Hazardous chemical substance transportation route planning method based on multi-objective optimization and related device

Abstract

The invention discloses a hazardous chemical substance transportation route planning method based on multi-objective optimization and a related device. The method comprises the following steps: acquiring hazardous chemical substance transportation task information and target road information of a plurality of iteration versions; comparing the transportation task and the road information of each iteration version to generate difference data; in response to the consistency of the difference data of each iteration version, training a route planning model based on the difference data; in response to inconsistency of the difference data of each iteration version, inputting the difference data of each version and the transportation demand into a model to generate final difference data; based on the final difference data and the current road information, generating an optimized transportation route through a multi-objective optimization algorithm; wherein the transportation task information comprises hazardous chemical substance types, transportation risk preferences and time windows, and the road information comprises safety facilities, real-time traffic flow, weather data and emergency rescue resource distribution. A transportation route considering safety, cost and efficiency can be effectively planned, and the actual requirements of transportation are met.

Description

Technical Field

[0001] The present application relates to the technical field of route planning, and in particular to a hazardous chemicals transportation route planning method and related devices based on multi-objective optimization. Background Art

[0002] With the rapid development of the economy and the growing prosperity of the chemical industry, the demand for the transportation of hazardous chemicals continues to grow. The transportation of hazardous chemicals is characterized by high risks and high requirements. If an accident occurs during transportation, it may have a serious impact on personnel safety, the environment and the economy.

[0003] The existing hazardous chemical transportation route planning mainly relies on manual experience, usually only focusing on basic factors such as distance or transportation cost, and cannot adapt to the growing demand and increasingly complex transportation scenarios in the field of hazardous chemical transportation. Summary of the Invention

[0004] The main technical problem solved by this application is to provide a hazardous chemical transportation route planning method and related devices based on multi-objective optimization, which comprehensively considers multiple objectives such as safety score, total transportation cost, and estimated transportation time, and can effectively and accurately plan a transportation route that takes into account safety, cost and efficiency.

[0005] In order to solve the above technical problems, a technical solution adopted in this application is: to provide a hazardous chemical transportation route planning method based on multi-objective optimization, the method including: obtaining multiple iterative versions of hazardous chemical transportation task information and target road information; comparing the transportation task and road information of each iterative version to generate difference data; in response to the consistency of the difference data of each iterative version, training a route planning model based on the difference data; in response to the inconsistency of the difference data of each iterative version, inputting the difference data of each version and the transportation demand into the model to generate final difference data; based on the final difference data and current road information, generating an optimized transportation route through a multi-objective optimization algorithm; wherein, the transportation task information includes the type of hazardous chemicals, transportation risk preference and time window, and the road information includes safety facilities, real-time traffic flow, weather data and emergency rescue resource distribution.

[0006] If the current iteration version is the initial iteration version, the route planning model is trained based on the difference data of the current iteration version.

[0007] The multiple iteration versions include the current iteration version and its historical iteration versions, and the historical iteration versions are all versions generated before the current iteration version.

[0008] Among them, the difference data consists of at least one difference element, which includes changes in the type of hazardous chemicals, adjustments to transportation risk preferences, updates to road safety facilities, and changes in traffic flow thresholds.

[0009] The difference data of each iterative version is input into the route planning model, including: calculating the initial matching probability of each difference element in different iterative versions, which is generated by the semantic similarity between the difference element and the transportation demand and the historical matching record; dynamically weighting each initial matching probability according to the transportation demand to generate the target matching probability; selecting the difference elements whose target matching probability exceeds the preset threshold and combining them to generate the final difference data.

[0010] Among them, dynamic weighting includes: if the transportation risk preference priority is higher than the cost, then the weight of the difference elements related to road safety facilities and emergency rescue resources will be increased; if the transportation time window priority is higher than the risk, then the weight of the difference elements related to traffic flow and weather data will be increased.

[0011] Among them, the optimization objectives of the multi-objective optimization algorithm include: safety score, total transportation cost, and estimated transportation time; among them, the safety score is calculated by weighted calculation of the completeness of road safety facilities, the coverage radius of emergency rescue resources and the matching degree of hazardous chemicals types, the total transportation cost includes path length, tolls and risk insurance costs, and the estimated transportation time is based on dynamic predictions of real-time traffic flow and weather data; among them, the optimized transportation route is selected from the candidate path set through multi-objective equilibrium analysis; the route planning model triggers iterative updates based on one or more of the following conditions: the change rate of real-time traffic flow data in the target road information exceeds the preset threshold; extreme weather warnings are added to the weather data; the distribution of emergency rescue resources undergoes structural adjustments.

[0012] In order to solve the above technical problems, another technical solution adopted by the present application is: to provide a hazardous chemical transportation route planning device based on multi-objective optimization, and the hazardous chemical transportation route planning device based on multi-objective optimization includes: an acquisition module, a comparison module, a first response module, a second response module and a generation module. The acquisition module is used to obtain multiple iterative versions of hazardous chemical transportation task information and target road information; the comparison module is used to compare the transportation task and road information of each iterative version to generate difference data; the first response module is used to respond to the consistency of the difference data of each iterative version, and then train the route planning model based on the difference data; the second response module is used to respond to the inconsistency of the difference data of each iterative version, input the difference data of each version and the transportation demand into the model, and generate the final difference data; the generation module is used to generate an optimized transportation route through a multi-objective optimization algorithm based on the final difference data and current road information; wherein, the transportation task information includes the type of hazardous chemicals, transportation risk preference and time window, and the road information includes safety facilities, real-time traffic flow, weather data and emergency rescue resource distribution.

[0013] In order to solve the above technical problems, another technical solution adopted in this application is: to provide a hazardous chemical transportation route planning device based on multi-objective optimization, and the hazardous chemical transportation route planning device based on multi-objective optimization includes: a memory and at least one processor, the memory stores instructions; at least one processor calls the instructions in the memory to enable the hazardous chemical transportation route planning device based on multi-objective optimization to execute the steps of the hazardous chemical transportation route planning method based on multi-objective optimization as described above.

[0014] In order to solve the above technical problems, another technical solution adopted in this application is: providing a computer-readable storage medium, on which instructions are stored, and when the instructions are processed and executed, the steps of the hazardous chemical transportation route planning method based on multi-objective optimization as described above are implemented.

[0015] Different from the existing technology, the beneficial effect of the present application is: by obtaining multiple iterative versions of hazardous chemical transportation task information and target road information, and comparing the transportation tasks and road information of each iterative version, differential data is obtained. If the differential data of each iterative version is consistent, the route planning model is trained based on the differential data, so that the model can effectively learn the relationship between transportation tasks, road information differences and route planning, which is convenient for subsequent efficient route planning based on different transportation scenarios, and realizes the effective use of transportation-related data; if the differential data of each iterative version is inconsistent, the differential data of each version and the transportation demand are input into the model to generate the final differential data, and then combined with the current road information, an optimized transportation route is generated through a multi-objective optimization algorithm. In this process, the multi-objective optimization algorithm comprehensively considers multiple objectives such as safety score, total transportation cost, and estimated transportation time, and can effectively and accurately plan a transportation route that takes into account safety, cost and efficiency, accurately meet the actual needs of hazardous chemical transportation, and facilitate improving the accuracy and safety of hazardous chemical transportation route planning, improving transportation efficiency and reducing costs. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 It is a flow chart of an implementation method of a hazardous chemicals transportation route planning method based on multi-objective optimization in this application.

