Dangerous goods transportation charging method based on risk assessment

Abstract

The invention discloses a dangerous goods transportation charging method based on risk assessment, and relates to the technical field of dangerous goods transportation charging, and the method comprises the following steps: collecting the data of traffic flow, accident occurrence frequency, environmental factors and the like of each road segment in a dangerous goods transportation network, and calculating a dangerous goods transportation charging result based on a population coverage risk model and an accident expected value calculation method; and evaluating the potential risk level of each road section. Through multi-dimensional data fusion and an advanced algorithm, real-time assessment and path optimization of transportation network risks are realized. A population coverage risk model and an accident expected value calculation method are adopted, high-risk road sections are accurately recognized, transportation tasks are reasonably distributed, and safety is improved. Meanwhile, the genetic algorithm is utilized to optimize the transportation path, balance the cost and risk, reduce the transportation cost and guarantee the safety. Global path distribution is optimized through a simulated annealing algorithm, excessive load of high-risk road sections is avoided, task distribution is dynamically adjusted according to risk prediction, reasonable resource allocation is ensured, and transportation efficiency and safety are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of dangerous goods transportation charges, and in particular to a dangerous goods transportation charging method based on risk assessment. Background Art

[0002] With the rapid development of modern industry, the transportation volume and demand of dangerous goods have increased year by year. In particular, road transportation, as the main mode of dangerous goods transportation, has become an indispensable part of social economy and industrial production. However, the safety of dangerous goods transportation has always been a difficult problem to be solved. Therefore, how to effectively reduce the risks in the transportation of dangerous goods and ensure the safety of road transportation has become an important issue to be solved.

[0003] At present, there have been certain studies and policies at home and abroad on the risk control of dangerous goods transportation. Traditional control strategies mainly rely on road bans, capacity restrictions and other means to control the risks of dangerous goods transportation. However, these traditional strategies often bring a series of problems. The ban policy will cause a large number of roads to be eliminated, affecting traffic smoothness and increasing transportation costs. It will also lead to excessively high regulatory costs and difficult to sustain in the long run. Although capacity restrictions can avoid overloading of certain sections of roads, they rely on fixed capacity settings and are difficult to adjust flexibly according to actual conditions. Therefore, they cannot fully respond to risks under different roads and different transportation conditions. In addition, the existing dangerous goods transportation route planning usually focuses on minimizing transportation costs or shortest time, while ignoring the comprehensive consideration of risks and lacking a flexible charging mechanism to regulate transportation behavior.

[0004] In response to these problems, in recent years, some studies have begun to try to introduce charging mechanisms into the design of dangerous goods transportation networks, in order to reduce the overall risk of the road network by charging specific road sections. For example, Marcotte P and Mercier A proposed a network design model based on congestion charging, and Wang J and Kang Y further proposed a dual charging model to reduce the risks of ordinary goods and dangerous goods transportation at the same time. Although these studies have proposed charging strategies, there are still some shortcomings. Existing charging strategy research mainly focuses on the comparison of different charging mechanisms with other strategies, and few scholars have deeply explored how to differentiate charging based on the risk characteristics of different roads, traffic flow and other factors. More importantly, the risk measurement in current research mostly depends on the probability and impact of accidents, and lacks an in-depth analysis of risk fairness (i.e., the fairness of risk distribution between different road sections). Therefore, how to formulate a scientific and reasonable charging strategy based on the risk characteristics of different road sections is still a blank area in the design of dangerous goods transportation networks.

[0005] The above information disclosed in this Background section is only for enhancement of understanding of the background of the present disclosure and therefore it may contain information that does not constitute the prior art that is already known to one of ordinary skill in the art. Summary of the invention

[0006] The purpose of the present invention is to provide a dangerous goods transportation charging method based on risk assessment, combining multidimensional data with advanced algorithms, and reasonably allocating transportation tasks through real-time evaluation of the risk level of each section in the transportation network to avoid excessive load on high-risk sections, thereby improving transportation safety. Genetic algorithms are used to optimize transportation routes, while reducing costs, taking into account the risks of each section to achieve a balance between cost and risk. At the same time, the system dynamically adjusts the charging standards to ensure that enterprises can flexibly respond to real-time changing risks. The transportation network is globally optimized through a simulated annealing algorithm, tasks are reasonably scheduled, section risks are balanced, excessive traffic pressure on high-risk sections is avoided, transportation efficiency and safety are improved, and reasonable allocation and efficient scheduling of resources are ensured to solve the problems in the above-mentioned background technology.

[0007] In order to achieve the above object, the present invention provides the following technical solution: a dangerous goods transportation charging method based on risk assessment, comprising the following steps:

[0008] By collecting data on traffic flow, accident frequency, environmental factors, etc. on each section of the dangerous goods transportation network, the potential risk level of each section is evaluated based on the population coverage risk model and accident expectation value calculation method;

[0009] Based on the risk assessment results, the road sections in the transportation network are divided into risk levels, and segmented charges are implemented for the road sections according to different risk levels. The risk level of each road section is clustered through the K-means clustering algorithm, and different charging standards are set based on the clustering results;

[0010] According to the given transportation starting point and end point, the transportation route is optimized by using genetic algorithm to minimize the transportation cost and reduce the total risk in the transportation process. The genetic algorithm optimizes the route selection and balances the cost and risk by simulating the natural selection process.

[0011] According to the risk level differences of transportation routes, a risk balance allocation strategy is implemented to avoid excessive burden of transportation tasks on high-risk sections. The global transportation network is searched through the simulated annealing algorithm to adjust the path distribution to achieve risk balance among different sections.

[0012] Based on real-time monitoring data, the charging standards of each section in the transportation network are dynamically adjusted. Machine learning algorithms are used to analyze historical transportation data, predict future risks of sections, and adjust charging strategies based on the prediction results to ensure that risks match charges.

