Dynamic risk prevention and control system and method for hazardous chemicals transportation routes based on time-space fusion

Abstract

The present invention relates to the field of intelligent transportation safety, and in particular to a dynamic risk prevention and control system and method for hazardous chemical transportation routes based on spatiotemporal fusion. The present invention proposes calculating a multidimensional risk index matrix through a risk index quantification module, and combining it with a spatiotemporal environmental sensitivity prediction matrix generated by an environmental sensitivity analysis module and a risk propagation probability cloud map of a risk diffusion prediction module to achieve accurate assessment of the dynamic risk of road sections; the system establishes a collaborative risk factor assessment model, adjusts safety thresholds based on historical accident data, and generates a spatiotemporal road network risk scoring matrix; the risk prevention and control decision-making module determines the warning level based on this, selects prevention and control measures, optimizes transportation routes, formulates emergency response plans for high-risk sections, and outputs an intelligent decision-making instruction set; the system achieves comprehensive quantification of hazardous chemical characteristics, improves the comprehensiveness of risk assessment and the intelligence of prevention and control decisions.

Description

Technical Field

[0001] The present invention relates to the field of intelligent transportation safety, and in particular to a dynamic risk prevention and control system and method for hazardous chemical transportation routes based on spatiotemporal fusion, which is used to predict, evaluate and prevent and control safety risks in the transportation of hazardous chemicals. Background Art

[0002] There are major safety risks in the transportation of hazardous chemicals. Once an accident such as leakage, fire or explosion occurs, it will not only cause direct economic losses, but may also cause serious environmental pollution and casualties.

[0003] In existing technology, Chinese patent CN 107591012 B discloses a hazardous chemical transportation early warning method. This method establishes a vehicle balance model for different road conditions and loads. Based on current vehicle state and road information, it estimates the vehicle's balance during future road travel and determines whether to issue an alert. This technology primarily focuses on the vehicle's physical stability, mitigating accidents such as rollovers by predicting the vehicle's balance under specific road conditions.

[0004] However, existing technologies have the following shortcomings: First, they focus solely on the physical stability of vehicles, ignoring the risks associated with the interaction between the properties of hazardous chemicals and the external environment; second, they fail to consider the differences in the physical and chemical properties of different hazardous chemicals and the impact range of a leak; third, they lack a weighted risk assessment for sensitive areas such as densely populated areas; fourth, they fail to provide specific alternative route planning options after early warnings; fifth, the system fails to integrate dynamic factors such as real-time traffic congestion and weather conditions; and sixth, data transmission and storage lack security mechanisms. These shortcomings severely limit the effectiveness of risk prevention and control, making it impossible to achieve systematic security assurance for the entire hazardous chemical transportation process. Summary of the Invention

[0005] The purpose of the present invention is to provide a dynamic risk prevention and control system and method for hazardous chemical transportation routes based on spatiotemporal fusion, aiming to overcome the shortcomings of the existing technology that only focuses on vehicle stability while ignoring the interaction between hazardous chemical characteristics and the environment, and to achieve a technological leap from simple vehicle physical balance warning to all-round and multi-dimensional risk perception and active prevention and control.

[0006] To achieve the above objectives, the present invention provides a dynamic risk prevention and control system for hazardous chemical transportation routes based on spatiotemporal fusion, including a risk index quantification module, an environmental sensitivity analysis module, a risk diffusion prediction module, a road section risk assessment module and a risk prevention and control decision-making module. Through the coordinated operation of the five modules, a complete closed-loop system from risk perception and assessment to active prevention and control is formed.

[0007] The present invention proposes a dynamic risk prevention and control system for hazardous chemicals transportation routes based on time-space fusion, including:

[0008] Risk index quantification module, used for:

[0009] Receive hazardous chemical property parameters, including physical state, chemical activity, toxicity level, explosiveness, flash point, vapor pressure, water solubility, corrosivity, reactivity, stability, volatility, carcinogenicity, teratogenicity, bioaccumulation, and environmental persistence;

[0010] Based on the hazardous chemical property parameters, calculate the leakage risk index, fire and explosion risk index, health hazard risk index, environmental pollution risk index and reaction hazard risk index;

[0011] Generate a multi-dimensional risk index matrix for hazardous chemicals;

[0012] The environmental sensitivity analysis module is connected to the risk index quantification module and is used to:

[0013] Acquiring environmentally sensitive data, including real-time meteorological data, geographic information data, population distribution data, and sensitive target data;

[0014] Construct a gridded space-time fusion model and calculate the environmental sensitivity index for each space-time grid;

[0015] A Bayesian spatiotemporal network prediction model is established based on historical data to predict the changes in environmental sensitivity of each grid in the future period;

[0016] Generate spatiotemporal environmental sensitivity prediction matrix;

[0017] The risk diffusion prediction module is connected to the risk index quantification module and the environmental sensitivity analysis module and is used to:

[0018] Constructing a multi-media migration model based on the multi-dimensional risk index matrix of hazardous chemicals and the spatiotemporal environmental sensitivity prediction matrix;

[0019] Preset multiple leakage scenarios and conduct Monte Carlo risk propagation simulations;

[0020] Generate risk propagation probability cloud map;

[0021] The road section risk assessment module is connected to the risk diffusion prediction module and is used to:

[0022] Receive road network data and discretize it into standard road segments;

[0023] Calculate the dynamic risk score of each road section at different time periods by combining the multi-dimensional risk index matrix of hazardous chemicals, the spatiotemporal environmental sensitivity prediction matrix, and the risk propagation probability cloud map;

[0024] Establish a collaborative risk factor assessment model;

[0025] Dynamically adjust safety thresholds based on historical accident data;

[0026] Generate a spatiotemporal road network risk score matrix;

[0027] The risk prevention and control decision module is connected to the road section risk assessment module and is used to:

[0028] Determining a warning level based on a comparison between the spatiotemporal road network risk score matrix and a safety threshold;

[0029] According to different warning levels, select corresponding prevention and control measures from the prevention and control measures library;

[0030] Execute multi-objective optimization decision to generate the optimal transportation route;

[0031] Generate emergency response plans for high-risk road sections;

[0032] Output risk intelligent decision-making instruction set.