[0017] Figure 2 It is a structural framework diagram of an implementation scheme of a hazardous chemicals transportation route planning device based on multi-objective optimization in this application.

[0018] Figure 3 It is a structural framework diagram of an implementation scheme of a hazardous chemicals transportation route planning device based on multi-objective optimization in this application. DETAILED DESCRIPTION

[0019] To make the objectives, technical solutions, and advantages of the embodiments of the present invention more clear, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the embodiments described are only part of the embodiments of the present invention, not all of them. All other embodiments obtained by ordinary technicians in this field based on the embodiments of the present invention without making any creative efforts shall fall within the scope of protection of the present invention.

[0020] The terms "including," "having," and any variations thereof, as used in the embodiments of the present invention are intended to cover non-exclusive inclusions. For example, a process, method, system, product, or device comprising a series of steps or units is not limited to the listed steps or units, but may optionally include other steps or units not listed, or may optionally include other steps or units inherent to the process, method, product, or device.

[0021] The term "and / or" in this article is only a description of the association relationship of associated objects, indicating that there may be three relationships. For example, A and / or B can mean: A exists alone, A and B exist at the same time, and B exists alone. In addition, the character " / " in this article generally indicates that the objects associated before and after are in an "or" relationship. In addition, "many" in this article means two or more than two. In addition, the term "at least one" in this article means any combination of at least two of any one or more of a plurality of. For example, including at least one of A, B, and C, can mean including any one or more elements selected from the set consisting of A, B, and C. In addition, the terms "first" and "second" are only used for descriptive purposes and cannot be understood as indicating or implying relative importance or implicitly indicating the number of technical features indicated. Therefore, the features defined as "first" and "second" may explicitly or implicitly include one or more features. In the description of this application, the meaning of "many" is two or more, unless otherwise clearly and specifically defined.

[0022] To facilitate understanding of this embodiment, a method for planning hazardous chemical transportation routes based on multi-objective optimization disclosed in an embodiment of the present invention is first introduced in detail. Figure 1 As shown, Figure 1 This is a flow chart of an implementation method of a hazardous chemicals transportation route planning method based on multi-objective optimization in this application, and this method includes the following steps.

[0023] Step S11: Acquire multiple iterations of hazardous chemical transport mission information and target road information. The transport mission information includes the hazardous chemical type, transport risk preference, and time window, while the road information includes safety facilities, real-time traffic flow, weather data, and emergency rescue resource distribution.

[0024] Specifically, acquiring multiple iterations of transport mission information and target road information allows for comprehensive consideration of hazardous chemical transport planning at different stages and under different circumstances. Hazardous chemicals can range from flammable and explosive to toxic and hazardous, each requiring different transport conditions and risks. Transport risk preferences reflect the transporter's risk tolerance; for example, some transporters may prefer longer but safer routes. Time windows define the time constraints of transport missions, such as requiring transport to be completed within a specific timeframe. Regarding road information, safety features such as guardrails and warning signs play a crucial role in ensuring transport safety. Real-time traffic flow can reflect road congestion. By understanding real-time traffic flow, congested sections can be avoided, improving transport efficiency. Weather data can impact road conditions and transport safety. For example, severe weather such as heavy rain and snow can make roads slippery and obstruct visibility. Emergency rescue resource distribution can include the locations of hospitals and fire stations along the route, ensuring timely assistance in the event of an emergency.

[0025] In some specific embodiments, information can be obtained through various means. Transport mission information can be obtained from transport order systems and relevant records of transport providers. Target road information can be obtained from channels such as traffic management databases, meteorological website websites, and real-time traffic monitoring equipment. For example, intelligent transportation systems can be used to collect real-time traffic flow data, while meteorological satellites and ground monitoring stations can be used to obtain weather data.

[0026] In some application scenarios, such as long-distance hazardous chemical transport, different iterations of information may reflect conditions at different dates and times. By comprehensively analyzing information from multiple iterations, better transport route planning can be achieved. For example, considering differences in traffic flow at different times of day, departures can be made during lighter traffic periods; or seasonal weather patterns can be used to avoid sections of road prone to inclement weather. Furthermore, by combining transport risk preferences and time windows, the needs of the transporter can be met to the greatest extent possible while ensuring transport safety.

[0027] In some embodiments, the multiple iterative versions include a current iterative version and its historical iterative versions, where the historical iterative versions are all versions generated before the current iterative version.

[0028] Specifically, differentiating multiple iterations into the current iteration and its historical iteration will help to conduct a more comprehensive analysis during the hazardous chemical transportation route planning process. The current iteration represents the latest collection of hazardous chemical transportation task information and target road information, which reflects the actual situation at the current moment. The historical iterations are all versions generated before the current iteration, which record transportation-related information at different time points in the past. By comparing and analyzing historical iterations with the current iteration, it is possible to discover how transportation tasks and road conditions change over time, thereby providing a basis for more accurate route planning. For example, it is possible to discover the traffic flow change trend of certain roads within a specific time period, or the impact of certain seasons on the risk of hazardous chemical transportation.

[0029] In some specific embodiments, historical iteration information can be obtained by establishing a dedicated database to store the data of each iteration. Each time a new iteration is generated, its information is fully recorded in the database. When route planning is required, data for the current iteration and historical iterations are extracted from the database. This data can be analyzed using techniques such as data mining and machine learning. For example, time series analysis methods can be used to predict future traffic flow changes, or classification algorithms can be used to assess transportation risk levels in different historical periods.

[0030] In some application scenarios, such as long-term cross-regional hazardous chemical transportation projects, road conditions and transportation requirements in different regions may change significantly over time. By referring to historical iterations, transportation planners can understand traffic flow, safety facility usage, and the deployment of emergency rescue resources on various roads during special periods such as different seasons and holidays. For example, during peak tourist season, traffic flow on certain roads may increase significantly, while the distribution of emergency rescue resources may be adjusted during holidays. By combining this historical information with data from the current iteration, more realistic, safer, and more efficient transportation routes can be developed.

[0031] Step S12: Compare the transportation tasks and road information of each iterative version to generate difference data.

[0032] Specifically, comparing transport missions and road information across iterations and generating differential data allows for a precise understanding of changes in transport-related factors between versions. For transport mission information, differences in hazardous chemical type, transport risk preference, and time window are compared. A change in hazardous chemical type can lead to different transport requirements and potential risks; changes in transport risk preference may adjust transport route selection; and shifts in time windows can affect the timeliness of transport arrangements. For road information, differences in safety features, real-time traffic flow, weather data, and the distribution of emergency rescue resources are compared. The addition or removal of safety features can affect road safety; changes in real-time traffic flow directly impact transport efficiency; differences in weather data can alter road conditions and impact transport safety; and differences in the distribution of emergency rescue resources can lead to varying response capabilities in the event of an emergency. Generating this differential data provides a clearer understanding of the dynamic changes in the transport environment, providing a more accurate basis for subsequent route planning.

[0033] In some specific embodiments, transport mission information can be compared through data matching and difference calculation. For example, hazardous chemical types from different iterations can be compared one by one, noting any changes in type. Transport risk preferences can be quantified into numerical values, and the differences between different versions can be calculated. When comparing road information, real-time traffic flow can be compared using data recorded by traffic monitoring systems at different time points to calculate increases or decreases in flow. Weather data can be compared for differences in temperature, humidity, rainfall, and other indicators. The distribution of safety facilities and emergency rescue resources can be compared using a Geographic Information System (GIS) to compare changes in their location and quantity. These comparison results are then organized into a difference data table or dataset.