[0013] Preferably, based on the population coverage risk model and the accident expectation value calculation method, the potential risk level of each road section is evaluated. The specific steps are as follows:

[0014] Collect relevant data from each section of the dangerous goods transportation network, including traffic flow, accident frequency, and environmental factors (such as climate, topography, population density, etc.). These data can be obtained through sensors, historical accident records, meteorological data interfaces, and traffic monitoring systems. The collected data must be preprocessed to ensure its accuracy and completeness. For example, for traffic flow data, outliers and noise need to be removed; for accident frequency, the data needs to be normalized for subsequent calculations;

[0015] The purpose of this step is to provide accurate basic data for subsequent risk assessment. Traffic flow and accident frequency are important factors affecting transportation risk, while environmental factors can reflect the additional risks that may be brought by certain special sections of road (such as mountainous areas or densely populated areas). By processing these data, we ensure that the input data for risk assessment is of high quality.

[0016] After collecting and preprocessing the data, we need to define a comprehensive risk assessment function that combines the population coverage risk model and the accident expectation value calculation method to assess the risk level of each road section. The risk level R i It is expressed as:

[0017] R i =α·T i +β·A i +γ·E i

[0018] , where: T i is the traffic flow index of road section i, reflecting the busyness of the road section. i is the accident frequency index of road section i, indicating the probability of past accidents, E i is the environmental risk factor of road section i, including factors such as geography, climate and population density. α, β and γ are weight coefficients, which are set according to actual conditions and represent the degree of influence of each factor on the risk.

[0019] By weighting and integrating various influencing factors, the comprehensive risk level R of each road section is calculated. i , providing a quantitative risk assessment standard. This function can adapt to the differences between different data sources and regions through weighted coefficients, making it easy to adjust and optimize risk assessment results according to actual conditions.

[0020] In order to more accurately quantify the potential risks in the transportation process, the accident expectation value is used to measure the possible accident losses in each section. The accident expectation value can be expressed as:

[0021]

[0022] , where: P(t) represents the probability of an accident occurring at time t, C(t) is the loss at the time of the accident (including casualties, property losses, etc.), and T is the period of the entire transportation time;

[0023] Through this step, the possible accident losses during transportation can be quantified for each road section, reflecting the potential safety hazards that different road sections may bring during transportation. This method combines the timeliness of the accident and the severity of the accident to more comprehensively evaluate potential risks.

[0024] In the previous step, the risk level R of each road section has been calculated. i and the expected value of accidents (EAL) i Next, we divide the road sections based on these evaluation results. Assuming that we classify the road sections according to the comprehensive risk, we can get the comprehensive risk classification of each road section using the following formula:

[0025]

[0026] , where C i is the comprehensive risk classification of road section i. By weighted average of risk level and accident expectation value, the comprehensive safety risk level of each road section can be reflected. By this method, the road sections in the transportation network can be classified according to risk level, for example: low risk (C i ≤0.3), medium risk (0.3 <C i ≤0.7), high risk (C i >0.7);

[0027] This step classifies the road sections based on the comprehensive risk assessment results, providing a basis for the subsequent toll collection strategy formulation. i , corresponding risk control measures can be formulated for each road section, thereby achieving differentiated charging.

[0028] Preferably, the risk level of each road section is clustered by a K-means clustering algorithm, and different charging standards are set based on the clustering results. The specific steps are as follows:

[0029] Before implementing the K-means clustering algorithm, it is necessary to first prepare and preprocess the data of the road segments in the transportation network. This includes collecting the risk level of each road segment (such as the risk level R calculated in the previous step) according to the results of the risk assessment. i or Comprehensive Risk Category C i), these data should cover all road sections that require toll assessment. In addition, the accuracy, completeness and consistency of the data need to be ensured. Outliers, missing values ​​and noise in the data need to be processed. For example, for abnormal data in the risk assessment results, data cleaning techniques such as median filling or deletion can be used to remove abnormal effects and ensure data stability during the K-means clustering process. The preprocessed data should be normalized to avoid the influence of data of different dimensions (such as traffic, accident frequency, and environmental factors) on the clustering algorithm. In this process, a commonly used normalization method is to normalize the data to the same interval range (such as 0 to 1) so that the clustering algorithm can process different types of data equally.

[0030] After the data is prepared, the K-means clustering algorithm is used to cluster the risk level of each road section. The core of the K-means algorithm is to divide the data set into a predetermined number of clusters. The risk level of the road sections in each cluster should be as similar as possible, and the difference in risk level between different clusters should be maximized. First, select a suitable K value (number of clusters), that is, the number of risk level groups divided. This K value can be selected by methods such as the Elbow Method or the Silhouette Coefficient. The Elbow Method plots the sum of squared errors (SSE) corresponding to different K values ​​and selects the point where the SSE drops sharply as the appropriate K value. In the K-means algorithm, the risk level data of each road section will be classified into the closest cluster according to its distance from the cluster center. Through multiple iterative calculations, the cluster center is continuously updated until the clustering results converge. The result after clustering is the risk category corresponding to each road section.

[0031] After clustering is completed, different charging standards will be set for each cluster according to its characteristics. Each cluster represents a section with similar risk levels, and these clusters can be divided into different risk levels, such as low risk, medium risk and high risk. The specific charging standards can be determined according to the risk level: for example, high-risk sections may need to set higher charging standards to compensate for higher transportation risks and potential accident losses; while low-risk sections set lower charging standards. In addition, the charging level of each risk level can also be flexibly adjusted according to the specific market environment and policy requirements, combined with factors such as cost and operational risk. At this time, sections of different risk levels will be charged in sections according to the preset charging standards, thereby realizing a risk-based dynamic charging strategy. This segmented charging can not only better reflect the risk differences of sections, but also provide a fairer and more flexible fee calculation method for the transportation of dangerous goods.

