Expressway hazardous chemical vehicle accident rescue and evacuation path planning method based on CVaR

Abstract

The invention discloses a CVaR-based highway hazardous chemical vehicle accident rescue and evacuation path planning method, and belongs to the technical field of traffic safety and emergency. The path planning method comprises the following steps: acquiring dangerous chemical vehicle transportation accident occurrence points, rescue vehicle rescue points, accident vehicle placement points (evacuation destinations) and road network related parameters; the rescue stage is a path from a rescue exit point to an accident occurrence point of a rescue vehicle, comprehensively considers that the response time of the rescue vehicle is shortest, the efficiency is highest and an upper-layer rescue vehicle travel time shortest model is established, considers that a secondary accident is extremely easy to occur in a dangerous chemical vehicle transportation accident on a highway, and constructs a lower-layer model with the minimum secondary accident risk based on CvaR; a highway dangerous chemical vehicle accident rescue double-layer game planning model is formed; in the evacuation stage, a dangerous chemical substance accident vehicle is a path from an accident occurrence point to a vehicle placement point, the possibility of pollution caused by dangerous chemical substance leakage in the evacuation stage is considered, and an accident vehicle evacuation path planning model based on the CVaR dangerous chemical substance leakage risk is established. According to the highway hazardous chemical vehicle accident rescue and evacuation path planning method provided by the invention, the road traffic safety level and the emergency rescue efficiency can be improved, and smooth and orderly operation of highway traffic is ensured.

Description

Technical Field

[0001] The present invention relates to the field of traffic safety and emergency technology, and in particular to a CVaR-based highway hazardous chemicals vehicle accident rescue and evacuation path planning method. Background Art

[0002] Due to the special properties of hazardous chemicals, any activities related to their use are accompanied by huge risks. After a hazardous chemical transportation accident occurs, reasonable route planning can effectively avoid high-risk areas such as traffic congestion and densely populated areas, ensuring rescue and evacuation via the route with the shortest travel time and the lowest risk for rescue vehicles.

[0003] Expressways have the functional characteristics of high speed, large traffic volume, large interruption loss, limited openness, large space, and environmental fragility. Compared with ordinary roads, traffic accidents on expressways are more destructive and have more serious consequences. Due to their closed nature, the difficulty of accident rescue also increases.

[0004] In the past, research on hazardous chemicals transportation often focused on the selection of transportation routes, lacking in-depth discussions on transportation accidents and emergency rescue in special scenarios. Therefore, in order to make up for the defects in previous studies, this study proposed a CVaR-based method for emergency rescue and evacuation route planning for hazardous chemicals on highways, targeting the problem of rescue and evacuation of hazardous chemicals accidents on highways. Summary of the invention

[0005] The purpose of the present invention is to provide a method for planning accident rescue and evacuation routes for hazardous chemicals transport vehicles on highways based on CVaR, so as to solve the problems of how to ensure that rescue vehicles can efficiently and safely carry out rescue in complex environments after an accident occurs for hazardous chemicals transport vehicles, and how to plan an evacuation route with the lowest risk for accident vehicles, so as to ensure the normal operation of highway network traffic and reduce the negative impact of hazardous chemicals vehicle accidents.

[0006] To achieve the above object, the present invention provides the following technical solutions:

[0007] A method for planning a rescue and evacuation path for a highway hazardous chemicals transport vehicle accident based on CVaR, comprising the following steps:

[0008] Step 1: Obtain the points where hazardous chemicals transport accidents occur, rescue vehicle exit points, accident vehicle resettlement points (evacuation destinations), and road network related parameters; convert the road network into a directed network topology graph. The nodes in the directed network topology graph include accident points, rescue points, resettlement points, and intersection points of various road sections, and the line segments represent the connected road sections connecting the nodes.

