 kW·m -2 Greater than 2.0 kW·m -2 The maximum fire radii are 68m, 98m, and 156m, respectively, from which the areas of the three damage zones can be calculated to be 14519.36m². 2 15637.2m 2 46258.48m 2 .

[0103] Figure 10 This represents the accident area where BLEVE occurs under different transport volumes. The mortality rate is calculated based on the maximum heat flux within each accident area, yielding the mortality rate and accident area under different transport volumes.

[0104] Example 6

[0105] To classify the consequences of accidents, directly clustering traffic volume and population density only indicates the classification, not the classification. To address this, this paper calculates the death toll classification based on traffic volume and population density, transforming two-dimensional clustering into one-dimensional clustering. The K-means algorithm is used to cluster the death toll under different traffic volumes and population densities. The accident consequence classification is clearly determined based on the number of deaths in each sample group. Then, the corresponding traffic volume and population density are found for each sample group, allowing for classification based on traffic volume and population density. The results are as follows: Figure 13 As shown.

[0106] like Figure 14 As shown, the mathematical idea of ​​the K-means clustering method can be summarized as follows: for a given dataset X = {X_n} containing n data points... i Given |i=1,2,3,...,n} and the values ​​K of the data cluster sets to be generated, the K-means clustering method divides the dataset into K partitions Y={Y j |j=1,2,...,K}, where each subset corresponds to a cluster center value C. For example... Figure 11 As shown, the specific steps are as follows.

[0107] Step 1: Standardize the units of measurement. Since the units of measurement for each element are different, to improve the computational efficiency of the algorithm and the accuracy of clustering, the indicators are normalized using the range method. The calculation method is as follows:

[0108]

[0109] Where: x ij y ij These represent the values ​​of the j-th indicator of the i-th data point before and after processing, respectively.

[0110] Step 2: Select the value K for the cluster set.

[0111] Step 3: Select the initial cluster center C. Randomly select a data point as C1, and select the next initial cluster center based on the principle of maximizing the P-value, until the number of initial cluster centers reaches K. The formula for calculating P is:

[0112]

[0113] Step 4: Calculate cluster centers. Calculate the Euclidean distance D(X) between the sample and the initial cluster centers. i C j Cluster centers are determined based on minimizing the Euclidean distance. The Euclidean distance is:

[0114]

[0115] Step 5: Update cluster centers. Calculate the arithmetic mean of each dataset and use the arithmetic mean to represent the new cluster centers.

[0116] Step 6: Calculate the total sum of squared errors. If the elements of the new clusters have not changed, the iteration ends; if they have changed, return to Step 3.

[0117] Cluster analysis was performed on population density. After four iterations, the elements within each group no longer changed, and the iteration ended. Population density was divided into five levels: 0–156 people / km². -2 Level 1, 156–1176 people / km -2 Level 2, 1176–4383 people·km -2 Level 3, 4383–18992 people·km -2 Level 4, 18,992–89,340 people / km -2 It is level 5.

[0118] Example 6

[0119] A three-dimensional comprehensive risk assessment matrix is ​​constructed based on accident type, traffic volume, and population density. Taking accident type level 2, population level 2, and traffic volume level 4 as an example, querying the three level matrices reveals that the corresponding risk levels are level 2, level 3, and level 3, respectively, and the comprehensive risk level corresponds to level 3.

[0120] Traditional methods for calculating transportation route risk values ​​mostly rely on summation. While simple and easy to understand, this method is susceptible to extreme values. If a route selection includes a segment with an extreme risk value, the cumulative risk value may be too high, affecting the overall route selection result. Triangular fuzzy numbers, due to their fuzziness and uncertainty, can effectively solve this problem. They represent an interval concept, given the upper and lower limits of the probability interval and the middle limit of the most probable value. This invention utilizes the concept of triangular fuzzy numbers to calculate the risk value of transportation routes. Let the fuzzy set of the triangular fuzzy numbers be B = (a, b, c), then its membership function is U. B(x) The calculation formula is as follows:

[0121]

[0122] Using triangular fuzzy numbers to calculate road segment risk values ​​can avoid the situation where high-risk road segments reduce the safety of the route risk value, which occurs when using the geometric mean method. The specific calculation steps are as follows:

[0123] Step 1: Calculate the risk value of the transportation segment. The risk level of the transportation segment is obtained through fuzzy clustering analysis. A distance factor is introduced, and the risk value of the segment is defined as the product of the segment level and the segment length.

[0124] r i-j =G i-j ·l i-j

[0125] Among them: G i-j Risk level of road segment ij; l i-j Let be the length of road segment ij.

[0126] Step 2: Construct the triangular fuzzy number. The minimum, maximum, and arithmetic mean of the risk values ​​of road segments along the transportation route are used to represent the average risk level of different road segments. The triangular fuzzy number is as follows:

[0127] B i =(a i ,b i ,c i ),and

[0128] Among them: B i Let be the triangular fuzzy number of transportation route i.

[0129] Step 3: Deblurring. Deblurring is performed using the center-of-gravity method. Its advantage lies in fully preserving the objectivity of the road segment risk value. After completion, the transportation route risk value can be obtained, and its calculation formula is as follows:

[0130]

[0131] Based on the comprehensive risk assessment model, the risk level and risk value of each road segment can be obtained, as shown in Tables 5-20 and 5-21. The risk values ​​of the road segments were calculated using triangular fuzzy numbers, and the risk values ​​were 3.82 and 3.04, respectively. From the perspective of accident type analysis, Route 2 has a lower accident type level, with an average value of 1.375, while Route 2's average accident type level is 2, which is higher than Route 1. From the perspective of accident consequences, with the same transport volume, comparing population density levels reveals that Route 1, despite its shorter route length, has extremely high population density around it, while Route 2, with some sections traversing urban areas, has lower population density. In summary, transport route 2 is superior to transport route 1.