[0032] After the initial setting of the charging standards, the charging strategy needs to be further optimized through dynamic monitoring and data feedback. During the transportation process, the risk level of the road section may fluctuate with changes in factors such as road conditions, weather changes, and accident frequency. Therefore, the charging strategy needs to be evaluated and adjusted regularly. For example, a periodic review mechanism (such as quarterly or annually) can be set up to adjust the charging standards in a timely manner through re-risk assessment and cluster analysis. If the risk level of a section of road changes (for example, from medium risk to high risk), the charging standard of the section may need to be increased to reflect the new risk level; conversely, if the risk level of a section of road decreases, the charge can be appropriately reduced. In addition, with the gradual accumulation of transportation data and accident data, the K value of the K-means algorithm, the clustering results, and the charging standards can also be further optimized based on the new data, making the charging strategy more accurate, flexible, and fair.

[0033] Through these four steps, the risk level of each section in the transportation network is divided based on the K-means clustering algorithm, and differentiated segmented charging standards are formulated based on the risk level. The data preparation and preprocessing in step one ensure the accuracy of the subsequent clustering analysis. Step two uses the K-means clustering algorithm to achieve a scientific division of the risks of each section. Step three formulates a reasonable charging standard based on the clustering results, and step four ensures the continuous optimization of the charging strategy through a dynamic adjustment mechanism. This risk-based charging method can not only adjust the charging standards according to real-time changes, but also improve the safety and rationality of dangerous goods transportation, and ultimately achieve a balance between risk and charging.

[0034] Preferably, according to the given transportation starting point and end point, a genetic algorithm is used to optimize the transportation route to minimize the transportation cost and reduce the total risk during the transportation process. The specific steps are as follows:

[0035] In the preliminary step of the genetic algorithm, the transport path needs to be encoded first. Each transport path can be regarded as an individual (chromosome), which consists of a series of road segments, each of which corresponds to a road segment in the transportation network. Assume that the starting point of the transportation is O and the end point is D, and each road segment of the transport path is composed of nodes in the network (such as the starting point and end point of the road segment). For the convenience of representation, the transport path can be encoded as a vector P = (p i )=(p 1 , p 2 ,…,p n ), where p i represents the ID of the i-th road segment, and n is the number of road segments contained in the path;

[0036] Next, we randomly generate the initial population P 0 = {P 1 , P2 ,…,P m}, where each individual p i is a possible transportation path, and the path length is n, and the population size is m (usually m takes a larger value, such as 100 or 200). When initializing the population, considering the diversity of the paths, each path in the population can be a different path from the starting point to the end point, and each section in the path is connected by random selection. The initialized population provides a diverse initial solution space for the algorithm.

[0037] In genetic algorithms, the fitness function is the key to measuring the quality of individuals. For the optimization problem of transportation routes, the goal is to minimize the sum of transportation costs and risks during transportation. Therefore, the fitness function f(P) can be expressed as the weighted sum of cost and risk:

[0038] f(P)=α′·C(P)+β′·R(P)

[0039] , where C(P) is the transportation cost of path P, R(P) is the sum of the risks of all sections on path P, and α′ and β′ are coefficients for adjusting the weights between cost and risk. In order to better adapt to the actual situation, these coefficients can be set according to specific needs. For example, if the risk of transporting dangerous goods is high, the weight of β′ can be increased.

[0040] The specific formula for calculating cost and risk can be defined as follows:

[0041] The transportation cost C(P) is the sum of the transportation costs of each road segment:

[0042]

[0043] , where c(p i ) is the road segment p i The transportation cost may be determined by factors such as traffic volume, section length, and vehicle type;

[0044] The transport risk R(P) is the sum of the risks of each road segment, which can be assessed based on the traffic flow, accident frequency and environmental factors of each road segment:

[0045]

[0046] , where r(p i ) is the road segment p i The risk value may involve traffic accident probability, environmental risk factor, etc.

[0047] After fitness evaluation, the selection operation selects better individuals through the process of natural selection. The selection operation usually uses methods such as Roulette Wheel Selection or Tournament Selection. According to the fitness value of the individual, the more suitable individuals are selected as the parents of the next generation. Individuals with higher fitness have a greater probability of being selected, thereby ensuring that the better path solution is passed on to the next generation.

[0048] After selecting the parent generation, the genetic algorithm generates a new generation of individuals through crossover and mutation operations. The crossover operation simulates the genetic recombination in nature and generates a new child path by exchanging sections of two parent paths. In the optimization problem of transportation paths, you can choose to cross some paths, such as exchanging paths through single-point crossover or multi-point crossover.

[0049] Single-point crossover: Assume there are two parent paths P a and P b , the crossover operation splits the path at a certain position k, and P a The front part of P b The rear part and P b The front part of P a The latter part of the crossover is exchanged to obtain a new child path. After the crossover, a local optimal solution may be obtained. Therefore, in order to avoid falling into the local optimal solution, the diversity of solutions must be further improved through mutation operations. Mutation operations usually make random adjustments to individuals, such as changing a section of the path. For path P = (p i )=(p 1 , p 2 ,…,p n ) The mutation operation can randomly select a road segment p i , replace it with another segment p that does not appear j , thereby changing the structure of the path. The mutation operation helps the algorithm escape from the local optimal solution and improve the global search capability.

[0050] By repeatedly performing crossover and mutation operations, the genetic algorithm continuously optimizes the selection of routes and eventually finds the optimal route that minimizes the sum of transportation costs and transportation risks. In each generation, the fitness value of the newly generated route is calculated, and selection, crossover and mutation operations are performed based on the fitness, thereby optimizing the route selection and balancing costs and risks. Ultimately, through enough iterations, an optimal or approximately optimal transportation route is found to solve the optimization problem of transportation costs and risks.

[0051] Preferably, the global transportation network is searched by a simulated annealing algorithm to adjust the path distribution. The specific steps are as follows:

[0052] Before implementing the risk balance allocation strategy, it is necessary to first perform a comprehensive state initialization of the transportation network, including risk level assessment of each road section. The risk level of each road section can be comprehensively assessed based on multiple indicators such as traffic flow, accident frequency, environmental factors, etc. to obtain a preliminary risk value. These assessment results provide basic data for subsequent optimization. At this stage, it is necessary to collect and store risk data for each road section to determine the transportation capacity and safety risk of each road section.