[0009] Step 2: The rescue stage is the path from the rescue point to the accident point of the rescue vehicle. Taking into account the shortest response time and highest efficiency of the rescue vehicle, a model with the shortest travel time for the upper-level rescue vehicle is established. Considering that highway hazardous chemical vehicle transportation accidents are prone to secondary accidents, a lower-level model with the lowest secondary accident risk based on CvaR is constructed, forming a two-level game planning model for highway hazardous chemical vehicle accident rescue to select the best rescue path.

[0010] The conditional value at risk (CVaR) represents the average loss above the value at risk (VaR) level, that is, it focuses on the tail risk that exceeds the VaR value, and avoids extreme events by controlling the tail risk, such as leakage, explosion and serious secondary accidents caused by hazardous chemical transportation accidents. The CVaR model is a consistent risk measurement model with good mathematical properties. Therefore, based on the subadditivity of the CVaR model, we conduct in-depth research on the secondary accident risk and leakage risk caused by hazardous chemical vehicle transportation accidents to select the best path.

[0011] Firstly, the shortest travel time model for rescue vehicles is proposed:

[0012]

[0013] In the formula, u represents the type of rescue vehicle, T u It represents the total travel time of the U-type rescue vehicle in the rescue phase, ω1 is an undetermined parameter, q is the traffic volume, and c is the road section capacity.

[0014] Among them, t ij represents the travel time of road section ij under free flow state, t ij and γ are calculated as follows:

[0015] t ij = l ij / v ij (2.2)

[0016]

[0017] In the formula, l ij represents the travel distance of road section ij, v ij represents the vehicle speed in the free flow state of road section ij, and ω2 and ω3 are the unknown parameters of the model.

[0018] The CVaR secondary accident risk value can be derived from the transportation risk correction coefficient and the CVaR accident risk value.

[0019] The calculation formula of CVaR accident risk value is:

[0020]

[0021] In the formula, αβ Represents the VaR value at the confidence level β; It represents the CVaR value under the confidence level β. + represents a positive real number, f(x,y) is the distribution function value of accident loss, and p(y) represents the average accident probability of the road section. β represents the confidence level in statistics, that is, the probability that the risk of hazardous chemicals transportation is lower than the threshold.

[0022] in, represents the risk probability, which is calculated as follows:

[0023]

[0024] Combining the above two formulas, solving the risk probability under the confidence level β is equivalent to solving F β , that is, the final CVaR value is obtained under different confidence levels β, and the calculation formula is:

[0025]

[0026] In the formula, R m Indicates that the value of y is in the range of positive real numbers.

[0027] Taking into account the changes in risk probability caused by external factors such as road linearity, emergency service level, weather, and driving time period, the transportation risk correction coefficient is expressed as follows:

[0028] F N =m·h ij ·c ij ·t d ·n ij (2.7)

[0029] In the formula, F N It represents the correction coefficient of the probability of road accidents from the vehicle rescue point to the accident point. ... It represents the correction coefficient of the service level of the unit where the rescue vehicle is located, h ij Represents the linear correction coefficient of the road section, c ij Indicates the climate factor correction coefficient of the road section, n ij Represents the road traffic flow correction coefficient, t d Indicates the correction factor for the vehicle's driving period.

[0030] Furthermore, the secondary accident risk model based on CVaR can be concluded as follows:

[0031]

[0032] In the formula, C ij It represents the loss value caused by an accident on the path from i to j.

[0033] Game theory studies the decisions made by decision-makers when they interact directly with each other, and solves the equilibrium problem between decisions. In the master-slave game, the upper decision-maker decides the path strategy with the shortest travel time and passes this strategy to the lower decision-maker; the lower decision-maker, as the follower of the upper decision-maker, unconditionally obeys the planning decision of the upper level and makes the path selection strategy with the least risk based on this decision. The best strategy of the lower level will be passed to the upper decision-maker, and the upper level will further update the best path decision based on the decision of the lower level and pass the new decision to the lower level. Repeat the above iterative process until the decision no longer changes, that is, the game equilibrium point is reached.