[0033] Preferably, the risk index quantification module includes:

[0034] Parameter characterization unit, used to standardize hazardous chemical property parameters into a scale of 0-10;

[0035] The risk calculation unit is used to calculate five types of risk indices based on standardized hazardous chemical property parameters;

[0036] Weight adjustment unit, used to dynamically adjust the weight coefficient according to different scenarios and calculate the comprehensive risk index :

[0037] ,

[0038] in, is the leakage risk index, is the fire and explosion risk index, is the health hazard risk index, is the environmental pollution risk index, To reflect the hazard risk index, to is the weight coefficient, and .

[0039] Preferably, the environmental sensitivity analysis module includes:

[0040] Data access unit, used to acquire and preprocess multi-source environmental data;

[0041] Grid division unit, used to construct a standard grid unit of 100m×100m and establish a 24-hour×7-day time matrix on each grid;

[0042] Sensitivity calculation unit, used to calculate the environmental sensitivity index ESI of each spatiotemporal grid:

[0043] ,

[0044] in, For Grid In time The population density index, For Grid The water sensitivity index, For Grid Special sensitive target index, For Grid In time The weather condition index, is the weight coefficient and 1;

[0045] The prediction modeling unit is used to build a Bayesian network based on historical data to predict the changes in the environmental sensitivity of each grid within the next 24 hours.

[0046] Preferably, the risk diffusion prediction module includes:

[0047] Migration model unit, used to establish migration and diffusion models of three media: atmosphere, water and soil;

[0048] Scenario preset unit, used to preset various leakage scenarios such as small leakage, medium leakage, large leakage, fire-associated leakage and explosion-associated leakage according to the characteristics of hazardous chemicals and environmental conditions;

[0049] The Monte Carlo simulation unit is used to perform 10,000 Monte Carlo simulations for each leakage scenario and generate a risk probability distribution cloud map.

[0050] Preferably, the migration model unit includes:

[0051] The atmospheric diffusion calculation subunit uses an improved Gaussian plume model to calculate the diffusion concentration of hazardous chemicals in the atmosphere;

[0052] The water diffusion calculation subunit uses the two-dimensional convection-diffusion equation to calculate the migration process of hazardous chemicals in water bodies;

[0053] The soil penetration calculation subunit simulates the penetration behavior of hazardous chemicals in soil based on the Richards equation.

[0054] Preferably, the road section risk assessment module includes:

[0055] The road segment discretization unit is used to discretize the road network into standard 500m length sections;

[0056] The risk scoring unit is used to calculate the dynamic risk score RSI of each road segment at different time periods:

[0057] ,

[0058] in, is the comprehensive risk index of hazardous chemicals, For road sections In time The environmental sensitivity index, is the impact amplification factor;

[0059] Collaborative risk assessment unit, used to evaluate the collaborative risk factors of vehicle status, road conditions, driving behavior patterns and hazardous chemicals characteristics;

[0060] Threshold adaptive unit, used to dynamically adjust safety thresholds based on historical accident data:

[0061] ,in, is the learning rate, TargetSafety is the target safety level, and CurrentSafety is the current safety level.

[0062] Preferably, the risk prevention and control decision module includes:

[0063] The warning determination unit is used to determine the four warning levels of green, yellow, orange or red based on the comparison of the road section risk score and the safety threshold;

[0064] The prevention and control measures unit is used to select corresponding driver operation suggestions, vehicle parameter adjustments and route planning changes based on the warning level;

[0065] Multi-objective optimization unit, used to make multi-objective optimization decisions for path planning by considering the three objectives of safety, timeliness, and economy;

[0066] The plan generation unit is used to generate emergency response plans for high-risk road sections.

[0067] Preferably, the optimization function of the multi-objective optimization unit is:

[0068] ,

[0069] in, is the safety target, reflecting the total risk of the path, For timeliness goals, reflect the transportation time, For economic purposes, it reflects transportation costs;

[0070] The constraints are:

[0071] Maximum road section risk score Allowable risk threshold,

[0072] Total transportation time Maximum permissible transit time,

[0073] Total shipping cost Budget cap.

[0074] As a preference, the data flow process between modules is as follows:

[0075] The multi-dimensional risk index matrix of hazardous chemicals generated by the risk index quantification module is transmitted to the environmental sensitivity analysis module and the risk diffusion prediction module;

[0076] The spatiotemporal environmental sensitivity prediction matrix generated by the environmental sensitivity analysis module is transmitted to the risk diffusion prediction module and the road section risk assessment module;

[0077] The risk propagation probability cloud map generated by the risk diffusion prediction module is transmitted to the road section risk assessment module;

[0078] The spatiotemporal road network risk scoring matrix generated by the road section risk assessment module is transmitted to the risk prevention and control decision module;

[0079] The risk intelligent decision instruction set of the risk prevention and control decision module is transmitted to the vehicle terminal and the monitoring center through the communication network.

[0080] The method corresponding to the dynamic risk prevention and control system for hazardous chemicals transportation routes based on spatiotemporal fusion as described above includes the following steps:

[0081] Receive hazardous chemical attribute parameters and calculate the multi-dimensional risk index matrix of hazardous chemicals;

[0082] Acquire environmental sensitive data, build a gridded spatiotemporal fusion model, and generate a spatiotemporal environmental sensitivity prediction matrix;

[0083] Based on the multi-dimensional risk index matrix of hazardous chemicals and the spatiotemporal environmental sensitivity prediction matrix, a multi-media migration model is constructed, a Monte Carlo risk propagation simulation is performed, and a risk propagation probability cloud map is generated;

[0084] Discrete the road network into standard road sections, and calculate the spatiotemporal road network risk score matrix by combining the multidimensional risk index matrix of hazardous chemicals, the spatiotemporal environmental sensitivity prediction matrix, and the risk propagation probability cloud map;

[0085] Based on the comparison between the spatiotemporal road network risk scoring matrix and the safety threshold, the warning level is determined, the prevention and control measures are selected, the multi-objective optimization decision is executed to generate the optimal transportation route, and the risk intelligent decision instruction set is output.