[0034] In some application scenarios, when transporting hazardous chemicals within a city, different iterations may reflect the conditions on different weekdays or during different time periods. For example, the real-time traffic flow during peak hours in the morning and evening on weekdays is quite different from that during non-peak hours. By comparing the generated differential data, transportation planners can choose to avoid congested sections during peak hours and choose relatively unobstructed routes. In addition, when encountering severe weather, the differences in weather data from different iterations can remind transportation personnel to prepare countermeasures in advance, such as adjusting transportation time or changing routes to avoid sections that may be more affected by the weather. In long-term transportation projects, as urban construction progresses, road safety facilities and the distribution of emergency rescue resources may change. Differential data can help detect these changes in a timely manner and ensure transportation safety.

[0035] In some embodiments, the difference data consists of at least one difference element, which includes changes in the type of hazardous chemicals, adjustments to transportation risk preferences, updates to road safety facilities, and changes in traffic flow thresholds.

[0036] Specifically, variance data, as the core element reflecting changes in transport-related information between different iterations, can be composed of one or more variance elements, which can intuitively reflect the dynamic adjustments to transport tasks or road conditions. A change in hazardous chemical type can involve switching the type of hazardous chemical within the same transport task, such as switching from a common corrosive chemical to a highly toxic one. Such a change directly affects transport packaging requirements, risk prevention and control measures, and route selection criteria. Adjustments to transport risk preferences may involve the transporter re-prioritizing safety, cost, and efficiency, such as shifting from prioritizing transport cost reduction to prioritizing transport safety, which in turn affects the weighting of route planning. Updates to road safety facilities can include adding crash barriers, emergency parking lanes, or removing outdated warning signs to specific sections of road. Such changes can alter the safety performance assessment of that section. Changes in traffic flow thresholds may refer to adjustments to the criteria for determining the degree of congestion on a particular road, such as changing "speeds below 30 km / h are considered congested" to "speeds below 20 km / h are considered congested," which can affect the judgment of traffic efficiency during route planning.

[0037] In some specific embodiments, identifying difference elements can be achieved through data comparison technology. For example, for changes in the type of hazardous chemicals, the system can parse the cargo classification code in the transport order, match the fields with the historical version, and mark the type inconsistency; for adjustments to transport risk preferences, they can be captured through the parameter change records of the user input interface, such as when the user modifies the risk weight coefficient in the management background, the difference mark is triggered; the update information of road safety facilities can be obtained by accessing the facility maintenance database of the traffic management department, regularly capturing change records and comparing them with historical versions; changes in traffic flow thresholds can be analyzed by analyzing the distribution pattern of historical traffic data, combining the personalized needs of the transporter to make dynamic adjustments and record differences. The above process can be automatically executed through a preset difference detection algorithm, or it can be completed through a combination of manual review and system verification.

[0038] In some application scenarios, when a transportation task involves transporting multiple batches of hazardous chemicals, the type of hazardous chemicals may change due to changes in customer demand. For example, a batch of transportation tasks was originally planned to transport flammable liquids, but was later adjusted to transport corrosive liquids. At this time, the hazardous chemical type change element in the difference data will trigger the system to re-evaluate the safety of the route; during holidays or extreme weather periods, the transporter may temporarily adjust the transportation risk preference, such as increasing the risk weight for the impact of severe weather. At this time, the risk preference adjustment element in the difference data will prompt the system to give priority to routes equipped with more emergency facilities; when a commonly used transportation route has a new tunnel escape passage or fire hydrant due to urban infrastructure construction, the updated elements of the road safety facilities will be identified and incorporated into the route planning model to improve the safety score of the route; during peak hours in the morning and evening, the system may dynamically adjust the traffic flow threshold based on real-time traffic data. At this time, the change element of the traffic flow threshold will affect the travel time prediction of the candidate routes, thereby optimizing route selection.

[0039] Step S13: In response to the difference data of each iterative version being consistent, a route planning model is trained based on the difference data.

[0040] Specifically, in response to the consistency of the differential data of each iterative version, it means that the changes in transportation tasks and road information between different versions show stability or regularity. At this time, training the route planning model based on the differential data can provide the route planning model with representative training samples, thereby improving the route planning model's ability to identify stable change patterns. The training process can include inputting the differential data into the route planning model, optimizing the parameters of the route planning model through an algorithm, and enabling the route planning model to learn the mapping relationship between the differential data and the transportation route planning. For example, when the differential data of multiple versions all show that a certain type of hazardous chemical change is associated with the update of specific road safety facilities, the route planning model can strengthen its recognition of this association through training, thereby giving priority to relevant factors in subsequent planning.

[0041] In some specific embodiments, the route planning model can be trained using machine learning algorithms, such as neural networks and genetic algorithms. First, the differential data is preprocessed, including data cleaning and feature extraction, to convert differential elements, such as changes in hazardous chemical types and updated road safety facilities, into feature vectors that the model can recognize. The feature vectors are then input into the model, and the model's weights and biases are adjusted through iterative calculations to gradually improve the match between the route planning results output by the model and the actual optimization objectives, such as safety scores and transportation costs. A verification mechanism can be introduced during the training process to test the model using a portion of the differential data to ensure the model's generalization capabilities in scenarios with consistent differential data.

[0042] In some application scenarios, when hazardous chemical transportation tasks in a certain area are primarily of a fixed type over a long period of time, and information such as road safety facilities and the distribution of emergency rescue resources remains stable over a period of time, the differential data of each iterative version may converge. For example, if a chemical park has been transporting the same type of hazardous chemicals for a long time and the surrounding roads have not undergone recent facility renovations, training the model based on consistent differential data can enable the model to focus on learning the impact of factors such as transportation risk preference adjustments and changes in traffic flow thresholds on route planning in this scenario, thereby improving the model's planning accuracy in specific scenarios. The trained model can be applied to daily transportation planning, quickly generating optimized routes that conform to historical patterns, and reducing the computational cost of real-time data processing.

[0043] Step S14: In response to inconsistency between the difference data of each iterative version, the difference data of each version and the transportation demand are input into the model to generate final difference data.

[0044] Specifically, when the difference data between iterations is inconsistent, it indicates that the changes in transportation tasks and road information between versions are complex and irregular. In this case, the difference data from each version is input into the model along with transportation demand to comprehensively consider all factors that may affect route planning. Transportation demand can include multiple aspects such as timeliness requirements, cost budgets, and tolerance for different risks. The model conducts in-depth analysis and integration of this input information. Using built-in algorithms and rules, it weighs the relationship between different difference data and transportation demand to generate a final difference data that reflects the overall situation. This final difference data more accurately reflects the actual changes in the current transportation scenario, providing a more reliable basis for subsequent route planning. For example, differences in hazardous chemical types, transportation risk preferences, road safety facilities, etc. between different versions may affect route selection. The model comprehensively considers these differences and the transporter's time and cost requirements to generate a comprehensive result.