[0034] The two-level game model in the rescue phase can be expressed as follows:

[0035]

[0036] In the above two-layer model, the * in the superscript represents the Nash equilibrium point of the master-slave game. The upper objective function (2.1) aims to minimize the rescue travel time of the rescue vehicle arriving at the scene; the lower objective function (2.8) aims to minimize the CVaR risk value of the secondary accident during the rescue vehicle arriving at the scene.

[0037] Subject to the following constraints:

[0038] T u ≤T uMax

[0039]

[0040] ΣNum u ≥0

[0041] Num u ≤Num uMax

[0042]

[0043] α,β∈R

[0044] The first constraint ensures that the rescue phase is completed within an acceptable time window, where T uMax The upper threshold limit for emergency response corresponding to U-type rescue vehicles;

[0045] The second constraint ensures the accessibility of the road from any starting point to the end point, where Q ij It represents the final selected path obtained according to the model in the rescue phase;

[0046] The third constraint ensures that rescue vehicles arrive at the accident scene for emergency response;

[0047] The fourth constraint represents the maximum number of different types of rescue vehicles responding.u Indicates the number of different types of rescue vehicles;

[0048] The fifth constraint represents the flow balance constraint, that is, for each road network node, each rescue vehicle can only visit it once at most;

[0049] The sixth constraint X ij represents a binary decision variable, ensuring that the selected road is within the interval;

[0050] The seventh constraint states that both α and β are real numbers.

[0051] Step 3: The evacuation phase is the process of evacuating the hazardous chemical accident vehicle to a safe resettlement point after the rescue vehicle responds to and handles the current accident. Since hazardous chemicals have special properties such as flammability, explosiveness, radiation, and toxicity, they are prone to leakage after a traffic accident, resulting in different types of serious accident consequences. In addition, the leakage of hazardous chemicals will cause varying degrees of environmental pollution, and the chemicals leaked in the air may also cause certain harm to human health. Therefore, the evacuation phase focuses on the risk of hazardous chemical leakage and establishes a CvaR-based hazardous chemical accident vehicle evacuation path planning model.

[0052] Considering the leakage risk probability of hazardous chemicals accident vehicles during the evacuation stage, as well as the superposition of external factors such as road linearity, emergency service level, weather, and driving time on the risk, the correction coefficient F of the probability of road accidents during the evacuation stage is S It is expressed by the following formula:

[0053] F S =h ij ·c ij ·t d ·n ij +l ij ·t d ·p h (3.1)

[0054] In the formula, p h represents the conditional probability correction coefficient of accident vehicle leakage, and the other parameters are as shown above.

[0055] The evacuation stage model refers to the CVaR model of the first stage and is as follows:

[0056]

[0057] Subject to the following constraints:

[0058]

[0059] α,β∈R

[0060] The first constraint ensures the reachability of any starting point to the end point, where W ij represents the final selected path obtained according to the model during the evacuation phase;

[0061] The second constraint represents the flow balance constraint, and for each road network node, the hazardous chemicals transportation accident vehicle can only visit it once at most;

[0062] The third constraint X ij represents a binary decision variable, ensuring that the selected road is within the interval;

[0063] The fourth constraint states that both α and β are real numbers.

[0064] Substantial effects of the present invention: The present invention aims at the emergency rescue and evacuation issues of hazardous chemical transportation accidents on highways. Considering that hazardous chemical transportation vehicle accidents are prone to cause secondary accidents, and external uncertainties are superimposed thereon, a CVaR-based rescue vehicle path planning method for hazardous chemical transportation accidents is established, with the highest rescue efficiency and the lowest emergency risk as the goal, a two-layer game model is constructed. In addition, considering the possibility of environmental pollution, casualties, etc. caused by the leakage of hazardous chemicals during the evacuation of hazardous chemical vehicles from the accident site to the resettlement site, a CVaR-conditional leakage risk accident vehicle evacuation path planning method is established to reduce the serious harm caused by the leakage of hazardous chemicals from accident vehicles. This method targets the entire rescue process of hazardous chemical transportation accidents on highways, greatly improves the level of emergency rescue, and reduces the dangerous consequences caused by hazardous chemical accidents and leaks. BRIEF DESCRIPTION OF THE DRAWINGS