[0086] The beneficial effects of this invention are as follows: First, it achieves multi-dimensional quantification of hazardous chemical characteristics and establishes five risk indices for leakage, fire and explosion, health hazards, environmental pollution, and reaction hazards, making risk assessment more comprehensive; second, it constructs a spatiotemporal environmental sensitivity model to dynamically assess changes in the sensitivity of the surrounding environment along the transportation route; third, through Monte Carlo simulation, it visualizes the potential impact of hazardous chemical leaks, supporting more accurate risk assessment; fourth, it innovatively integrates vehicle status, road conditions, and driving behavior to construct a collaborative risk assessment model; and fifth, based on a multi-objective optimization algorithm, it achieves a balance between safety, timeliness, and economy to generate the optimal transportation route. Experimental data shows that this system can reduce exposure time on high-risk sections by 55%, advance emergency response preparation time by 12 minutes, and effectively reduce the incidence of hazardous chemical transportation accidents. BRIEF DESCRIPTION OF THE DRAWINGS

[0087] Figure 1 This is the overall architecture diagram of the dynamic risk prevention and control system for hazardous chemicals transportation routes based on time-space fusion of the present invention;

[0088] Figure 2 This is a schematic diagram of the structure of the risk index quantification module of the present invention;

[0089] Figure 3 Schematic diagram of the structure of the environmental sensitivity analysis module of the present invention;

[0090] Figure 4 Schematic diagram of the structure of the risk diffusion prediction module of the present invention;

[0091] Figure 5 Schematic diagram of the structure of the road section risk assessment module of the present invention;

[0092] Figure 6 This is a schematic diagram of the structure of the risk prevention and control decision module of the present invention;

[0093] Figure 7 This is a flow chart of the method for dynamic risk prevention and control of hazardous chemicals transportation routes of the present invention;

[0094] Figure 8 This is a comparison diagram of the application effects of the present invention. DETAILED DESCRIPTION

[0095] Please refer to Figures 1-8 The present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0096] Example 1: Overall Architecture of a Dynamic Risk Prevention and Control System for Hazardous Chemical Transportation Routes Based on Spatiotemporal Fusion

[0097] like Figure 1As shown, the dynamic risk prevention and control system for hazardous chemicals transportation routes based on time-space fusion provided by the present invention includes a risk index quantification module 1, an environmental sensitivity analysis module 2, a risk diffusion prediction module 3, a road section risk assessment module 4 and a risk prevention and control decision module 5.

[0098] The risk index quantification module 1 is used to receive hazardous chemical attribute parameters, calculate the leakage risk index, fire and explosion risk index, health hazard risk index, environmental pollution risk index, and reaction hazard risk index based on these parameters, and generate a multi-dimensional risk index matrix for hazardous chemicals. Hazardous chemical attribute parameters include physical state, chemical activity, toxicity level, explosiveness, flash point, vapor pressure, water solubility, corrosivity, reactivity, stability, volatility, carcinogenicity, teratogenicity, bioaccumulation, and environmental persistence.

[0099] Environmental sensitivity analysis module 2 is connected to risk index quantification module 1 to acquire environmental sensitivity data, construct a gridded spatiotemporal fusion model, calculate the environmental sensitivity index for each spatiotemporal grid, and establish a Bayesian spatiotemporal network prediction model based on historical data to predict changes in the environmental sensitivity of each grid in the future and generate a spatiotemporal environmental sensitivity prediction matrix. Environmental sensitivity data includes real-time meteorological data, geographic information data, population distribution data, and sensitive target data.

[0100] The risk diffusion prediction module 3 is connected to the risk index quantification module 1 and the environmental sensitivity analysis module 2. It is used to construct a multi-media migration model based on the multi-dimensional risk index matrix of hazardous chemicals and the spatiotemporal environmental sensitivity prediction matrix, preset multiple leakage scenarios and perform Monte Carlo risk propagation simulation to generate a risk propagation probability cloud map.

[0101] The road section risk assessment module 4 is connected to the risk diffusion prediction module 3, and is used to receive road network data and discretize it into standard sections. It combines the multi-dimensional risk index matrix of hazardous chemicals, the spatiotemporal environmental sensitivity prediction matrix and the risk propagation probability cloud map to calculate the dynamic risk score of each road section at different time periods, establish a collaborative risk factor assessment model, dynamically adjust the safety threshold based on historical accident data, and generate a spatiotemporal road network risk score matrix.

[0102] The risk prevention and control decision module 5 is connected to the road section risk assessment module 4, and is used to determine the warning level based on the comparison of the spatiotemporal road network risk score matrix and the safety threshold, select corresponding prevention and control measures according to different warning levels, perform multi-objective optimization decisions to generate the optimal transportation route, generate emergency response plans for high-risk sections, and output risk intelligent decision-making instruction sets.

[0103] Each module is connected via a standardized data interface, supporting real-time data exchange and processing. The system adopts a distributed architecture, with the core algorithm deployed on a cloud server and data collection and early warning functions deployed on the vehicle terminal device, enabling real-time data transmission and processing via 4G / 5G networks.

[0104] Example 2: Specific implementation of the risk index quantification module

[0105] like Figure 2 As shown, the risk index quantification module 1 includes a parameter characterization unit 11 , a risk calculation unit 12 and a weight adjustment unit 13 .

[0106] The parameter characterization unit 11 receives hazardous chemical property parameters and standardizes them, converting each indicator into a 0-10 scale. For example, for physical state, solids are assigned a value of 1-3, liquids are assigned a value of 4-7, and gases are assigned a value of 8-10. For flash point, the actual temperature is linearly mapped to a scale of 0-10, with lower flash points assigned higher values. This standardization facilitates comparison and comprehensive evaluation of different property parameters.