[0045] In some specific embodiments, a model of a multi-objective optimization algorithm can be used to process these inputs. First, the difference data of each version is standardized, and different types of difference elements are converted into a unified numerical representation so that the model can perform effective calculations. For transportation demand, it can be quantified into specific indicators, such as time cost, economic cost, risk cost, etc. Then, these standardized difference data and quantified transportation demand are input into the model. The model will calculate and analyze the input according to the preset weights and algorithms, and finally generate the final difference data through continuous iteration and optimization. For example, in a transportation network including multiple cities, the difference data of the iterative versions of different cities may be different, and the transportation demand may also vary depending on the urgency of the goods and customer requirements. The model can generate the final difference data by calculating the comprehensive cost of different routes, such as time, cost, etc.

[0046] In some application scenarios, such as cross-regional transportation of hazardous chemicals, road conditions, traffic flow, and weather conditions in different regions may change significantly over time, resulting in inconsistent differential data across different iterative versions. For example, a transportation route passes through multiple cities, and one city is undergoing road construction, resulting in changes in traffic flow and safety facilities on that section of the road, while the situation in other cities remains relatively stable. At the same time, transporters may have different transportation requirements, such as requiring goods to be delivered within a specific timeframe or wanting to minimize transportation costs. In this case, the differential data and transportation requirements of each version are input into the model. The resulting differential data can help transport planners select an optimal route that both meets transportation needs and adapts to changes in different regions.

[0047] In some embodiments, if the current iteration version is the initial iteration version, the route planning model is trained based on the difference data of the current iteration version.

[0048] Specifically, when the current iteration version is the initial iteration version, since there is no historical iteration version for comparison, the difference data of the current iteration version reflects the characteristics of the transportation task and road information in the initial state. Training the route planning model based on these difference data can enable the route planning model to initially learn the basic laws and patterns under the transportation scenario. The route planning model will use these difference data as samples and try to establish a mapping relationship between the difference data and reasonable transportation routes by adjusting its own parameters and structure. For example, the difference data of the initial iteration version shows that a certain type of hazardous chemicals has a certain correlation with the traffic flow of a specific road. The route planning model will capture this correlation during the training process and take this factor into account in subsequent route planning.

[0049] In some specific embodiments, supervised learning methods can be used for model training. The difference data of the current iteration version is used as input features, and some reasonable route planning results are set as target outputs based on the actual transportation situation or experience. The model will continuously adjust its own weights and biases to make the model output as close to the target output as possible. For example, using a neural network model, difference data such as hazardous chemical types and road safety facilities are used as nodes in the input layer. After calculation and processing in the hidden layer, a route planning prediction result is finally output. By comparing with the target output, the error is calculated and back-propagated to update the model parameters.

[0050] In some application scenarios, such as newly launched hazardous chemical transportation operations, the initial iteration is the initial version. In this case, training a model based on the differential data from the current iteration can provide a foundational reference for subsequent transportation planning. For example, when a new chemical park begins hazardous chemical transportation, the initial iteration records information such as the initial conditions of the park's surrounding roads and the basic requirements of the transportation mission. By training the model based on this information, the route planning model can be adjusted and optimized based on new circumstances during subsequent transportation operations, gradually improving the accuracy and rationality of route planning.

[0051] In some embodiments, the route planning model triggers iterative updates based on one or more of the following conditions: the rate of change of real-time traffic flow data in the target road information exceeds a preset threshold; new extreme weather warnings are added to the weather data; and there are structural adjustments to the distribution of emergency rescue resources.

[0052] Specifically, the iterative update of the route planning model is used to ensure the accuracy and safety of hazardous chemical transportation route planning. When the rate of change of real-time traffic flow data in the target road information exceeds a preset threshold, it means that the road's traffic conditions have changed significantly. For example, a previously unobstructed road may experience a sharp increase in traffic due to an accident or special event. If the rate of change exceeds the preset threshold, continuing to transport according to the original planned route may result in a significant increase in transportation time and even increase transportation risks. At this time, triggering an iterative update of the model allows the model to replan a more appropriate route based on the new traffic flow conditions.

[0053] New extreme weather warnings in weather data, such as heavy rain, snow, and high winds, can have a significant impact on road conditions and transportation safety. Heavy rain can cause waterlogging on roads, compromising driving safety; heavy snow can cause icy roads, increasing the risk of skidding. Models must be updated promptly based on these extreme weather warnings, adjusting route planning to avoid sections of road most affected by extreme weather and ensuring transportation safety.

[0054] Structural changes in the distribution of emergency rescue resources, such as the relocation of fire stations or the redeployment of hospital emergency resources, can alter emergency response capabilities during transportation. The model iteratively updates based on these structural changes, fully factoring in the new distribution of emergency rescue resources when planning routes, ensuring timely and effective rescue support in the event of an emergency.

[0055] In some specific embodiments, monitoring the rate of change in real-time traffic flow data can be accomplished by collecting traffic flow information in real time through sensors, traffic cameras, and other devices installed on the road, and transmitting the data to the model. The model calculates the rate of change between the current traffic flow and the previous moment or the historical average traffic flow. When the rate of change exceeds a preset threshold, such as 30%, the model is automatically triggered to iterate and update. For example, during rush hour in the morning and evening, traffic flow on certain main roads changes frequently. The model will continuously monitor the rate of change of traffic flow on these sections of road, and adjust the route planning once the threshold is exceeded.

[0056] New extreme weather warnings added to weather data can be monitored through real-time integration with the meteorological department's database. When an extreme weather warning is issued, the route planning model can promptly access relevant information and reassess and adjust the route plan based on the warning's type and impact. For example, if a warning of impending heavy rain is received for a certain road section, the route planning model will prioritize routes that avoid that section.

[0057] To address structural adjustments in the distribution of emergency rescue resources, a database can be established and regularly updated. Whenever structural adjustments occur to the information in the database, the model automatically captures these changes and recalculates the safety and emergency response capabilities of routes based on the new distribution. For example, if a hospital in a certain area adds a specialized hazardous chemicals emergency department, the model will consider that hospital in its emergency rescue resource allocation when planning routes.

[0058] In some application scenarios, during large-scale urban events, such as concerts and sporting events, traffic around the event venue can surge. If the model detects that the rate of change in real-time traffic flow in that area exceeds a preset threshold, it triggers an iterative update to plan routes for hazardous chemical transport vehicles that avoid the event area, preventing congestion and reducing transportation risks.

[0059] During periods of seasonal extreme weather, such as summer rainstorms and winter snowstorms, the model closely monitors weather data. Upon receiving an extreme weather warning, it quickly adjusts route planning to ensure the safety of transport vehicles. For example, during heavy rain, the model will choose higher, well-drained roads to prevent vehicles from being trapped by water.

[0060] When cities undergo infrastructure construction or adjust regional planning, this can lead to structural adjustments in the distribution of emergency rescue resources. For example, if a new fire station is built in a certain area, the model will consider this change when planning routes, taking the fire station's location into account, enabling transportation routes to more quickly access rescue support in the event of an emergency.

[0061] In some embodiments, the difference data of each iterative version is input into the route planning model, including: calculating the initial matching probability of each difference element in different iterative versions, the initial matching probability is generated by the semantic similarity between the difference element and the transportation demand and the historical matching record; dynamically weighting each initial matching probability according to the transportation demand to generate a target matching probability; selecting the difference elements whose target matching probability exceeds a preset threshold and combining them to generate the final difference data.