[0132] The above-described embodiments further illustrate the purpose, technical solution, and advantages of the present invention. It should be understood that the above-described embodiments are merely preferred embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made to the present invention within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for risk classification and evaluation of the consequences of hazardous materials transportation leakage accidents based on accident area simulation, characterized in that, include: Data information after a traffic accident is acquired; this data is input into an accident area simulation model to obtain accident consequence indicators; the K-means algorithm is used to perform cluster analysis on the accident consequence indicators to obtain the risk classification results of hazardous materials transportation leakage accidents; the risk classification results are evaluated; the severity and scope of the accident consequences are determined based on the evaluation results; a three-dimensional comprehensive risk assessment matrix is ​​constructed based on accident type, transport volume, and population density; the three classification matrices are queried to obtain the corresponding risk levels and a comprehensive risk level assessment is performed; based on the risk level assessment results, fuzzy cluster analysis is used to obtain the risk classification of transport segments, introducing a distance factor, and defining the segment risk value as the product of the segment level and the segment length; the risk value of the transport route is calculated using the concept of triangular fuzzy numbers. Based on the risk value, the transportation routes are screened to obtain the optimal transportation routes for dangerous goods; The process of simulating the accident area using an accident scope simulation model includes: calculating the number of people in the danger zone caused by the hazardous materials leak; calculating the mortality rate of people in the danger zone based on the risk factors of the hazardous materials leak; simulating the accident scope based on the mortality rate; and calculating accident consequence indicators using an accident consequence calculation model.

2. The method for risk classification and evaluation of the consequences of hazardous materials transportation leakage accidents based on accident area simulation as described in claim 1, characterized in that, The calculation of the number of people in a hazardous area due to a hazardous materials leak includes: the number of people in a hazardous area includes the on-street population and the off-street population; the on-street population is calculated using a traffic flow-based method, that is, a traffic volume-vehicle density model is constructed, and an on-street population-traffic volume function is constructed based on the traffic volume-vehicle density model; the on-street population is calculated based on the on-street population-traffic volume function; the off-street population is calculated using building influence factors.

3. The method for risk classification and evaluation of the consequences of hazardous materials transportation leakage accidents based on accident area simulation as described in claim 2, characterized in that, The formula for calculating the population on the street is: P on =δL on k / K D Among them, P on Let L be the roadside population, δ be the average number of people per vehicle, and L be the average number of people per vehicle. on Let K be the length of the road segment affected by the accident, and k be the relationship between traffic density and traffic volume. D k is the coefficient of directional non-uniformity. j For congestion density, AADT is the annual average daily traffic volume, u f This refers to the maximum speed of a motor vehicle.

4. The method for risk classification and evaluation of the consequences of hazardous materials transportation leakage accidents based on accident area simulation as described in claim 2, characterized in that, The formula for calculating the population under the road is: P off =p off S(0.8λ+0.2) Among them, P off For the population below the road, p off Let S be the underground population density, S be the accident impact range, and λ be the proportion of the population outdoors.

5. The method for risk classification and evaluation of the consequences of hazardous materials transportation leakage accidents based on accident area simulation as described in claim 2, characterized in that, Calculating the mortality rate of people in hazardous locations includes factors contributing to death, such as the spread of toxic gases, jet fire, and BLEVE shockwaves.

6. The method for risk classification and evaluation of the consequences of hazardous materials transportation leakage accidents based on accident area simulation as described in claim 5, characterized in that, The calculation of mortality rate caused by toxic gas diffusion includes: determining the type of leak source and the duration of the leak; if it is an instantaneous leak or a partially continuous leak, the point source concentration is calculated using a plume diffusion model; if the leak type is a continuous source, or the leak time is greater than or equal to the diffusion time, the point source concentration is calculated using a plume model; and the mortality rate of personnel is calculated based on the point source concentration.

7. The method for risk classification and evaluation of the consequences of hazardous materials transportation leakage accidents based on accident area simulation as described in claim 6, characterized in that, The point source concentration calculated by the smoke plume diffusion model is: The point source concentration calculated by the plume model is: Where c(x,y,z,T) is the concentration of toxic gas at a point in space, and Q c Let σ be the leakage source intensity, u be the average wind speed at the height of the leakage point, T be the time it takes for the plume to travel from the leakage point to that location, and σ be the average wind speed at the leakage source. x σ y σ z , respectively, are the gas diffusion coefficients in the downwind, crosswind, and vertical directions, and H is the effective height of the leak point.

8. The method for risk classification and evaluation of the consequences of hazardous materials transportation leakage accidents based on accident area simulation as described in claim 1, characterized in that, The accident consequence calculation model calculates accident consequence indicators including: Among them, C i–j f represents the death rate index for road segment ij. K The mortality rate for Class K accidents, AADT i–j Let be the average daily traffic volume of road segment ij.

9. The method for risk classification and evaluation of the consequences of hazardous materials transportation leakage accidents based on accident area simulation as described in claim 1, characterized in that, Cluster analysis of accident consequence indicators using the K-means algorithm includes: S101. Normalize the accident consequence indicators; S102. Select the initial cluster value K; S103. Randomly select a data point as C1, and select the next initial cluster center based on the principle of maximizing the P value, until the number of initial cluster centers reaches K; S104. Calculate the Euclidean distance D(X) between the sample and the initial cluster center. i C j Cluster centers are determined based on the principle of minimizing Euclidean distance. S105. Calculate the arithmetic mean of each dataset and use the arithmetic mean to represent the new cluster centers; S106. If the elements of the new clusters have not changed, the iteration ends; if they have changed, return to S103.

Images