[0107] The risk calculation unit 12 calculates five risk indices based on the standardized hazardous chemical attribute parameters:

[0108] The formula for calculating the leakage risk index (LRI) is:

[0109] ,

[0110] in, is the physical state parameter value, is the vapor pressure parameter value, is the volatility parameter value, is the stability parameter value, to is the weight coefficient, and .

[0111] Similarly, the Fire and Explosion Risk Index (FERI), Health Hazard Risk Index (HHRI), Environmental Pollution Risk Index (EPRI) and Reaction Hazard Risk Index (RHRI) are also calculated by weighting the corresponding parameters.

[0112] The weight adjustment unit 13 dynamically adjusts the weight coefficient according to different scenarios and calculates the comprehensive risk index CRI:

[0113] ,

[0114] in, to is the weight coefficient, and The weight coefficient will be adjusted dynamically according to the environmental characteristics, such as in densely populated areas. Increase by 30%, in water source protection areas Increase by 40%, in transportation hub areas Increase by 25% on expressway sections Increase by 20%, in chemical industry parks Increased by 35%.

[0115] Preferably, the risk index quantification module 1 further includes a data storage unit 14 for storing attribute parameters of known hazardous chemicals and pre-calculated risk indices to form a hazardous chemical risk knowledge base to support rapid query and risk comparison.

[0116] Example 3: Specific implementation of the environmental sensitivity analysis module

[0117] like Figure 3 As shown, the environmental sensitivity analysis module 2 includes a data access unit 21 , a grid division unit 22 , a sensitivity calculation unit 23 and a prediction modeling unit 24 .

[0118] The data access unit 21 is responsible for acquiring and preprocessing multi-source environmental data, including real-time meteorological data (wind direction, wind speed, temperature, humidity, air pressure, precipitation, with a sampling frequency of 5 minutes / time), geographic information data (topography elevation, water system distribution, vegetation cover, soil type), population distribution data (permanent population density, floating population density, time period distribution characteristics) and sensitive target data (schools, hospitals, water sources, nature reserves, etc., including spatial coordinates and sensitivity levels).

[0119] Gridding unit 22 divides the spatial area into standard 100m x 100m grid cells and creates a 24-hour x 7-day time matrix on each grid, forming a spatiotemporal grid data structure. This fine-grained spatiotemporal division helps accurately capture the spatiotemporal variations in environmental sensitivity.

[0120] The sensitivity calculation unit 23 calculates the environmental sensitivity index ESI of each spatiotemporal grid:

[0121] ,

[0122] in, For Grid In time The population density index, For Grid The water sensitivity index, For Grid Special sensitive target index, For Grid In time The weather condition index, is the weight coefficient and .

[0123] Predictive modeling unit 24 constructs a Bayesian network based on historical data to predict changes in environmental sensitivity for each grid cell over the next 24 hours. This Bayesian network primarily considers factors such as time characteristics (hours, weekdays / holidays, and seasons), historical environmental sensitivity, and weather forecast data, achieving predictions through conditional probabilistic reasoning.

[0124] Preferably, the sensitivity calculation unit 23 also considers temporal fluctuations. For example, during rush hour on weekdays, the population density index weight increases accordingly; during rainy or snowy weather, the water source sensitivity index weight increases accordingly. The prediction modeling unit 24 uses a sliding time window to update the prediction model, updating the prediction results every hour to ensure the timeliness and accuracy of the prediction.

[0125] Example 4: Specific implementation of the risk diffusion prediction module

[0126] like Figure 4 As shown, the risk diffusion prediction module 3 includes a migration model unit 31 , a scenario preset unit 32 and a Monte Carlo simulation unit 33 .

[0127] Migration Model Unit 31 establishes migration and diffusion models for three media: atmosphere, water, and soil, based on the physical and chemical properties of hazardous chemicals and environmental conditions. Atmospheric diffusion uses a modified Gaussian plume model, water diffusion uses a two-dimensional convection-diffusion equation, and soil infiltration uses a model based on the Richards equation.

[0128] The scenario presetting unit 32 presets various leak scenarios, including small leaks, medium leaks, large leaks, fire-related leaks, and explosion-related leaks, based on the characteristics of the hazardous chemicals and environmental conditions. For example, a small leak is defined as a leak volume of 0-100 kg and a leak rate of 0-5 kg / min; a medium leak is defined as a leak volume of 100-1000 kg and a leak rate of 5-50 kg / min; and a large leak is defined as a leak volume greater than 1000 kg and a leak rate greater than 50 kg / min. Fire-related leaks and explosion-related leaks also take into account the impact range of thermal radiation and shock waves.

[0129] Monte Carlo simulation unit 33 performs multiple Monte Carlo simulations for each leak scenario, based on a diffusion model under random meteorological conditions, to generate a risk probability distribution cloud map. The simulation process first randomly generates a set of meteorological conditions that match the local meteorological characteristics. It then calculates the diffusion concentration distribution, determines whether each grid cell exceeds the hazard threshold, calculates the number of people affected and the extent of environmental damage, and ultimately generates a risk probability distribution.

[0130] To improve simulation efficiency, Monte Carlo simulation unit 33 preferably employs a stratified sampling strategy, focusing on sampling key parameters (such as wind direction and speed), while also utilizing parallel computing technology to accelerate the simulation process. For common hazardous chemicals, the system pre-calculates and stores risk propagation probability cloud maps under standard conditions. In actual application, only parameter adjustments and rapid calibration are required, significantly improving response speed.

[0131] Example 5: Specific implementation of the migration model unit in the risk diffusion prediction module

[0132] like Figure 4 As shown, the migration model unit 31 includes an atmospheric diffusion calculation subunit 311 , a water body diffusion calculation subunit 312 and a soil penetration calculation subunit 313 .

[0133] The atmospheric diffusion calculation subunit 311 uses the improved Gaussian plume model to calculate the diffusion concentration of hazardous chemicals in the atmosphere:

[0134] ,

[0135] in, Indicates time point The pollutant concentration at For the source of strength, and are the diffusion parameters in the horizontal and vertical directions, which are related to the atmospheric stability and the distance from the source. is the wind speed, is the effective source height, is the downwind distance, is the crosswind distance, is the vertical height.