[0062] Specifically, the process of inputting differential data into the route planning model is essentially to screen out the differential elements most relevant to the current transportation demand through quantitative analysis. The calculation of the initial matching probability can be combined with semantic similarity and historical matching records. Semantic similarity can be used through natural language processing technologies such as TF-IDF and Word2Vec to assess the degree of semantic relevance between differential elements and the transportation demand description. For example, when the transportation demand emphasizes "high safety", the differential element "road safety facility update" has a high semantic similarity with the demand, and the initial matching probability increases accordingly. Historical matching records can include the frequency or effectiveness scores of similar differential elements adopted in past transportation tasks, providing a reference for current matching through statistical analysis.

[0063] The dynamic weighting process adjusts the importance of each differential element based on the priority of transportation needs. For example, if the "time window" in transportation needs is given a higher priority, the weight of differential elements related to traffic flow and weather data will be increased; if the "transport risk preference" focuses more on safety, the weight of differential elements related to road safety facilities and emergency rescue resources will be strengthened. Ultimately, differential elements with a high probability of target matching are screened using a preset threshold, ensuring that the elements included in the final differential data significantly influence route planning decisions and preventing redundant information from interfering with model judgment.

[0064] In some specific embodiments, when calculating the initial matching probability, a semantic similarity matrix can be constructed, the descriptive text of each differential element and the transportation demand text can be vectorized, and the correlation between the two can be calculated using cosine similarity. Historical matching records can be stored in a database, recording the actual effect of each differential element in past transportation tasks. For example, the extent to which transportation risk was reduced after a change in the type of hazardous chemicals was adopted in historical planning can be used as a correction factor for the initial matching probability. Dynamic weighting can use the Analytic Hierarchy Process (AHP) or a machine learning algorithm such as random forest to automatically adjust the weight of each differential element based on the input parameters of the transportation demand, such as the weight coefficients of time, cost, and risk. The preset threshold can be set by empirical value, such as 0.6, or it can be dynamically determined through training on historical data.

[0065] For example, when the difference element is "traffic flow threshold change", if the weight of "estimated transportation time" in the transportation demand is 0.7, the initial matching probability of this element of 0.5 will be weighted to 0.8, and after exceeding the preset threshold of 0.7, it will be included in the final difference data; and if the initial matching probability of "hazardous chemical type change" is 0.4, it will still be lower than the threshold after weighting, and it will not be included for the time being.

[0066] In some application scenarios, such as cross-regional hazardous chemical transport, if a batch of transport requires "low risk, strict time window," the difference data includes "hazardous chemical type changed from ordinary flammable to highly toxic," indicating high semantic similarity. Historical matching records indicate a need for an increased safety weight. A "traffic flow threshold reduction on a certain road section" is related to the time window, and historical matching records indicate a significant impact on traffic efficiency. Calculations show that the initial match probability of 0.6 for the former is dynamically weighted to 0.9 with the safety weight, while the initial match probability of 0.5 for the latter is weighted to 0.8 with the time weight. Both exceed the threshold of 0.7. The final difference data contains these two elements, prompting the model to prioritize routes with higher safety facilities and stable traffic flow.

[0067] For example, when the transportation demand prioritizes "low cost," safety-related variance elements like "updated road safety facilities," even if they have high semantic similarity, will be dynamically weighted below the threshold due to their low cost weight and thus not included in the final variance data. This ensures the model prioritizes cost-related factors like tolls and route length. Through this mechanism, the model can adapt to different transportation demands, generating final variance data that better reflects actual scenarios, improving the relevance and effectiveness of route planning.

[0068] In some embodiments, dynamic weighting includes: if the transportation risk preference priority is higher than the cost, then the weight of the difference elements related to road safety facilities and emergency rescue resources is increased; if the transportation time window priority is higher than the risk, then the weight of the difference elements related to traffic flow and weather data is increased.

[0069] Specifically, the core of the dynamic weighting mechanism is to differentiate the importance of differentiating elements based on the priorities of different dimensions in transportation demand. When transportation risk preference takes precedence over cost, it means that the transporter is more concerned about safety during transportation. At this time, the weights of differentiating elements related to road safety facilities, such as the completeness of guardrails and the density of warning signs, and emergency rescue resources, such as the coverage of fire stations and hospital response times, can be increased through preset rules or algorithms. For example, the update of road safety facilities may directly affect the probability of accidents, and the distribution of emergency rescue resources is related to the efficiency of handling after an accident. Increasing the weights of these elements can prompt the model to prioritize safer paths when planning routes.

[0070] If the transport time window is prioritized over risk, indicating that the transporter has stricter timeliness requirements, the weighting of differential elements related to traffic flow and weather data will be increased. For example, changes in traffic flow directly affect the estimated transport time, and inclement weather may cause route travel time to increase. Increasing the weighting of these elements can make the model more inclined to select routes with high traffic efficiency and low weather impact, ensuring that the transport task is completed within the time window.

[0071] In some specific embodiments, dynamic weighting can be achieved by: A priority-weight mapping table is established within the system. When users enter their transportation needs and select a mode such as "risk-first," "time-first," or "cost-first," the system automatically retrieves the corresponding weighting parameters. For example, when selecting "risk-first" mode, the weight coefficient for road safety facilities-related differentiation factors increases from the default 0.3 to 0.6, and the weight for emergency rescue resources increases from 0.2 to 0.5. When selecting "time-first" mode, the weights for traffic flow and weather data increase from 0.3 to 0.7 and 0.6, respectively.

[0072] Dynamic calculations based on user input allow users to customize the priorities of each requirement dimension. For example, using sliders to set the priority ratios for risk, time, and cost, the system dynamically adjusts the weights of these differentiating elements based on the input ratios. For example, if a user sets risk preference at 60%, time window at 30%, and cost at 10%, the combined weights of safety facilities and emergency resources will be no less than 60%, while the combined weights of traffic flow and weather data will not exceed 30%.

[0073] Machine learning dynamically adjusts the weighting model using historical transportation data. When it detects a higher priority for a particular dimension in transportation demand, the model automatically optimizes the weights of the relevant differentiating elements. For example, historical data shows that in a "risk-first" scenario, increasing the weight of safety features can reduce the accident rate by 20%. The model translates this experience into weighting adjustment rules and automatically applies them to new tasks.

[0074] In some application scenarios, such as high-risk hazardous chemical transportation, when transporting highly toxic chemicals or flammable and explosive materials, the transporter usually sets the risk preference priority to the highest. In this case, if the difference data includes "a new dedicated emergency lane for hazardous materials on a certain road section," which means an update of road safety facilities, or "a new professional hazardous chemical rescue team along the route," which means an adjustment of emergency rescue resources, the dynamic weighting mechanism will significantly increase the weight of these difference elements. For example, the difference element "distribution of emergency rescue resources," which originally had a weight of 0.4, is increased to 0.8 in the risk-priority mode, prompting the model to give priority to routes passing through the area covered by the rescue team, even if the cost of that route is slightly higher.

[0075] Emergency supply transportation: When the transport time window is tight, such as when hazardous medical supplies need to be delivered within four hours, the transporter prioritizes this time window. In this case, if the variance data includes information such as a 20% drop in real-time traffic flow on a particular road section (i.e., a change in the traffic flow threshold) or a rainstorm warning issued for the original route area (i.e., an update in weather data), the weights of variance factors related to traffic flow and weather will be increased from the default 0.3 to 0.9. The model will prioritize routes with smooth traffic flow and moderate safety over more detours and higher safety routes, ensuring that the transport time meets the requirements.