[0053] At this point, the choice of transport path is not fixed, but is reasonably configured based on the risk information in the current network. Through a preliminary analysis of the risk differences of each section, the differences between some high-risk sections and low-risk sections can be found, providing data support for subsequent path optimization and risk balance distribution. The core of the network initialization phase is to ensure that the risk assessment of each section can accurately reflect the actual situation through comprehensive data collection.

[0054] Before optimizing the simulated annealing algorithm, it is necessary to define the objective function. The core of the objective function is to achieve a balanced risk distribution of transportation tasks and avoid excessive transportation tasks on certain high-risk sections. The objective function can be designed as: maximizing the risk balance of the entire transportation network, that is, increasing the transportation tasks on low-risk sections, thereby avoiding excessive burdens on high-risk sections.

[0055] The main idea of ​​the simulated annealing algorithm is to start from an initial state and gradually find a low-energy state by continuously adjusting the path distribution, so that the transportation tasks can be reasonably distributed throughout the network. To achieve this goal, the "temperature" parameter in the annealing process needs to be gradually reduced to ensure the stability of the search process and prevent the generation of local optimality. Through a reasonable temperature decay strategy, it is ensured that the global search can be fully explored until the optimal path distribution that balances the risk is finally found.

[0056] In the simulated annealing search process, path adjustment is the core step of the algorithm. The simulated annealing algorithm continuously makes local adjustments to the existing path to find a better path distribution. Specifically, the simulated annealing algorithm will select a section of a path for random changes and calculate the risk level of the adjusted path. If the new path configuration can improve the overall risk balance, the algorithm will accept this change; if the risk is not improved, the algorithm will accept a worse solution based on a certain probability to avoid falling into the local optimal solution.

[0057] During the path search process, adjustments can be made by changing the path selection, for example, replacing certain high-risk sections with low-risk sections, or adjusting the transportation task volume of each path. Through these local adjustments, the simulated annealing algorithm can gradually improve the path configuration to achieve the goal of global risk balance. The key to this step is how to ensure that the path selection balances the task burden between high-risk and low-risk sections as much as possible through appropriate adjustment strategies.

[0058] The iterative process of the simulated annealing algorithm is the key to achieving global search and finding the optimal path distribution. As the algorithm proceeds, the temperature will gradually decrease, which means that the probability of accepting a bad solution will gradually decrease, making the algorithm's search more stable. In each round of iteration, a new risk balance will be calculated based on the current path adjustment situation, and this value will be used to determine whether to continue path adjustment.

[0059] As the temperature decreases, the algorithm will tend to local optimization and gradually find the optimal solution. Therefore, in this process, in addition to path adjustment, it is also necessary to control the depth and breadth of the search according to the set temperature decay strategy. For example, in the initial stage, a larger temperature can be selected to conduct a wide range of path searches, while in the later stage, the temperature can be gradually reduced to focus on local optimization. Eventually, the algorithm will stop near a certain temperature value, thereby obtaining a relatively stable path distribution, ensuring that the risks of different sections can be evenly distributed.

[0060] Through the above steps, the simulated annealing algorithm can effectively adjust the path distribution in the global transportation network, realize the reasonable allocation of transportation tasks, avoid over-reliance on high-risk sections, and ensure the safety and stability of the entire transportation process.

[0061] Preferably, a machine learning algorithm is used to analyze historical transportation data, predict future risks of road sections, and adjust charging strategies based on the prediction results to ensure that risks match charges. The specific steps are as follows:

[0062] Before dynamically adjusting the charging standards for each section of the transportation network, the first step is to ensure the comprehensive collection and preprocessing of real-time monitoring data. These data mainly include traffic flow, accident rate, environmental factors (such as weather, road conditions, etc.), type and speed of transport vehicles, etc. The operational data of each section is collected in real time through sensors, GPS equipment, surveillance cameras and other monitoring means. After the data is collected, it must be preprocessed to remove noise, fill in missing values, and perform normalization or standardization to ensure that the data quality meets the requirements of the machine learning model. The purpose of the preprocessing stage is to ensure that the data can effectively reflect the real-time status of the current transportation network and provide accurate input for subsequent predictions and decisions.

[0063] In the process of data preprocessing, historical data also needs to be sorted out, including past traffic flow, accident records, weather data, etc., and key features analyzed through data mining technology to identify potential patterns and laws. The analysis of these historical data will help to establish a prediction model based on historical experience and provide a basis for the dynamic adjustment of charging standards. Through this comprehensive data collection and processing, it is ensured that the status information of the current transportation network can be obtained in a timely and accurate manner, providing accurate risk prediction for subsequent adjustments to charging strategies.

[0064] Based on the pre-processed real-time data and historical transportation data, the next step is to use machine learning algorithms to predict risks. By analyzing the trends and features in historical data, machine learning models can predict the future risk level of each road section. Common machine learning algorithms, such as regression analysis, support vector machine (SVM), decision tree, random forest, and neural network, can be used for risk prediction.

[0065] At this stage, the model will use multiple input features such as traffic flow, accident rate, weather, environmental factors, etc., and build a model that can predict the risks of each road section in the future through training and learning. By continuously optimizing model parameters and selecting appropriate feature variables, machine learning algorithms can effectively capture complex nonlinear relationships and potential risk patterns. The prediction results based on historical and real-time data will help to estimate the risks of each road section in the future time period and provide a basis for adjusting the charging strategy. For example, if the accident frequency rate or traffic flow forecast value of a certain road section is high, the corresponding risk level and charging standards will be increased accordingly.

[0066] Based on the future risk level of each road section predicted by the machine learning algorithm, the next step is to adjust the toll standards of each road section according to the risk level. Road sections with higher risks require higher toll standards to ensure that transportation tasks are carried out on safer roads, while also providing financial support for the management and maintenance of high-risk roads. For roads with lower risks, lower toll standards can be set to promote a balanced distribution of vehicles and avoid all vehicles being concentrated on low-toll roads, resulting in excessive traffic pressure.