[0136] The water diffusion calculation subunit 312 uses a two-dimensional convection-diffusion equation to calculate the migration process of hazardous chemicals in water bodies:

[0137] ,

[0138] in, is the pollutant concentration, For time, and They are and Flow velocity in the direction, and They are and The diffusion coefficient in the direction, is the degradation coefficient.

[0139] The soil permeability calculation subunit 313 simulates the permeability behavior of hazardous chemicals in soil based on the Richards equation:

[0140] ,

[0141] in, is the soil moisture content, is the soil water potential, is the unsaturated hydraulic conductivity function, is the vertical coordinate, is the source-sink term.

[0142] Optimally, atmospheric diffusion calculations account for topographic influences and building obstruction, water diffusion calculations consider river velocity variations and seasonal hydrological characteristics, and soil infiltration calculations consider the effects of varying soil types and groundwater levels. Each subunit utilizes adaptive meshing technology, with finer meshes used in key areas to improve calculation accuracy.

[0143] Example 6: Specific implementation of the road section risk assessment module

[0144] like Figure 5 As shown, the road segment risk assessment module 4 includes a road segment discretization unit 41 , a risk scoring unit 42 , a collaborative risk assessment unit 43 and a threshold adaptation unit 44 .

[0145] The road segment discretization unit 41 receives road network data and discretizes the road network into standard 500m sections. For complex road structures such as overpasses and tunnels, flexible segmentation is performed based on the actual structural characteristics to ensure that each section has relatively homogeneous characteristics.

[0146] The risk scoring unit 42 calculates the dynamic risk score RSI of each road segment at different time periods:

[0147] ,

[0148] in, is the comprehensive risk index of hazardous chemicals, For road sections In time Environmental Sensitivity Index To influence the amplification factor, In time Traffic conditions and weather conditions.

[0149] The collaborative risk assessment unit 43 assesses the collaborative risk factors of vehicle status, road characteristics, driving behavior patterns and hazardous chemical characteristics :

[0150] (vehicle status, road conditions, driving behavior patterns, hazardous chemicals characteristics),

[0151] Vehicle status includes parameters such as vehicle age, technical condition, load ratio, tire wear, and braking system response time; road condition characteristics include parameters such as road surface type, friction coefficient, slope, curve radius, and road surface condition; driving behavior patterns include parameters such as sudden acceleration frequency, sudden braking frequency, steering stability, and vehicle speed fluctuation.

[0152] The threshold adaptive unit 44 dynamically adjusts the safety threshold based on historical accident data:

[0153] ,in, is the learning rate, which is generally set to 0.05-0.1, TargetSafety is the target safety level, which is set by regulatory requirements, and CurrentSafety is the current safety level, which is calculated by the accident rate and approximate accident rate.

[0154] Preferably, the risk scoring unit 42 considers the time-varying nature of road segment characteristics, such as the impact of rain and snow on road friction and the impact of nighttime visibility. The collaborative risk assessment unit 43 uses fuzzy logic reasoning to address uncertainties and improve the robustness of the assessment. The threshold adaptive unit 44 sets upper and lower constraints to ensure that the safety threshold is adjusted within a reasonable range.

[0155] Example 7: Specific implementation of the risk prevention and control decision module

[0156] like Figure 6 As shown, the risk prevention and control decision module 5 includes an early warning judgment unit 51, a prevention and control measures unit 52, a multi-objective optimization unit 53 and a plan generation unit 54.

[0157] The warning determination unit 51 determines the warning level based on a comparison of the road segment risk score (RSI) with the safety threshold. If the road segment risk score (RSI) is below the first threshold (Threshold_1), the warning level is determined to be green (safe). If the RSI is between the first threshold (Threshold_1) and the second threshold (Threshold_2), the warning level is determined to be yellow (caution). If the RSI is between the second threshold (Threshold_2) and the third threshold (Threshold_3), the warning level is determined to be orange (warning). If the RSI exceeds the third threshold (Threshold_3), the warning level is determined to be red (dangerous).

[0158] Based on the warning level, the prevention and control measures unit 52 selects appropriate prevention and control measures from the prevention and control measures library. These measures include driver operation suggestions (such as deceleration, avoidance, and parking), vehicle parameter adjustments (such as maximum speed limit, maximum steering angle velocity, and brake pressure adjustment), and route planning changes (such as whether to recommend a route change, a list of alternative routes, and a recommended departure time adjustment).

[0159] The multi-objective optimization unit 53 considers the three objectives of safety, timeliness and economy to make a multi-objective optimization decision for path planning:

[0160] ,

[0161] in, is the safety target, reflecting the total risk of the path, For timeliness goals, reflect the transportation time, This is an economic objective that reflects transportation costs. Constraints include ensuring that the maximum road segment risk score does not exceed the permitted risk threshold, that the total transportation time does not exceed the maximum permitted transportation time, and that the total transportation cost does not exceed the budget cap.

[0162] The plan generation unit 54 generates emergency response plans for high-risk road sections. The plan generation process includes: identifying a set of high-risk road sections, extracting sensitive points in the surrounding environment of the road sections, predicting the impact range based on the risk propagation model, generating evacuation routes and safe areas, calculating the location of the nearest emergency resources and response time, and forming a structured plan document.

[0163] Preferably, the early warning determination unit 51 can adjust the early warning trigger threshold based on factors such as the type of transport vehicle and driver experience. The prevention and control measures unit 52 adopts a hierarchical progressive strategy, prioritizing mild intervention measures and escalating to mandatory measures when necessary. The multi-objective optimization unit 53 uses an improved NSGA-II algorithm to solve multi-objective optimization problems, improving solution efficiency. The emergency plan generation unit 54 interfaces with the emergency management department system to enable automatic push and response of emergency plans.