[0065] Figure 1 The modeling basis and steps of the embodiment of the present invention are shown in FIG. DETAILED DESCRIPTION

[0066] The present invention will be further described below. It should be noted that this embodiment is based on the technical solution and provides a detailed implementation method and a specific operation process, but the protection scope of the present invention is not limited to this embodiment.

[0067] This embodiment provides a method for planning a rescue and evacuation path for a highway hazardous chemicals transport vehicle accident based on CVaR. The specific implementation method of the present invention is further described in detail in conjunction with the following drawings:

[0068] Step 1: Figure 1 As shown, the points where hazardous chemicals transport accidents occur, the rescue points for rescue vehicles, the resettlement points for accident vehicles (evacuation destinations) and related parameters of the road network are obtained; the road network is converted into a directed network topology diagram, in which the nodes include accident points, rescue points, resettlement points and intersection points of various road sections, and the line segments represent the connected road sections connecting the nodes.

[0069] Step 2: If Figure 1 As shown in the figure, the rescue stage is the path of the rescue vehicle from the rescue point to the accident point. Taking into account the shortest response time and highest efficiency of the rescue vehicle, a model with the shortest travel time of the upper-level rescue vehicle is established. Considering that highway hazardous chemical vehicle transportation accidents are prone to secondary accidents, a lower-level model with the lowest secondary accident risk based on CvaR is constructed, forming a two-level game planning model for highway hazardous chemical vehicle accident rescue to select the best rescue path.

[0070] The conditional value at risk (CVaR) represents the average loss above the value at risk (VaR) level, that is, it focuses on the tail risk that exceeds the VaR value, and avoids extreme events by controlling the tail risk, such as leakage, explosion and serious secondary accidents caused by hazardous chemical transportation accidents. The CVaR model is a consistent risk measurement model with good mathematical properties. Therefore, based on the subadditivity of the CVaR model, we conduct in-depth research on the secondary accident risk and leakage risk caused by hazardous chemical vehicle transportation accidents to select the best path.

[0071] Firstly, the shortest travel time model for rescue vehicles is proposed:

[0072]

[0073] In the formula, u represents the type of rescue vehicle, T u It represents the total travel time of the U-type rescue vehicle in the rescue phase, ω1 is an undetermined parameter, q is the traffic volume, and c is the road section capacity.

[0074] Among them, t ij represents the travel time of road section ij under free flow state, t ij and γ are calculated as follows:

[0075] t ij = l ij / v ij (2.2)

[0076]

[0077] In the formula, l ij represents the travel distance of road section ij, v ij represents the vehicle speed in the free flow state of road section ij, and ω2 and ω3 are the unknown parameters of the model.

[0078] The CVaR secondary accident risk value can be derived from the transportation risk correction coefficient and the CVaR accident risk value.

[0079] The calculation formula of CVaR accident risk value is:

[0080]

[0081] In the formula, α β Represents the VaR value at the confidence level β; It represents the CVaR value under the confidence level β. + represents a positive real number, f(x,y) is the distribution function value of accident loss, and p(y) represents the average accident probability of the road section. β represents the confidence level in statistics, that is, the probability that the risk of hazardous chemicals transportation is lower than the threshold.

[0082] in, represents the risk probability, which is calculated as follows:

[0083]

[0084] Combining the above two formulas, solving the risk probability under the confidence level β is equivalent to solving F β , that is, the final CVaR value is obtained under different confidence levels β, and the calculation formula is:

[0085]

[0086] In the formula, R m Indicates that the value of y is in the range of positive real numbers.