[0067] The adjustment of charging standards is not only based on risk prediction results, but also needs to be combined with real-time traffic conditions, policies and regulations, market demand and other factors. For example, in high-risk sections with very large traffic flows, it is possible to consider dynamically adjusting charging strategies and implementing high charges during peak hours to control traffic pressure and ensure transportation safety. At the same time, the system should also consider the differences in various types of transportation tasks and provide flexible charging plans for different types of goods and transportation needs. Through this dynamically adjusted charging mechanism, it is ensured that the charging standards can flexibly respond to changes in risks in various sections of the transportation network, and achieve more efficient resource allocation and risk control.

[0068] Finally, to ensure that the charging standards always match the actual risks, the system needs to provide real-time feedback and optimization adjustments. The prediction results of the machine learning model are continuously updated based on real-time monitoring data and new transportation data. If the new data deviates from the previous prediction, the model needs to update the future risk prediction through incremental learning or retraining. The real-time feedback mechanism ensures that the system continues to adapt to the dynamic characteristics of risk changes in the transportation network.

[0069] In actual operation, dynamic adjustments based on real-time monitoring data can effectively identify emergencies such as accidents, weather changes or traffic jams, so as to timely correct forecasts and charging standards. For example, if a serious accident occurs on a certain road section, causing the risk level of the road section to rise sharply, the system should quickly adjust the charging standards for the road section and notify the transporter to choose other routes with lower risks for detours. Through real-time feedback and optimization adjustments, the system can minimize the risks in the transportation process and keep the matching of charges and risks in different time periods and situations, ensuring a balance between transportation safety and cost-effectiveness.

[0070] In the above technical solution, the technical effects and advantages provided by the present invention are:

[0071] The present invention combines multidimensional data, uses population coverage risk models and accident expectation calculation methods to evaluate the risk level of each section in the transportation network in real time. By efficiently identifying high-risk sections, transportation tasks can be reasonably allocated according to different risk levels to avoid excessive load on high-risk sections, significantly improving the safety of the transportation process. The system can achieve accurate risk prediction and dynamically adjust charging standards, thereby providing safer and more scientific decision-making support for transportation companies and reducing safety hazards caused by improper path selection.

[0072] The present invention optimizes the transport route through genetic algorithms, comprehensively considers the transport cost and the risks of each section, and seeks the optimal balance between cost and risk. When selecting a route, the genetic algorithm can screen out routes with lower risks, thereby reducing unnecessary risk exposure and reducing transport costs. At the same time, the system dynamically adjusts the charging strategy, which changes in real time according to the predicted risk level, ensuring that transport companies can respond flexibly, further reducing transport costs and ensuring the safety and stability of transport.

[0073] The present invention uses a simulated annealing algorithm to globally optimize the transportation network, reasonably schedule tasks and optimize path allocation to avoid overloading of high-risk sections. The algorithm can evenly allocate transportation tasks in a global search, reduce traffic pressure on high-risk sections, and reduce the probability of accidents. By predicting future risk changes, the system can adjust task allocation in advance, achieve risk balance of transportation tasks, ensure reasonable allocation of resources and efficient scheduling of transportation tasks, and improve overall transportation efficiency and safety. BRIEF DESCRIPTION OF THE DRAWINGS

[0074] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in the present invention. For ordinary technicians in this field, other drawings can also be obtained based on these drawings.

[0075] Figure 1 The present invention is a flow chart of the method for charging dangerous goods transportation based on risk assessment. DETAILED DESCRIPTION

[0076] Example embodiments will now be described more fully with reference to the accompanying drawings. However, example embodiments can be implemented in a variety of forms and should not be construed as limited to the examples set forth herein; rather, these example embodiments are provided so that the description of the present disclosure will be more comprehensive and complete, and the concept of the example embodiments will be fully conveyed to those skilled in the art.

[0077] The present invention provides Figure 1 The risk-based dangerous goods transportation charging method shown includes the following steps:

[0078] By collecting data on traffic flow, accident frequency, environmental factors, etc. on each section of the dangerous goods transportation network, the potential risk level of each section is evaluated based on the population coverage risk model and accident expectation value calculation method;

[0079] Based on the risk assessment results, the road sections in the transportation network are divided into risk levels, and segmented charges are implemented for the road sections according to different risk levels. The risk level of each road section is clustered through the K-means clustering algorithm, and different charging standards are set based on the clustering results;

[0080] According to the given transportation starting point and end point, the transportation route is optimized by using genetic algorithm to minimize the transportation cost and reduce the total risk in the transportation process. The genetic algorithm optimizes the route selection and balances the cost and risk by simulating the natural selection process.

[0081] According to the risk level differences of transportation routes, a risk balance allocation strategy is implemented to avoid excessive burden of transportation tasks on high-risk sections. The global transportation network is searched through the simulated annealing algorithm to adjust the path distribution to achieve risk balance among different sections.

[0082] Based on real-time monitoring data, the charging standards of each section in the transportation network are dynamically adjusted. Machine learning algorithms are used to analyze historical transportation data, predict future risks of sections, and adjust charging strategies based on the prediction results to ensure that risks match charges.

[0083] Specific implementation 1: Risk assessment and path optimization system based on multi-dimensional data fusion;

[0084] In order to implement a risk-based dangerous goods transportation charging method, it is first necessary to develop a multi-dimensional data fusion platform that is designed to comprehensively collect, process and analyze data from different sources, including traffic monitoring systems, meteorological monitoring, historical accident data, road section terrain data and real-time road condition information. Through this platform, the system can combine multiple data dimensions to conduct a comprehensive risk assessment of each section in the transportation network. In terms of data collection, the system needs to continuously access real-time sensor data, such as traffic flow sensors, accident monitoring cameras and weather warning systems, to ensure that all factors affecting transportation risks can be reflected in a timely and accurate manner.