[0076] Conventional cost-sensitive transport: When transporting common hazardous chemicals and prioritizing cost, the dynamic weighting mechanism reduces the weight of safety and time-related factors. For example, the weight of the "updated road safety facilities" differential factor is reduced from 0.5 to 0.2, while the weight of cost-related factors such as "route length" and "tolls" is increased. In this case, the model may select a route with moderate safety but the lowest tolls, balancing cost and risk.

[0077] Through this dynamic weighting mechanism, the route planning model can adapt to the core needs of different transportation scenarios, find the optimal balance between safety, timeliness and cost, and improve the practicality and flexibility of hazardous chemical transportation route planning.

[0078] Step S15: Based on the final difference data and current road information, an optimized transportation route is generated through a multi-objective optimization algorithm.

[0079] Specifically, the core of the multi-objective optimization algorithm is to strike a balance between multiple interrelated objectives, such as safety, transportation costs, and estimated transportation time. The final difference data can reflect the dynamic changes in transportation tasks and road conditions, such as the special safety protection requirements posed by changes in hazardous chemical types and the impact of adjustments in transportation risk preferences on route selection. Current road information provides real-time or up-to-date road condition data, such as the status of safety facilities, traffic flow distribution, and weather impact range. The algorithm transforms this information into a quantifiable objective function. For example, it sets the safety score as a weighted value of the completeness of road safety facilities and the coverage of emergency rescue resources; decomposes transportation costs into the sum of path length, tolls, and risk insurance costs; and links estimated transportation time to a dynamic prediction model based on real-time traffic flow and weather data. Then, through mathematical programming or heuristic algorithms, it generates a set of candidate routes and selects the optimal route that satisfies the multi-objective balance.

[0080] In some specific embodiments, the objective function is constructed as follows: Safety score: It can be calculated by weighting the completeness of road safety facilities such as guardrail density, warning sign coverage, emergency rescue resource coverage radius and the matching degree of hazardous chemical types, and the weight can be dynamically adjusted according to the transportation risk preference. Total transportation cost: includes path length, tolls and risk insurance costs. Estimated transportation time: Dynamic prediction is made based on real-time traffic flow data and weather data, and historical travel time data can be introduced for correction. Among them, the completeness of road safety facilities such as anti-collision guardrail density and special warning sign coverage rate for hazardous goods transportation. The matching degree of emergency rescue resource coverage radius and hazardous chemical types, such as the requirement that highly toxic chemicals have a professional rescue team within 50 kilometers. The path length can be calculated through GIS maps, and tolls can include highway tolls, bridge tolls, etc. The risk insurance cost can fluctuate according to the type of hazardous chemicals and the risk level of the route.

[0081] Multi-objective optimization algorithms can employ heuristic algorithms such as the non-dominated sorting genetic algorithm (NSGA-II) and particle swarm optimization (PSO) to iteratively optimize candidate routes. For example, in NSGA-II, each candidate route corresponds to a three-dimensional objective vector containing safety, cost, and time. Selection, crossover, and mutation operations are used to generate a Pareto-optimal solution set. Ultimately, the route with the highest overall score is selected based on transportation demand priorities, such as a safety weight of 0.4, a cost weight of 0.3, and a time weight of 0.3.

[0082] In some application scenarios, such as long-distance transportation of high-risk hazardous chemicals, if the final difference data indicates that the hazardous chemical has changed to flammable and explosive, and the transportation risk preference is prioritized, the multi-objective optimization algorithm will significantly increase the safety score weight, such as 0.6. In this case, the algorithm may prioritize a detour via a highway with more dedicated hazardous materials emergency facilities. Although this increases transportation costs by 15%, the safety score improves by 30%. Furthermore, by using real-time traffic flow data to avoid peak hours in the morning and evening, the estimated transportation time increases by only 10%.

[0083] Emergency urban delivery of hazardous medical supplies: When the transport time window is prioritized, such as a 2-hour delivery requirement, and variance data indicates a change in traffic flow thresholds due to temporary construction on a particular road section, indicating increased congestion risk, the algorithm increases the weight of the estimated transport time to 0.7. In this case, the model may select a route with a 5-kilometer increase in length but smoother real-time traffic. Based on weather data, if there are no extreme weather impacts, the speed prediction model is dynamically adjusted to ensure the transport time meets the requirement. Safety objectives are also balanced by optimizing the presence of safety features (for example, prioritizing routes with continuous emergency lanes).

[0084] Conventional chemical raw material transportation is often cost-sensitive. If cost is the primary transport requirement, a weight of 0.5 can be set. If the variance data does not include high-risk factors, the algorithm will prioritize minimizing total transportation costs. For example, choosing the shortest national highway route, despite a slightly lower safety score, can reduce the safety weight to, for example, 0.3. By combining dynamic weighting with real-time traffic flow to avoid daily congestion, the total cost can be reduced by 20% compared to the original route, with an estimated transportation time increase of only 5%, meeting cost control requirements.

[0085] In some embodiments, the optimization objectives of the multi-objective optimization algorithm include: safety score, total transportation cost, and estimated transportation time; wherein, the safety score is calculated by weighted calculation of the completeness of road safety facilities, the coverage radius of emergency rescue resources, and the matching degree of hazardous chemical types; the total transportation cost includes path length, tolls, and risk insurance costs; the estimated transportation time is based on dynamic predictions of real-time traffic flow and weather data; wherein, the optimized transportation route is selected from a set of candidate paths through multi-objective equilibrium analysis.

[0086] Specifically, the core of the multi-objective optimization algorithm is to transform key factors influencing the transportation process into quantifiable objective functions, then generate the optimal route through weighted calculation and equilibrium analysis. A safety score can be achieved by integrating the degree of compatibility between road safety facilities and emergency resources. For example, for the transportation of highly toxic chemicals, the completeness of road safety facilities and the coverage radius of emergency rescue resources can be assigned weights of 0.6 and 0.4, respectively, and a comprehensive score can be obtained through linear weighting. The total transportation cost can be calculated by including the route length, tolls, and risk insurance costs, which can be weighted and summed in a ratio of 0.5:0.3:0.2. Dynamic predictions of estimated transportation times can be generated by combining real-time traffic flow and weather data and calibrating them with historical travel time data. Ultimately, the algorithm selects from the candidate route set the route that achieves Pareto optimality in terms of safety, cost, and time. This means that no alternative route is superior in any one objective and not inferior to the others.

[0087] In some specific embodiments, safety score calculation involves constructing a road safety facility completeness indicator system. For example, parameters such as guardrail integrity, warning sign density, and emergency lane spacing are quantified for each candidate road, with a value range of 0-1, where 0 indicates no relevant facilities and 1 indicates complete facilities. The matching degree between the coverage radius of emergency rescue resources and the type of hazardous chemicals can be determined through buffer zone analysis. For example, for flammable liquids, a requirement is that there must be at least one fire station equipped with a foam fire truck within every 50 kilometers of the route. The matching degree is assigned based on the actual coverage, with full coverage being 1, partial coverage being 0.5, and no coverage being 0.

[0088] Among them, the weighted formula is: Safety score = 0.6 × facility completeness + 0.4 × emergency matching degree.