[0087] Taking into account the changes in risk probability caused by external factors such as road linearity, emergency service level, weather, and driving time period, the transportation risk correction coefficient is expressed as follows:

[0088] F N =m·h ij ·c ij ·t d ·n ij (2.7)

[0089] In the formula, F N It represents the correction coefficient of the probability of road accidents from the rescue point to the accident point. m represents the correction coefficient of the service level of the rescue vehicle unit, h ij Represents the linear correction coefficient of the road section, c ij Indicates the climate factor correction coefficient of the road section, n ij Represents the road traffic flow correction coefficient, t d Indicates the correction factor for the vehicle's driving period.

[0090] Furthermore, the secondary accident risk model based on CVaR can be concluded as follows:

[0091]

[0092] In the formula, C ij It represents the loss value caused by the accident on the path from i to j.

[0093] Game theory studies the decisions made by decision-makers when they interact directly with each other, and solves the equilibrium problem between decisions. In the master-slave game, the upper decision-maker decides the path strategy with the shortest travel time and passes this strategy to the lower decision-maker; the lower decision-maker, as the follower of the upper decision-maker, unconditionally obeys the planning decision of the upper level and makes the path selection strategy with the least risk based on this decision. The best strategy of the lower level will be passed to the upper decision-maker, and the upper level will further update the best path decision based on the decision of the lower level and pass the new decision to the lower level. Repeat the above iterative process until the decision no longer changes, that is, the game equilibrium point is reached.

[0094] The two-level game model in the rescue phase can be expressed as follows:

[0095]

[0096] In the above two-layer model, the * in the superscript represents the Nash equilibrium point of the master-slave game. The upper objective function (2.1) aims to minimize the rescue travel time of the rescue vehicle arriving at the scene; the lower objective function (2.8) aims to minimize the CVaR risk value of the secondary accident during the rescue vehicle arriving at the scene.

[0097] Subject to the following constraints:

[0098] T u ≤T uMax

[0099]

[0100] ∑Num u ≥0

[0101] Num u ≤Num uMax

[0102]

[0103] α,β∈R

[0104] The first constraint ensures that the rescue phase is completed within an acceptable time window, where T uMax The upper threshold limit for emergency response corresponding to U-type rescue vehicles;

[0105] The second constraint ensures the accessibility of the road from any starting point to the end point, where Q ij It represents the final selected path obtained according to the model in the rescue phase;

[0106] The third constraint ensures that rescue vehicles arrive at the accident scene for emergency response;

[0107] The fourth constraint represents the maximum number of different types of rescue vehicles responding. u Indicates the number of different types of rescue vehicles;

[0108] The fifth constraint represents the flow balance constraint, that is, for each road network node, each rescue vehicle can only visit it once at most;

[0109] The sixth constraint X ij represents a binary decision variable, ensuring that the selected road is within the interval;

[0110] The seventh constraint states that both α and β are real numbers.

[0111] Step 3: If Figure 1 As shown in the figure, the evacuation phase is the process of evacuating the hazardous chemical accident vehicle to a safe resettlement point after the rescue vehicle responds to and handles the current accident. Since hazardous chemicals have special properties such as flammability, explosiveness, radiation, and toxicity, they are prone to leakage after a traffic accident, resulting in different types of serious accident consequences. In addition, the leakage of hazardous chemicals will bring about different degrees of environmental pollution, and the chemicals leaked in the air may also cause certain harm to human health. Therefore, the evacuation phase focuses on the risk of hazardous chemical leakage and establishes a CvaR-based hazardous chemical accident vehicle evacuation path planning model.

[0112] Considering the leakage risk probability of hazardous chemicals accident vehicles during the evacuation stage, as well as the superposition of external factors such as road linearity, emergency service level, weather, and driving time on the risk, the correction coefficient F of the probability of road accidents during the evacuation stage is S It is expressed by the following formula:

[0113] F S =h ij ·c ij ·t d ·n ij +l ij ·t d ·p h (3.1)

[0114] In the formula, p h represents the conditional probability correction coefficient of accident vehicle leakage, and the other parameters are as shown above.