[0164] Example 8: Specific implementation of the multi-objective optimization unit

[0165] like Figure 6 As shown, the optimization function of the multi-objective optimization unit 53 is:

[0166] ,

[0167] in, is the safety target, and the calculation formula is:

[0168] ,

[0169] For the timeliness goal, the calculation formula is:

[0170] ,

[0171] For the economic goal, the calculation formula is:

[0172] ,

[0173] in, For road sections In time The risk score, For road sections length, For road sections In time The travel time, For road sections In time The waiting time, For road sections Fuel consumption rate For time cost, For road sections tolls, The number of road segments contained in the route.

[0174] The constraints are:

[0175] ,

[0176] ,

[0177] ,

[0178] in, .

[0179] The multi-objective optimization unit 53 uses an improved NSGA-II algorithm to solve the multi-objective optimization problem. The algorithm process includes: initializing the population, non-dominated sorting, calculating the crowding distance, selection, crossover, mutation, merging the populations, re-non-dominated sorting and crowding calculation, selecting the next generation of populations using an elite retention strategy, and iterating until convergence or the maximum number of iterations is reached.

[0180] To improve the efficiency of the algorithm, adaptive crossover and mutation operations are adopted to dynamically adjust the crossover rate and mutation rate according to the population diversity; a greedy strategy is used to generate the initial population to improve the quality of the initial solution; a local search strategy is used to optimize the non-dominated solution set to improve the quality of the solution.

[0181] Example 9: System data flow process

[0182] like Figure 1As shown in the figure, the data flow process between modules is as follows:

[0183] Risk Index Quantification Module 1 receives hazardous chemical attribute parameters, generates a multidimensional risk index matrix for hazardous chemicals, and transmits this matrix to Environmental Sensitivity Analysis Module 2 and Risk Diffusion Prediction Module 3. This matrix contains five risk indices and a comprehensive risk index for hazardous chemicals, stored in matrix form for rapid access and processing by subsequent modules.

[0184] Environmental Sensitivity Analysis Module 2 acquires environmental sensitivity data, generates a spatiotemporal environmental sensitivity prediction matrix, and transmits it to Risk Diffusion Prediction Module 3 and Road Section Risk Assessment Module 4. The spatiotemporal environmental sensitivity prediction matrix contains the environmental sensitivity index of each grid within the prediction period and is stored in a three-dimensional array (x-coordinate, y-coordinate, time), supporting efficient spatiotemporal queries.

[0185] Based on the multidimensional risk index matrix for hazardous chemicals and the spatiotemporal environmental sensitivity prediction matrix, the risk diffusion prediction module 3 generates a risk diffusion probability cloud map and transmits it to the road section risk assessment module 4. The risk diffusion probability cloud map is stored in a two-dimensional grid format, with each grid cell containing probability values ​​of varying degrees of impact, supporting spatial correlation analysis.

[0186] The road segment risk assessment module 4 combines the multidimensional hazardous chemicals risk index matrix, the spatiotemporal environmental sensitivity prediction matrix, and the risk propagation probability cloud map to generate a spatiotemporal road network risk score matrix and transmit it to the risk prevention and control decision module 5. The spatiotemporal road network risk score matrix contains the risk scores of each road segment at different time periods and is stored in a two-dimensional table indexed by the road segment ID and time.

[0187] Based on the spatiotemporal road network risk scoring matrix, the risk prevention and control decision module 5 generates a risk intelligent decision-making instruction set and transmits it to the vehicle terminal and monitoring center via the communication network. The risk intelligent decision-making instruction set uses a structured JSON format and contains warning information, prevention and control recommendations, route optimization solutions, and emergency response plans.

[0188] The data flow utilizes a distributed event-driven architecture, supporting real-time data processing and push. Caching mechanisms are implemented at key data nodes to reduce computational latency. Data transmission utilizes encrypted channels to ensure data security. The entire system forms a closed-loop feedback loop, with the output of each module serving as the input for other modules, enabling collaboration and improving the safety of hazardous chemical transportation.

[0189] Example 10: Dynamic risk prevention and control method for hazardous chemicals transportation routes

[0190] like Figure 7 As shown, the method for dynamic risk prevention and control of hazardous chemicals transportation routes of the present invention includes the following steps:

[0191] First, the system receives hazardous chemical attribute parameters and calculates a multidimensional risk index matrix. The system obtains 15 attribute parameters, including the physical state, chemical activity, and toxicity level of hazardous chemicals, through electronic waybills or manual input. After standardization, it calculates five risk indices and a comprehensive risk index to form a multidimensional risk index matrix.

[0192] Secondly, environmental sensitivity data is acquired, a gridded spatiotemporal fusion model is constructed, and a spatiotemporal environmental sensitivity prediction matrix is ​​generated. The system uses interfaces to access real-time meteorological data, geographic information data, population distribution data, and sensitive target data. It then divides the spatial area into standard grid cells, calculates the environmental sensitivity index for each spatiotemporal grid cell, and uses a Bayesian network to predict future environmental sensitivity changes.

[0193] Then, based on the multidimensional risk index matrix of hazardous chemicals and the spatiotemporal environmental sensitivity prediction matrix, a multi-media migration model was constructed, and Monte Carlo risk propagation simulations were conducted to generate a risk propagation probability cloud map. The system established migration and diffusion models for the atmosphere, water, and soil, pre-set various leakage scenarios, and conducted multiple Monte Carlo simulations to generate a visual risk probability distribution.

[0194] Next, the road network is discretized into standard sections. A spatiotemporal network risk score matrix is ​​calculated by combining a multidimensional hazardous chemical risk index matrix, a spatiotemporal environmental sensitivity prediction matrix, and a risk propagation probability cloud map. The system then calculates a dynamic risk score for each section at different times, taking into account the synergistic risk factors of vehicle status, road conditions, and driving behavior, and dynamically adjusts safety thresholds based on historical accident data.

[0195] Finally, based on a comparison of the spatiotemporal road network risk score matrix and safety thresholds, the system determines the warning level, selects prevention and control measures, executes multi-objective optimization decisions to generate the optimal transportation route, and outputs a set of risk-based intelligent decision-making instructions. Based on the risk score, the system determines four warning levels: green, yellow, orange, or red. It then selects corresponding prevention and control measures, optimizes routes based on safety, timeliness, and economy, and generates emergency response plans for high-risk sections.