[0085] Data processing and evaluation is one of the core tasks of the system. First, the system uses a population coverage risk model to assess the potential risk of road sections in the transportation network. The model takes into account factors such as population density, traffic flow, and the capacity of the road section to derive a basic risk index for each road section. In addition, the accident expectation value calculation method takes into account factors such as historical accident frequency, accident type, and environmental impact, thereby further optimizing the risk level assessment of each road section. By using these advanced calculation methods, the system is able to conduct a scientific risk assessment of each road section in the transportation network to ensure the accuracy and comprehensiveness of the data.

[0086] After completing the risk assessment of each road section, the system will divide the transportation network according to different risk levels and classify the risk level of each road section through the K-means clustering algorithm. This process divides the road sections into high-risk, medium-risk and low-risk areas. The K-means clustering algorithm will automatically classify road sections with similar risk levels according to the comprehensive risk index of each road section, thereby ensuring that high-risk sections are charged high fees, while low-risk sections remain at a lower level. In this way, the system can set scientific charging standards for each road section to ensure a balance between transportation costs and risks.

[0087] Next, the system uses genetic algorithms to optimize the transport routes. Genetic algorithms can find the best path in a large-scale transport network by simulating the natural selection process. The system uses genetic coding to represent the path selection, and uses all potential transport paths as the initial population. After multiple generations of crossover and mutation operations, the genetic algorithm gradually optimizes the path and finds the optimal solution. These paths not only consider the transportation cost, but also comprehensively consider the risk level of each section of the road to minimize the overall transportation cost and risk. Through this path optimization process, the system can provide the most suitable transportation route for transportation companies, avoid high-risk sections, and improve transportation safety.

[0088] Finally, the platform will also adjust the charging strategy in real time according to sudden changes in the transportation process. For example, if a section of road becomes a high-risk area due to weather or accidents, the system will identify this situation in real time and re-evaluate the charging standards for this section of road. If the charging standards for high-risk sections have been raised, the system will also recommend that transportation companies choose low-risk alternative routes to minimize the risks that may be brought about during transportation. This data-driven dynamic adjustment mechanism not only improves the safety of the transportation process, but also provides accurate decision-making support for transportation management departments.

[0089] Through this risk assessment and route optimization system based on multi-dimensional data fusion, the system can not only respond to external factors such as traffic flow, weather changes and accidents in real time, but also dynamically adjust charging standards and transportation routes according to different transportation needs, ensuring a balance between safety and cost-effectiveness during transportation.

[0090] Specific implementation 2: Dynamic pricing system based on real-time monitoring and predictive analysis;

[0091] Another specific implementation method is a dynamic pricing system based on real-time monitoring and predictive analysis technology. The core of this system is to predict and evaluate the risk changes of each transportation section through in-depth analysis of historical transportation data combined with real-time monitoring data, so as to adjust the charging standards. First, the system needs to collect real-time data of each section in the transportation network through a variety of sensors (such as traffic flow sensors, meteorological monitoring equipment, etc.). These data include traffic flow, traffic accident frequency, meteorological conditions (such as temperature, precipitation, wind speed, etc.) and environmental pollution level of the section.

[0092] After data collection, the system uses machine learning algorithms to analyze historical transportation data and identify potential risk patterns for each road section. Through algorithms such as support vector machines (SVM), the system can learn and predict the risk level of each road section. For example, by learning past traffic accident data, the system can predict the frequency of accidents that may occur on certain roads under specific climate or traffic conditions, thereby making dynamic risk predictions for these roads. At the same time, the system will also use methods such as regression analysis to estimate the risks of each road section in the future based on changes in real-time factors such as traffic flow and weather, thereby providing real-time risk analysis reports for each transportation task.

[0093] After completing the risk prediction for each road section, the system will dynamically adjust the charging standards based on the prediction results. If the system predicts that a certain road section will have high risks in the future (such as frequent traffic accidents, bad weather, etc.), the system will immediately adjust the transportation costs of the section to reflect the changes in its risk level. For example, if the frequency of accidents on a certain section increases, the system will increase the transportation charges for that section to encourage transportation companies to choose other routes with lower risks. Conversely, for low-risk sections, the system will reduce charges to attract transportation companies to choose these sections for transportation.

[0094] In addition, the system will optimize the transportation route using prediction algorithms based on historical data and real-time monitoring data to avoid the concentration of transportation tasks on high-risk sections. By simulating risk assessment and cost calculation of different transportation routes, the system can recommend the best route to transportation companies to ensure that transportation tasks can be completed with the lowest cost and the least risk. The system can also automatically generate real-time risk reports to help transportation companies understand road conditions in a timely manner and adjust transportation plans based on prediction results.

[0095] This dynamic pricing system based on real-time monitoring and predictive analysis is highly adaptable and flexible, and can dynamically adjust transportation costs according to changing risk factors, ensuring transportation safety while minimizing transportation costs. The system can also perform personalized pricing and route selection according to the needs of different transportation tasks, thereby optimizing the transportation process and improving overall transportation efficiency.

[0096] Specific implementation 3: Transportation scheduling platform based on global optimization and risk balance strategy;

[0097] In order to achieve a balanced risk distribution of transportation routes and avoid excessive burden of transportation tasks on high-risk sections, a transportation scheduling platform based on global optimization and risk balance strategy can be adopted. The core task of this platform is to dynamically adjust the paths in the transportation network through real-time monitoring and risk assessment to ensure that high-risk sections are not over-utilized, thereby dispersing the risks of the overall transportation network. The system will first identify high-risk sections through a comprehensive risk assessment of each section, and perform global optimization through a simulated annealing algorithm. The simulated annealing algorithm is an optimization algorithm that simulates the natural cooling process. By introducing appropriate randomness in each iteration, it gradually optimizes the path selection and finally obtains the global optimal solution.

[0098] Specifically, the system will take into account the specific needs of the transportation task (such as the types of dangerous goods transported, transportation time requirements, etc.), and combine the risk level of each section to select and adjust the path. The simulated annealing algorithm will traverse all possible paths in the transportation network and continuously evaluate the overall risk and cost of the path. In this process, the system will reasonably allocate transportation tasks according to the risk level of each path to avoid concentrating too many transportation tasks on high-risk sections. The system will adopt an adaptive strategy to continuously adjust the path planning according to the real-time changing road conditions and risk conditions during the transportation process to ensure that the transportation tasks of each section are reasonably allocated and avoid frequent accidents due to excessive load on a certain high-risk section.