[0115] The evacuation stage model refers to the CVaR model of the first stage and is as follows:

[0116]

[0117] Subject to the following constraints:

[0118]

[0119]

[0120] α,β∈R

[0121] The first constraint ensures the reachability of any starting point to the end point, where W ij represents the final selected path obtained according to the model during the evacuation phase;

[0122] The second constraint represents the flow balance constraint, and for each road network node, the hazardous chemicals transportation accident vehicle can only visit it once at most;

[0123] The third constraint X ij represents a binary decision variable, ensuring that the selected road is within the interval;

[0124] The fourth constraint states that both α and β are real numbers.

Claims

1. A method for planning a rescue and evacuation path for a highway hazardous chemicals transport vehicle accident based on CVaR, characterized in that: Step 1: Obtain the points where hazardous chemicals transport accidents occur, rescue vehicle exit points, accident vehicle resettlement points (evacuation destinations), and road network related parameters; convert the road network into a directed network topology graph. The nodes in the directed network topology graph include accident points, rescue points, resettlement points, and intersection points of various road sections, and the line segments represent the connected road sections connecting the nodes. Step 2: The rescue stage is the path from the rescue point to the accident point of the rescue vehicle. Taking into account the shortest response time and highest efficiency of the rescue vehicle, a model with the shortest travel time for the upper-level rescue vehicle is established. Considering that highway hazardous chemical vehicle transportation accidents are prone to secondary accidents, a lower-level model with the lowest secondary accident risk based on CvaR is constructed, forming a two-level game planning model for highway hazardous chemical vehicle accident rescue to select the best rescue path. The conditional value at risk (CVaR) represents the average loss above the value at risk (VaR) level, that is, it focuses on the tail risk that exceeds the VaR value, and avoids extreme events by controlling the tail risk, such as leakage, explosion and serious secondary accidents caused by hazardous chemical transportation accidents. The CVaR model is a consistent risk measurement model with good mathematical properties. Therefore, based on the subadditivity of the CVaR model, we conduct in-depth research on the secondary accident risk and leakage risk caused by hazardous chemical vehicle transportation accidents to select the best path. Firstly, the shortest travel time model for rescue vehicles is proposed: In the formula, u represents the type of rescue vehicle, T u It represents the total travel time of the U-type rescue vehicle in the rescue phase, ω1 is an undetermined parameter, q is the traffic volume, and c is the road section capacity. Among them, t ij represents the travel time of road section ij under free flow state, t ij and γ are calculated as follows: t ij =l ij / v ij (2.2) In the formula, l ij represents the travel distance of road section ij, v ij represents the vehicle speed in the free flow state of road section ij, and ω2 and ω3 are the unknown parameters of the model. The CVaR secondary accident risk value can be derived from the transportation risk correction coefficient and the CVaR accident risk value. The calculation formula of CVaR accident risk value is: In the formula, α β Represents the VaR value at the confidence level β; It represents the CVaR value under the confidence level β. + represents a positive real number, f(x,y) is the distribution function value of accident loss, and p(y) represents the average accident probability of the road section. β represents the confidence level in statistics, that is, the probability that the risk of hazardous chemicals transportation is lower than the threshold. in, represents the risk probability, which is calculated as follows: Combining the above two formulas, solving the risk probability under the confidence level β is equivalent to solving F β , that is, the final CVaR value is obtained under different confidence levels β, and the calculation formula is: In the formula, R m Indicates that the value of y is in the range of positive real numbers. Taking into account the changes in risk probability caused by external factors such as road linearity, emergency service level, weather, and driving time period, the transportation risk correction coefficient is expressed as follows: F N =m·h ij ·c ij ·t d ·n ij (2.7) In the formula, F N It represents the correction coefficient of the probability of road accidents from the rescue point to the accident point. m represents the correction coefficient of the service level of the rescue vehicle unit, h ij Represents the linear