[0089] Total transportation cost calculation: Route length is obtained using the GIS map API to obtain the shortest distance between two points, in kilometers. Tolls, including highway tolls calculated by vehicle type and mileage, and temporary road control fees, are available in real time through the transportation department's public data interface. Risk insurance costs are calculated using a model provided by the insurance company. For example, risk insurance cost = base premium × hazardous chemical risk factor × route accident rate. The hazardous chemical risk factor can be set as: 2.0 for highly toxic, 1.5 for flammable, and 1.0 for common. Route accident rates can be calculated based on historical data.

[0090] Estimated travel time calculation: Real-time traffic flow data can be converted into average speeds for road sections, such as 80 km / h in smooth traffic and 20 km / h in congested traffic. This is then combined with route length to calculate basic travel time. Weather data influencing factors can be preset, such as a 30% speed reduction on all road sections in heavy rain and a 50% speed reduction on bridge sections in icy and snowy weather. Estimated travel time is dynamically adjusted through overlay calculations.

[0091] In some application scenarios, such as the cross-regional transportation of high-risk hazardous chemicals, such as liquefied natural gas (LNG), the safety score can be increased to 0.7, requiring the route to pass through areas equipped with cryogenic storage tank emergency treatment stations. The multi-objective optimization algorithm prioritizes routes with a safety facility completeness rating ≥ 0.8 and an emergency coverage radius ≤ 30 kilometers. Even if this route is 15% longer than the shortest path, it can reduce risk insurance costs by 25%, resulting in better overall benefits.

[0092] In an urban delivery scenario involving emergency medical and hazardous chemicals, if the delivery window is 2 hours, the estimated delivery time weight is increased to 0.6. The algorithm captures real-time traffic flow data and dynamically adjusts candidate routes through the city. For example, it might choose a route that is 3 kilometers longer but with a 15% higher real-time speed. It also uses weather data to avoid sections of road expected to experience heavy rain, ensuring that the delivery time is met and the safety score remains above a baseline threshold, such as 0.5.

[0093] In the low-cost chemical raw material transportation scenario, for common corrosive chemicals, the cost weight is set to 0.6, and the algorithm prioritizes national highways with the shortest route length and lowest tolls. Even if the emergency rescue coverage radius of this route is 80 kilometers, with a matching degree of 0.5, the total cost is 20% lower than the highway option due to the lower risk insurance cost (e.g., a risk factor of 1.0), achieving the optimal cost within an acceptable safety range.

[0094] Through the above-mentioned multi-objective optimization mechanism, transportation route planning can dynamically adjust the weights of each objective according to different types of hazardous chemicals and transportation needs, generate optimization plans that take into account safety, efficiency, and cost under complex road conditions, and effectively improve the comprehensive benefits of hazardous chemical transportation.

[0095] This approach, by integrating multiple iterations of transport tasks and road information and performing variance analysis, dynamically captures changes in key factors such as hazardous chemical type, transport preferences, and road conditions, providing more comprehensive training data and decision-making basis for the route planning model. The variance data generation and screening mechanism effectively filters redundant information, focusing on core variables that significantly impact transport safety and efficiency. Combined with dynamic weighting adjustments to transport demand, the model can adaptively prioritize different transport scenarios. The introduction of a multi-objective optimization algorithm systematically balances safety, transport costs, and estimated transport times. By quantifying safety indicators such as road safety facilities and emergency resource matching, and integrating dynamic forecasts of real-time traffic flow and weather data, a Pareto-optimal solution is selected from candidate routes. This approach avoids the limitations of single-objective optimization while enhancing safety measures for the high-risk nature of hazardous chemical transport, while maintaining a balanced balance between timeliness and cost. Furthermore, the model's iterative update mechanism responds in real time to dynamic events such as sudden changes in traffic flow, extreme weather warnings, and adjustments to emergency resource allocation, ensuring that route planning results are continuously optimized as the external environment changes, effectively mitigating potential risks and delays during transport. Overall, this solution improves the scientificity, flexibility and safety of hazardous chemical transportation route planning through data-driven intelligent decision-making and multi-dimensional goal coordination, providing efficient and reliable technical support for transportation scheduling in complex scenarios.

[0096] See also Figure 2 , Figure 2 This is a schematic diagram of the structural framework of an embodiment of a hazardous chemicals transportation route planning device based on multi-objective optimization. Figure 2As shown, a hazardous chemical transportation route planning device 20 based on multi-objective optimization includes: an acquisition module 21, a comparison module 22, a first response module 23, a second response module 24, and a generation module 25. The acquisition module 21 is configured to acquire multiple iterations of hazardous chemical transportation task information and target road information. The comparison module 22 is configured to compare the transportation task and road information of each iteration and generate difference data. The first response module 23 is configured to train a route planning model based on the difference data if the difference data of each iteration is consistent. The second response module 24 is configured to input the difference data of each iteration and the transportation demand into the model to generate final difference data if the difference data of each iteration is inconsistent. The generation module 25 is configured to generate an optimized transportation route using a multi-objective optimization algorithm based on the final difference data and current road information. The transportation task information includes the hazardous chemical type, transportation risk preference, and time window, and the road information includes safety facilities, real-time traffic flow, weather data, and emergency rescue resource distribution. Specifically, the difference data is composed of at least one difference element, which includes a change in hazardous chemical type, an adjustment in transportation risk preference, an update to road safety facilities, or a change in traffic flow threshold.

[0097] In some specific embodiments, if the current iteration version is the initial iteration version, the route planning model is trained based on the difference data of the current iteration version. The multiple iteration versions include the current iteration version and its historical iteration versions, and the historical iteration versions are all versions generated before the current iteration version.

[0098] In some embodiments, the second response module 24 inputs the difference data from each iteration into the route planning model, including: calculating the initial matching probability of each difference element in different iterations, which is generated based on the semantic similarity between the difference element and the transportation demand and historical matching records; dynamically weighting each initial matching probability based on the transportation demand to generate a target matching probability; and selecting difference elements whose target matching probability exceeds a preset threshold and combining them to generate the final difference data. The dynamic weighting includes: if transportation risk preference prioritizes cost over cost, then increasing the weight of difference elements related to road safety facilities and emergency rescue resources; if transportation time window prioritizes risk over risk, then increasing the weight of difference elements related to traffic flow and weather data.

[0099] In some embodiments, the optimization objectives of the multi-objective optimization algorithm of the generation module 25 include: safety score, total transportation cost, and estimated transportation time; wherein, the safety score is calculated by weighted calculation of the completeness of road safety facilities, the coverage radius of emergency rescue resources and the matching degree of hazardous chemicals types, the total transportation cost includes path length, tolls and risk insurance costs, and the estimated transportation time is based on dynamic prediction of real-time traffic flow and weather data; wherein, the optimized transportation route is selected from the candidate path set through multi-objective equilibrium analysis; and / or, the route planning model triggers iterative updates based on one or more of the following conditions: the rate of change of real-time traffic flow data in the target road information exceeds a preset threshold; extreme weather warnings are added to the weather data; and the distribution of emergency rescue resources undergoes structural adjustments.