[0196] The entire approach forms a closed-loop control process, mitigating safety risks during hazardous chemical transportation through real-time data collection, dynamic risk assessment, and proactive prevention and control measures. The approach utilizes a modular design, allowing each step to be flexibly configured based on actual needs, adapting to hazardous chemical transportation scenarios of varying scales and types.

[0197] To verify the effectiveness of this invention, a chemical company's liquid chlorine transportation task was selected as an application case. The company needed to transport 15 tons of liquid chlorine from its production plant to a user factory 300 kilometers away. The transportation route required passing through multiple towns and water source protection areas.

[0198] First, the system receives the properties of liquid chlorine: physical state: liquefied gas (9 points), chemical activity: 8 points, toxicity: 9 points, explosiveness: 5 points, flash point: not applicable, vapor pressure: 8 points, water solubility: 7 points, corrosiveness: 9 points, reactivity: 8 points, stability: 6 points, volatility: 9 points, carcinogenicity: no, teratogenicity: no, bioaccumulation: 3 points, and environmental persistence: 5 points. The system calculates the leakage risk index of liquid chlorine to be 8.5, the fire and explosion risk index to be 5.5, the health hazard risk index to be 9.0, the environmental pollution risk index to be 6.8, the reaction hazard risk index to be 8.2, and the overall risk index to be 7.6.

[0199] Next, the system acquires environmental sensitivity data surrounding the transportation route and constructs a grid-based spatiotemporal fusion model. Analysis revealed that the transportation route passes through two densely populated areas (with peak traffic between 8:00 AM and 9:00 AM, and 5:00 PM and 6:00 PM, respectively) and a key water source protection area. The system then predicts changes in environmental sensitivity for each grid cell over the next 24 hours, generating a spatiotemporal environmental sensitivity prediction matrix.

[0200] The system then constructed a multi-media migration model based on the characteristics of liquid chlorine and environmental conditions. It then performed 10,000 Monte Carlo simulations based on a pre-set medium-scale leak scenario (500 kg, 30 kg / min leak rate). The results showed that under adverse meteorological conditions, a liquid chlorine leak could affect an area within a 3.2-kilometer radius, potentially impacting 21,000 people.

[0201] The system then discretized the transport routes into 604 standard sections. Combining the liquid chlorine risk index, environmental sensitivity, and risk transmission probability, it calculated a dynamic risk score for each section at different times. The analysis revealed that the risk scores of 28 sections exceeded safety thresholds during specific periods, with 12 of these sections receiving a red alert level.

[0202] Finally, the system uses a risk-scoring matrix to perform multi-objective optimization and generate the optimal transport route. This optimized route avoids peak hours in densely populated areas, increasing the distance by 27 kilometers (a 9% increase), but reduces exposure time on high-risk sections by 62% and the potential number of people affected by 78%. The system also generates emergency response plans for high-risk sections, including spill management, evacuation measures, and environmental protection measures, which are shared in advance with emergency response departments along the route.

[0203] like Figure 8 As shown, the traditional method focuses primarily on transportation distance and time, while ignoring risk factors. The method of the present invention significantly reduces safety risks while appropriately increasing transportation time and costs. Practical applications have shown that this system can effectively improve the safety level of hazardous chemical transportation and reduce the incidence and potential impact of accidents.

[0204] While embodiments of the present invention have been shown and described, it will be apparent to those skilled in the art that various changes, modifications, substitutions, and variations may be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the claims and their equivalents.

[0205] The foregoing description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. Those skilled in the art will readily appreciate that various modifications and variations of the present invention are possible. Any modifications, equivalent substitutions, or improvements made within the spirit and principles of the present invention are intended to be within the scope of protection of the present invention.

Claims

1. A dynamic risk prevention and control system for hazardous chemicals transportation routes based on time-space fusion, characterized by: include: Risk index quantification module, used for: Receive hazardous chemical property parameters, including physical state, chemical activity, toxicity level, explosiveness, flash point, vapor pressure, water solubility, corrosivity, reactivity, stability, volatility, carcinogenicity, teratogenicity, bioaccumulation, and environmental persistence; Based on the hazardous chemical property parameters, calculate the leakage risk index, fire and explosion risk index, health hazard risk index, environmental pollution risk index and reaction hazard risk index; Generate a multi-dimensional risk index matrix for hazardous chemicals; The environmental sensitivity analysis module is connected to the risk index quantification module and is used to: Acquiring environmentally sensitive data, including real-time meteorological data, geographic information data, population distribution data, and sensitive target data; Construct a gridded space-time fusion model and calculate the environmental sensitivity index for each space-time grid; A Bayesian spatiotemporal network prediction model is established based on historical data to predict the changes in environmental sensitivity of each grid in the future period; Generate spatiotemporal environmental sensitivity prediction matrix; The risk diffusion prediction module is connected to the risk index quantification module and the environmental sensitivity analysis module and is used to: Constructing a multi-media migration model based on the hazardous chemicals multidimensional risk index matrix and the spatiotemporal environmental sensitivity prediction matrix; Preset multiple leakage scenarios and conduct Monte Carlo risk propagation simulations; Generate risk propagation probability cloud map; The road section risk assessment module is connected to the risk diffusion prediction module and is used to: Receive road network data and discretize it into standard road segments; Calculate the dynamic risk score of each road section at different time periods by combining the multi-dimensional risk index matrix of hazardous chemicals, the spatiotemporal environmental sensitivity prediction matrix, and the risk propagation probability cloud map; Establishing a collaborative risk factor assessment model; the collaborative risk factor assessment model is used to assess collaborative risk factors of vehicle status, road condition characteristics, driving behavior patterns, and hazardous chemical characteristics; Dynamically adjust safety thresholds based on historical accident data; Based on the dynamic risk score of each road section at different time periods and the collaborative risk factor evaluation results, a spatiotemporal road network risk score matrix is ​​generated with the road section ID and time as the index; The risk prevention and control decision module is connected to the road section risk assessment module and is used to: Determining a warning level based on a comparison between the spatiotemporal road network risk score matrix and a safety threshold; According to different warning levels, select corresponding prevention and control measures from the prevention and control measures library; Execute multi-objective optimization decision to generate the optimal transportation route; Generate emergency response plans for high-risk road sections; Output risk intelligent decision-making instruction set.