correction coefficient of the road section, c ij Indicates the climate factor correction coefficient of the road section, n ij Represents the road traffic flow correction coefficient, t d Indicates the correction factor for the vehicle's driving period. Furthermore, the secondary accident risk model based on CVaR can be concluded as follows: In the formula, C ij It represents the loss value caused by an accident on the path from i to j. Game theory studies the decisions made by decision-makers when they interact directly with each other, and solves the equilibrium problem between decisions. In the master-slave game, the upper decision-maker decides the path strategy with the shortest travel time and passes this strategy to the lower decision-maker; the lower decision-maker, as the follower of the upper decision-maker, unconditionally obeys the planning decision of the upper level and makes the path selection strategy with the least risk based on this decision. The best strategy of the lower level will be passed to the upper decision-maker, and the upper level will further update the best path decision based on the decision of the lower level and pass the new decision to the lower level. Repeat the above iterative process until the decision no longer changes, that is, the game equilibrium point is reached. The two-level game model in the rescue phase can be expressed as follows: In the above two-layer model, the * in the superscript represents the Nash equilibrium point of the master-slave game. The upper objective function (2.1) aims to minimize the rescue travel time of the rescue vehicle arriving at the scene; the lower objective function (2.8) aims to minimize the CVaR risk value of the secondary accident during the rescue vehicle arriving at the scene. Subject to the following constraints: ∑Num u ≥0 In a u ≤Num uMax α,β∈R The first constraint ensures that the rescue phase is completed within an acceptable time window, where T uMax The upper threshold limit for emergency response corresponding to U-type rescue vehicles; The second constraint ensures the accessibility of the road from any starting point to the end point, where Q ij It represents the final selected path obtained according to the model in the rescue phase; The third constraint ensures that rescue vehicles arrive at the accident scene for emergency response; The fourth constraint represents the maximum number of different types of rescue vehicles responding. u Indicates the number of different types of rescue vehicles; The fifth constraint represents the flow balance constraint, that is, for each road network node, each rescue vehicle can only visit it once at most; The sixth constraint X ij represents a binary decision variable, ensuring that the selected road is within the interval; The seventh constraint states that both α and β are real numbers. Step 3: The evacuation phase is the process of evacuating the hazardous chemical accident vehicle to a safe resettlement point after the rescue vehicle responds to and handles the current accident. Since hazardous chemicals have special properties such as flammability, explosiveness, radiation, and toxicity, they are prone to leakage after a traffic accident, resulting in different types of serious accident consequences. In addition, the leakage of hazardous chemicals will cause varying degrees of environmental pollution, and the chemicals leaked in the air may also cause certain harm to human health. Therefore, the evacuation phase focuses on the risk of hazardous chemical leakage and establishes a CvaR-based hazardous chemical accident vehicle evacuation path planning model. Considering the leakage risk probability of hazardous chemicals accident vehicles during the evacuation stage, as well as the superposition of external factors such as road linearity, emergency service level, weather, and driving time on the risk, the correction coefficient F of the probability of road accidents during the evacuation stage is S It is expressed by the following formula: F S=h ij ·c ij ·t d ·n ij +l ij ·t d ·p h (3.1) In the formula, p h represents the conditional probability correction coefficient of accident vehicle leakage, and the other parameters are as shown above. The evacuation stage model refers to the CVaR model of the first stage and is as follows: Subject to the following constraints: α,β∈R The first constraint ensures the reachability of any starting point to the end point, where W ij represents the final selected path obtained according to the model during the evacuation phase; The second constraint represents the flow balance constraint, and for each road network node, the hazardous chemicals transportation accident vehicle can only visit it once at most; The third constraint X ij represents a binary decision variable, ensuring that the selected road is within the interval; The fourth constraint states that both α and β are real numbers.

Images