[0100] This approach, by integrating multiple iterations of transport tasks and road information and performing variance analysis, dynamically captures changes in key factors such as hazardous chemical type, transport preferences, and road conditions, providing more comprehensive training data and decision-making basis for the route planning model. The variance data generation and screening mechanism effectively filters redundant information, focusing on core variables that significantly impact transport safety and efficiency. Combined with dynamic weighting adjustments to transport demand, the model can adaptively prioritize different transport scenarios. The introduction of a multi-objective optimization algorithm systematically balances safety, transport costs, and estimated transport times. By quantifying safety indicators such as road safety facilities and emergency resource matching, and integrating dynamic forecasts of real-time traffic flow and weather data, a Pareto-optimal solution is selected from candidate routes. This approach avoids the limitations of single-objective optimization while enhancing safety measures for the high-risk nature of hazardous chemical transport, while maintaining a balanced balance between timeliness and cost. Furthermore, the model's iterative update mechanism responds in real time to dynamic events such as sudden changes in traffic flow, extreme weather warnings, and adjustments to emergency resource allocation, ensuring that route planning results are continuously optimized as the external environment changes, effectively mitigating potential risks and delays during transport. Overall, this solution improves the scientificity, flexibility and safety of hazardous chemical transportation route planning through data-driven intelligent decision-making and multi-dimensional goal coordination, providing efficient and reliable technical support for transportation scheduling in complex scenarios.

[0101] Figure 3 This is a schematic diagram of the structural framework of an embodiment of a hazardous chemical transportation route planning device based on multi-objective optimization of the present application. The hazardous chemical transportation route planning device 30 based on multi-objective optimization may have relatively large differences due to different configurations or performances, and may include one or more processors 31 and memory 32. The processor 31 may be configured to communicate with the memory 32 to execute a series of instruction operations in the memory on the hazardous chemical transportation route planning device based on multi-objective optimization to implement the steps of the hazardous chemical transportation route planning method based on multi-objective optimization. Those skilled in the art will understand that Figure 3The structure of the hazardous chemical transportation route planning equipment based on multi-objective optimization shown does not constitute a limitation on the hazardous chemical transportation route planning equipment based on multi-objective optimization provided by the present invention, and may include more or fewer components than shown in the figure, or a combination of certain components, or a different arrangement of components.

[0102] The present invention also provides a computer-readable storage medium, which may be a non-volatile computer-readable storage medium or a volatile computer-readable storage medium. The computer-readable storage medium stores instructions, which, when executed on a computer, enable the computer to execute the steps of a method for planning a hazardous chemical transportation route based on multi-objective optimization.

[0103] Those skilled in the art will clearly understand that, for the convenience and brevity of description, the specific working processes of the above-described systems, devices, and units can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0104] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the portion that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions for causing a computer device (which can be a personal computer, server, or network device, etc.) to perform all or part of the steps of the various embodiments of the method of the present invention. The aforementioned storage medium includes various media that can store program code, such as USB flash drives, mobile hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0105] The above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the various embodiments of the present invention.

Claims

1. A hazardous chemicals transportation route planning method based on multi-objective optimization, characterized in that: The method comprises: Obtain multiple iterations of hazardous chemical transportation mission information and target road information; Compare the transportation tasks and road information of each iteration to generate difference data; In response to the difference data of each iterative version being consistent, training a route planning model based on the difference data; In response to inconsistencies in the difference data of each iterative version, the difference data of each version and the transportation demand are input into the model to generate final difference data; Based on the final difference data and current road information, generating an optimized transportation route through a multi-objective optimization algorithm; Among them, the transportation task information includes the type of hazardous chemicals, transportation risk preference and time window, and the road information includes safety facilities, real-time traffic flow, weather data and emergency rescue resource distribution.

2. The method for planning hazardous chemicals transportation routes based on multi-objective optimization according to claim 1 is characterized in that: If the current iteration version is the initial iteration version, the route planning model is trained based on the difference data of the current iteration version.

3. The method for planning hazardous chemicals transportation routes based on multi-objective optimization according to claim 1 is characterized in that: The multiple iterative versions include a current iterative version and its historical iterative versions, where the historical iterative versions are all versions generated before the current iterative version.

4. The method for planning hazardous chemicals transportation routes based on multi-objective optimization according to claim 1 is characterized in that: The difference data consists of at least one difference element, which includes changes in hazardous chemical types, adjustments to transportation risk preferences, updates to road safety facilities, and changes in traffic flow thresholds.

5. The method for planning hazardous chemicals transportation routes based on multi-objective optimization according to claim 4 is characterized in that: Inputting the difference data of each iterative version into the route planning model includes: Calculating the initial matching probability of each difference element in different iteration versions, wherein the initial matching probability is generated based on the semantic similarity between the difference element and the transportation demand and the historical matching records; Dynamically weight each initial matching probability according to transportation demand to generate a target matching probability; The difference elements whose target matching probability exceeds a preset threshold are selected and combined to generate the final difference data.

6. The method for planning hazardous chemicals transportation routes based on multi-objective optimization according to claim 5 is characterized in that: The dynamic weighting includes: If transport risk preference takes priority over cost, then the weight of differentiation elements related to road safety facilities and emergency rescue resources will be increased; If the transportation time window has a higher priority than the risk, the weight of the difference elements related to traffic flow and weather data will be increased.

7. The method for planning hazardous chemicals transportation routes based on multi-objective optimization according to claim 1 is characterized in that: The optimization objectives of the multi-objective optimization algorithm include: safety score, total transportation cost, and estimated transportation time; wherein the safety score is calculated by weighting the completeness of road safety facilities, the coverage radius of emergency rescue resources, and the matching degree of hazardous chemical types; the total transportation cost includes the path length, tolls, and risk insurance costs; and the estimated transportation time is based on dynamic predictions of real-time traffic flow and weather data; wherein the optimized transportation route is selected from a set of candidate paths through a multi-objective equilibrium analysis; and / or, The route planning model triggers iterative updates based on one or more of the following conditions: The change rate of real-time traffic flow data in the target road information exceeds the preset threshold; New extreme weather warnings have been added to weather data; There has been a structural adjustment in the distribution of emergency rescue resources.

8. A hazardous chemicals transportation route planning device based on multi-objective optimization, characterized in that: The hazardous chemicals transportation route planning device based on multi-objective optimization includes: Acquisition module: used to obtain multiple iterations of hazardous chemicals transportation task information and target road information; Comparison module: used to compare the transportation tasks and road information of each iteration version and generate difference data; A first response module: configured to, in response to the difference data of each iterative version being consistent, train a route planning model based on the difference data; The second response module is used to respond to inconsistencies in the difference data of each iterative version, input the difference data of each version and the transportation demand into the model, and generate the final difference data; A generation module is configured to generate an optimized transportation route based on the final difference data and current road information through a multi-objective optimization algorithm; Among them, the transportation task information includes the type of hazardous chemicals, transportation risk preference and time window, and the road information includes safety facilities, real-time traffic flow, weather data and emergency rescue resource distribution.

9. A hazardous chemicals transportation route planning device based on multi-objective optimization, characterized in that: The hazardous chemicals transportation route planning device based on multi-objective optimization includes: a memory and at least one processor, wherein the memory stores instructions; The at least one processor calls the instructions in the memory so that the hazardous chemicals transportation route planning device based on multi-objective optimization performs the steps of the hazardous chemicals transportation route planning method based on multi-objective optimization as described in any one of claims 1-7.

10. A computer-readable storage medium having instructions stored thereon, characterized in that: When the instruction is processed and executed, the steps of the hazardous chemical transportation route planning method based on multi-objective optimization as described in any one of claims 1 to 7 are implemented.

Images