2. The system according to claim 1, wherein: The risk index quantification module includes: Parameter characterization unit, used to standardize hazardous chemical property parameters into a scale of 0-10; The risk calculation unit is used to calculate five types of risk indices based on standardized hazardous chemical property parameters; Weight adjustment unit, used to dynamically adjust the weight coefficient according to different scenarios and calculate the comprehensive risk index : , in, is the leakage risk index, is the fire and explosion risk index, is the health hazard risk index, is the environmental pollution risk index, To reflect the hazard risk index, to is the weight coefficient, and .

3. The system according to claim 1, wherein: The environmental sensitivity analysis module includes: Data access unit, used to acquire and preprocess multi-source environmental data; Grid division unit, used to construct a standard grid unit of 100m×100m and establish a 24-hour×7-day time matrix on each grid; Sensitivity calculation unit, used to calculate the environmental sensitivity index ESI of each spatiotemporal grid: ， in, For Grid In time The population density index, For Grid The water sensitivity index, For Grid Special sensitive target index, For Grid In time The weather condition index, is the weight coefficient and 1; The prediction modeling unit is used to build a Bayesian network based on historical data to predict the changes in the environmental sensitivity of each grid within the next 24 hours.

4. The system according to claim 1, wherein: The risk diffusion prediction module includes: Migration model unit, used to establish migration and diffusion models of three media: atmosphere, water and soil; Scenario preset unit, used to preset various leakage scenarios including small leakage, medium leakage, large leakage, fire-associated leakage and explosion-associated leakage according to the characteristics of hazardous chemicals and environmental conditions; The Monte Carlo simulation unit is used to perform 10,000 Monte Carlo simulations for each leakage scenario and generate a risk probability distribution cloud map.

5. The system according to claim 4, characterized in that The migration model unit includes: The atmospheric diffusion calculation subunit uses an improved Gaussian plume model to calculate the diffusion concentration of hazardous chemicals in the atmosphere; The water diffusion calculation subunit uses the two-dimensional convection-diffusion equation to calculate the migration process of hazardous chemicals in water bodies; The soil penetration calculation subunit simulates the penetration behavior of hazardous chemicals in soil based on the Richards equation.

6. The system according to claim 1, wherein: The road section risk assessment module includes: The road segment discretization unit is used to discretize the road network into standard 500m length sections; The risk scoring unit is used to calculate the dynamic risk score RSI of each road segment at different time periods: ， in, is the comprehensive risk index of hazardous chemicals, For road sections In time The environmental sensitivity index, is the impact amplification factor; Collaborative risk assessment unit, used to evaluate the collaborative risk factors of vehicle status, road conditions, driving behavior patterns and hazardous chemicals characteristics; Threshold adaptive unit, used to dynamically adjust safety thresholds based on historical accident data: ,in, is the learning rate, TargetSafety is the target safety level, and CurrentSafety is the current safety level.

7. The system according to claim 1, wherein: The risk prevention and control decision module includes: The warning determination unit is used to determine the four warning levels of green, yellow, orange or red based on the comparison of the road section risk score and the safety threshold; The prevention and control measures unit is used to select corresponding driver operation suggestions, vehicle parameter adjustments and route planning changes based on the warning level; Multi-objective optimization unit, used to make multi-objective optimization decisions for path planning by considering the three objectives of safety, timeliness, and economy; The plan generation unit is used to generate emergency response plans for high-risk road sections.

8. The system according to claim 7, characterized in that The optimization function of the multi-objective optimization unit is: , in, is the safety target, reflecting the total risk of the path, For timeliness goals, reflect the transportation time, For economic purposes, it reflects transportation costs; The constraints are: Maximum road section risk score Allowable risk threshold, Total transportation time Maximum permissible transit time, Total shipping cost Budget cap.

9. The system according to claim 1, wherein: The data flow process between modules is as follows: The multi-dimensional risk index matrix of hazardous chemicals generated by the risk index quantification module is transmitted to the environmental sensitivity analysis module and the risk diffusion prediction module; The spatiotemporal environmental sensitivity prediction matrix generated by the environmental sensitivity analysis module is transmitted to the risk diffusion prediction module and the road section risk assessment module; The risk propagation probability cloud map generated by the risk diffusion prediction module is transmitted to the road section risk assessment module; The spatiotemporal road network risk scoring matrix generated by the road section risk assessment module is transmitted to the risk prevention and control decision module; The risk intelligent decision instruction set of the risk prevention and control decision module is transmitted to the vehicle terminal and the monitoring center through the communication network.

10. The method corresponding to the dynamic risk prevention and control system for hazardous chemical transportation routes based on time-space fusion according to any one of claims 1 to 9, characterized in that: The following steps are involved: Receive hazardous chemical attribute parameters and calculate the multi-dimensional risk index matrix of hazardous chemicals; Acquire environmental sensitive data, build a gridded spatiotemporal fusion model, and generate a spatiotemporal environmental sensitivity prediction matrix; Based on the multi-dimensional risk index matrix of hazardous chemicals and the spatiotemporal environmental sensitivity prediction matrix, a multi-media migration model is constructed, a Monte Carlo risk propagation simulation is performed, and a risk propagation probability cloud map is generated; Discrete the road network into standard road sections, and calculate the spatiotemporal road network risk score matrix by combining the multidimensional risk index matrix of hazardous chemicals, the spatiotemporal environmental sensitivity prediction matrix, and the risk propagation probability cloud map; Based on the comparison between the spatiotemporal road network risk scoring matrix and the safety threshold, the warning level is determined, the prevention and control measures are selected, the multi-objective optimization decision is executed to generate the optimal transportation route, and the risk intelligent decision instruction set is output.

Images