[0099] During the implementation process, the system not only takes into account existing transportation tasks, but also needs to plan for future transportation needs. Through historical data analysis and risk prediction, the platform can foresee in advance which sections of road may face higher risks in the future and adopt a more cautious strategy in allocating transportation tasks to these sections. For example, if a section of road is about to become a high-risk area due to bad weather, the system will adjust the transportation task allocation of that section in advance to ensure that tasks are not over-concentrated, thereby avoiding the accumulation of high risks.

[0100] In addition, the system will also use a variety of optimization algorithms and risk balance strategies to dynamically adjust the path selection of transport vehicles to ensure that the route of each transport vehicle can avoid high-risk sections as much as possible and meet the timeliness requirements of the transportation task. The combination of simulated annealing algorithm and global optimization mechanism can not only optimize path selection and avoid excessive burden in high-risk areas, but also ensure the cost-effectiveness of the transportation process, minimize the risks in the transportation process, and ensure the safety and efficiency of dangerous goods transportation.

[0101] This transportation scheduling platform based on global optimization and risk balance strategy not only improves the safety of the transportation process, but also optimizes the allocation of transportation resources, ensuring that the transportation network can make the best scheduling decisions when dealing with complex risk situations.

[0102] The present invention can collect and analyze transportation data (such as traffic flow, weather conditions, historical accident data, etc.) from different sources in real time through a risk assessment and path optimization system based on multi-dimensional data fusion, thereby realizing accurate risk assessment of each section in the transportation network. By using advanced algorithms such as population coverage risk model and accident expectation value calculation method, the system can accurately assess the risk level of each section and timely identify potential high-risk sections. This evaluation method based on big data and intelligent algorithms effectively improves the safety of the transportation process and ensures that transportation tasks can be reasonably allocated according to sections of different risk levels, thereby avoiding excessive load on high-risk sections. Through real-time monitoring and dynamic adjustment of charging standards, the system provides transportation companies with more scientific and safe transportation decisions, reduces safety hazards caused by improper path selection, and greatly improves the safety of transportation tasks.

[0103] The present invention optimizes the transportation route by using a genetic algorithm to minimize transportation costs and reduce risks, and generally achieves an effective balance between cost and risk. During the optimization process, the genetic algorithm simulates the natural selection process, evaluates and selects the most suitable transportation route, and not only considers the transportation cost, but also comprehensively considers the risk level of each section. This allows transportation companies to choose the route with the lowest cost and lower risk, thereby optimizing the transportation route configuration and reducing unnecessary risk exposure. Furthermore, when the system dynamically adjusts the charging standards in real time, it adjusts the charging strategy according to the predicted risk level, thereby ensuring that the transportation company can make flexible decisions based on the real-time risk situation. This dynamic pricing and route optimization mechanism effectively reduces transportation costs while also ensuring the safety and stability of transportation.

[0104] The present invention achieves global optimization of the transportation network through a simulated annealing algorithm, solving the problem of excessive burden of transportation tasks on high-risk sections. The system reasonably schedules transportation tasks and optimizes path allocation according to the risk level of the section and the demand for transportation tasks. The simulated annealing algorithm can effectively avoid excessive burden in high-risk areas in global search, ensuring that transportation tasks between different sections are evenly distributed. This strategy can not only reduce traffic pressure on high-risk sections, but also avoid too many transportation tasks being concentrated in high-risk areas, reducing the probability of traffic accidents. In addition, the system can also adjust task allocation based on future risk change predictions, avoid possible risk points in advance, thereby effectively balancing the risk distribution of transportation tasks and ensuring reasonable allocation of resources and efficient scheduling during transportation.

[0105] The above description is only by way of illustration of certain exemplary embodiments of the present invention. It is undoubted that those skilled in the art can modify the described embodiments in various ways without departing from the spirit and scope of the present invention. Therefore, the above drawings and descriptions are illustrative in nature and should not be construed as limiting the scope of protection of the claims of the present invention.

Claims

1. A dangerous goods transportation charging method based on risk assessment, characterized by: The following steps are involved: By collecting data on each section of the dangerous goods transportation network, the potential risk level of each section is evaluated based on the population coverage risk model and accident expectation value calculation method; Based on the risk assessment results, the road sections in the transportation network are divided into risk levels, and segmented charges are implemented for the road sections according to different risk levels. The risk level of each road section is clustered through the K-means clustering algorithm, and different charging standards are set based on the clustering results; According to the given transportation starting point and end point, the transportation route is optimized by using genetic algorithm to minimize the transportation cost and reduce the total risk in the transportation process. The genetic algorithm optimizes the route selection and balances the cost and risk by simulating the natural selection process. According to the risk level differences of transportation routes, a risk balance allocation strategy is implemented to avoid excessive burden of transportation tasks on high-risk sections. The global transportation network is searched through the simulated annealing algorithm to adjust the path distribution to achieve risk balance among different sections. Based on real-time monitoring data, the charging standards of each section in the transportation network are dynamically adjusted. Machine learning algorithms are used to analyze historical transportation data, predict future risks of sections, and adjust charging strategies based on the prediction results to ensure a match between risks and charges.

2. The dangerous goods transportation charging method based on risk assessment according to claim 1 is characterized in that: Based on the population coverage risk model and accident expectation value calculation method, the potential risk level of each road section is evaluated. The specific steps are as follows: Collect relevant data from each section of the dangerous goods transportation network, including traffic flow, accident frequency, and environmental factors. The collected data is pre-processed to ensure the accuracy and completeness of the data; After collecting and preprocessing the data, a comprehensive risk assessment function is defined, which combines the population coverage risk model and the accident expectation value calculation method to evaluate the risk level of each road section. The risk level R i It is expressed as: R i =α·T i +β·A i +γ·E i Where: T i is the traffic flow index of road section i, reflecting the busyness of the road section. i is the accident frequency index of road section i, indicating the probability of past accidents, E i is the environmental risk factor of road section i, α, β, and γ are weight coefficients, representing the degree of influence of each factor on the risk; In order to more accurately quantify the potential risks in the transportation process, the accident expectation value is used to measure the accident loss of each road section. The expression of the accident expectation value is: EAL i =∫0 T P(t)·C(t)dt Where: P(t) represents the probability of an accident occurring at time t, C(t) is the loss at the time of the accident, and T is the period of the entire transportation time; By calculating the risk level R of each road section i and the expected value of accidents (EAL) i Divide the road sections and classify them according to the comprehensive risks to obtain the comprehensive risk classification of each road section using the following formula: Among them, C i It is the comprehensive risk classification of road section i, which reflects the comprehensive safety risk level of each road section by weighted average of risk level and accident expectation value.

3. The dangerous goods transportation charging method based on risk assessment according to claim 2 is characterized in that: The K-means clustering algorithm is used to cluster the risk level of each road section, and different charging standards are set based on the clustering results. The specific steps are as follows: Before implementing the K-means clustering algorithm, it is necessary to first prepare and preprocess the data of the road segments in the transportation network; After the data preparation is completed, the K-means clustering algorithm is used to cluster the risk level of each road section. The K-means algorithm divides the data set into a predetermined number of clusters. The risk level of the road sections in each cluster is similar, and the difference in risk level between different clusters is maximized. First, a K value is selected, that is, the number of risk level groups divided; In the K-means algorithm, the risk level data of each road section will be classified into the closest cluster according to its distance to the cluster center. Through multiple iterative calculations, the cluster center is continuously updated until the clustering result converges. The result after clustering is the risk category corresponding to each road section. After clustering is completed, different charging standards will be set for each cluster based on its characteristics. Each cluster represents a road section with a similar risk level. After initially setting the charging standards, the charging strategy is further optimized through dynamic monitoring and data feedback.

4. The dangerous goods transportation charging method based on risk assessment according to claim 3 is characterized in that: According to the given transportation starting point and end point, the genetic algorithm is used to optimize the transportation route to minimize the transportation cost and reduce the total risk during the transportation process. The specific steps are as follows: Encode the transport path. Each transport path is regarded as an individual. The individual consists of a series of road sections. Each road section corresponds to a road section in the transport network. Let the transport starting point be O and the end point be D. Each road section of the transport path is composed of nodes in the network. Encode the transport path as a vector P = (p i )=(p1,p2,…,p n ), where p i represents the ID of the i-th road segment, and n is the number of road segments contained in the path; Next, we randomly generate an initial population P0 = {P1, P2, ..., P m }, where each individual p i is a transport path, and the path length is n, the population size is m, and when the population is initialized, the diversity of the paths is taken into account. Each path in the population is a different path from the starting point to the end point, and each section in the path is connected by random selection; In the genetic algorithm, the fitness function is used to measure the quality of individuals. For the optimization problem of the transportation path, the goal is to minimize the sum of the transportation cost and the risk in the transportation process. Therefore, the fitness function f(P) is expressed as the weighted sum of the cost and the risk. The expression of the fitness function f(P) is: f(P)=α′·C(P)+β′·R(P) Among them, C(P) is the transportation cost of path P, R(P) is the sum of the risks of all sections on path P, and α′ and β′ are coefficients for adjusting the weights between cost and risk; After fitness evaluation, the selection operation selects better individuals through the process of natural selection.

5. The dangerous goods transportation charging method based on risk assessment according to claim 4 is characterized in that: The transportation cost C(P) is the sum of the transportation costs of each road segment: Among them, c(p i ) is the road segment p i transportation costs.

6. The dangerous goods transportation charging method based on risk assessment according to claim 4 is characterized in that: The transport risk R(P) is the sum of the risks of each road segment, which is assessed based on the traffic volume, accident frequency and environmental factors of each road segment: Among them, r(p i ) is the road segment p i risk value.

7. The dangerous goods transportation charging method based on risk assessment according to claim 1 is characterized in that: The global transportation network is searched through the simulated annealing algorithm to adjust the path distribution. The specific steps are as follows: Before implementing the risk balance allocation strategy, it is necessary to first perform a comprehensive state initialization of the transportation network, including risk level assessment of each road segment; Before optimizing the simulated annealing algorithm, the objective function is defined. The core of the objective function is to achieve a balanced risk distribution of transportation tasks and avoid excessive transportation tasks on certain high-risk sections. The objective function is designed to maximize the risk balance of the entire transportation network, that is, to increase the transportation tasks on low-risk sections and avoid excessive burden on high-risk sections. During the search process of simulated annealing, the path is adjusted. Path adjustment is the core step of the algorithm. The simulated annealing algorithm continuously makes local adjustments to the existing path to find a better path distribution. As the algorithm progresses, the temperature will gradually decrease. In each round of iteration, a new risk balance is calculated based on the current path adjustment situation, and this value is used to determine whether to continue path adjustment.

8. The dangerous goods transportation charging method based on risk assessment according to claim 1 is characterized in that: Machine learning algorithms are used to analyze historical transportation data, predict future risks of road sections, and adjust charging strategies based on the prediction results to ensure that risks match charges. The specific steps are as follows: Before dynamically adjusting the toll standards for each section of the transportation network, ensure the comprehensive collection and pre-processing of real-time monitoring data to effectively reflect the real-time status of the current transportation network and provide accurate input for subsequent predictions and decisions; Based on the pre-processed real-time data and historical transportation data, risk prediction is performed using machine learning algorithms. The machine learning model predicts the future risk level of each road section by analyzing the trends and features in the historical data; Based on the future risk level of each road section predicted by the machine learning algorithm, the toll standards of each road section are adjusted according to the risk level; Finally, to ensure that charging standards always match actual risks, real-time feedback and optimization adjustments are carried out, and the prediction results of the machine learning model are continuously updated based on real-time monitoring data and new transportation data.

Images