A decision support method for approval of large-scale transportation activities

Abstract

The present invention relates to the field of safety management technology, and in particular to a decision support method for the approval of large-scale transportation activities. The method comprises the following steps: obtaining historical transportation data of large-scale transportation companies and evaluating them to obtain credit evaluation data of transportation vehicles; obtaining real-time transportation vehicle data, and performing vehicle behavior detection of reporting critical values ​​to obtain reporting critical value behavior detection data; performing driver license credit evaluation based on license scanning data to obtain driver license credit evaluation data; constructing a risk assessment model based on transportation vehicle credit evaluation data, reporting critical value behavior detection data, and driver license credit evaluation data; dividing real-time transportation vehicle data into high-risk transportation vehicles to obtain high-risk transportation vehicle data; analyzing high-risk transportation vehicle data to obtain high-risk mitigation transportation strategies, and performing risk reduction to obtain low-risk transportation strategies. The present invention can improve the safety management level of large-scale transportation activities.

Description

Technical Field

[0001] The invention relates to the technical field of safety management, and in particular to a decision support method for approval of large-scale transportation activities. Background Art

[0002] With the development of the global economy and the increase in trade activities, the demand for large-scale transportation activities is growing. Large-scale transportation activities involve goods that are overlong, overwide, overhigh or overweight, such as engineering equipment, mechanical equipment, construction materials, etc. The transportation of these goods usually requires special vehicles, equipment and route planning to ensure the safe and efficient completion of transportation tasks. However, there are a series of challenges in large-scale transportation activities, including but not limited to route selection, traffic control, safety risks and environmental impacts. In order to manage and monitor these transportation activities, many countries and regions have formulated relevant laws and regulations requiring that large-scale transportation activities must be approved before they can be carried out. In the past, the approval of large-scale transportation activities was usually handled manually and with paper documents, which was cumbersome, inefficient, and prone to information asymmetry. Summary of the invention

[0003] Based on this, it is necessary for the present invention to provide a decision support method for the approval of large-scale transportation activities to solve at least one of the above technical problems.

[0004] To achieve the above purpose, a decision support method for large-scale transportation activity permit approval includes the following steps:

[0005] Step S1: Acquire historical transportation data of large-scale transportation companies, and conduct transportation vehicle credit assessment on the historical transportation data of large-scale transportation companies, thereby obtaining transportation vehicle credit assessment data;

[0006] Step S2: acquiring real-time transport vehicle data, extracting vehicle sensor data and license scanning data from the real-time transport vehicle data, thereby obtaining real-time vehicle sensor data and real-time license scanning data; performing declaration threshold vehicle behavior detection based on the real-time vehicle sensor data and the real-time license scanning data, thereby obtaining declaration threshold behavior detection data;

[0007] Step S3: performing a driver's license credit assessment based on the license scan data, thereby obtaining driver's license credit assessment data;

[0008] Step S4: Conducting a large-scale transportation risk assessment on the transport vehicle credit assessment data, the declared critical value behavior detection data, and the driver's license credit assessment data, thereby obtaining large-scale transportation risk assessment data, and constructing a risk assessment model based on the large-scale transportation risk assessment data;

[0009] Step S5: Based on the large-scale transportation risk assessment data, the real-time transportation vehicle data is classified into high-risk transportation vehicles, thereby obtaining high-risk transportation vehicle data, and uploading the high-risk transportation vehicle data to the large-scale transportation activity management cloud platform to perform high-risk early warning tasks;

[0010] Step S6: Perform risk mitigation strategy analysis on high-risk transport vehicle data based on real-time transport vehicle data to obtain a high-risk mitigation transport strategy, and iteratively reduce the transport strategy risk of the high-risk mitigation transport strategy through a risk assessment model to obtain a low-risk transport strategy, and upload the low-risk transport strategy to the large-scale transport activity management cloud platform to perform decision support tasks.

[0011] The present invention can evaluate the credit of its transport vehicles by analyzing the historical transport data of large-scale transport companies. This helps to determine which companies' vehicles can transport large-scale goods more safely, thereby improving the efficiency and accuracy of approval. By acquiring real-time transport vehicle data and sensor data, the system can monitor the status and behavior of the vehicle in a timely manner, such as vehicle speed, route, etc. Combined with license scanning data, the driver's identity and qualifications can be verified. This helps to detect abnormal situations and violations in a timely manner, and improve the safety and accuracy of the approval process. By evaluating the driver's license information, his driving record and qualifications can be determined. This helps to screen out qualified drivers, reduce safety hazards caused by driver problems, and improve the safety and reliability of approval. Combined with information such as transport vehicle credit evaluation, behavior detection data, and driver license credit evaluation, the system can perform a comprehensive large-scale transportation risk assessment. This helps to identify potential risks and safety hazards and provide a basis for subsequent decision-making. By identifying and classifying high-risk transport vehicles, the system can issue early warnings to relevant departments in a timely manner, take necessary measures to reduce potential risks and safety issues, and ensure the smooth progress of the approval process. By analyzing high-risk transport vehicle data, the system can formulate corresponding risk mitigation strategies and iterate and optimize through risk assessment models. This helps reduce risks and uncertainties in the approval process and improve the success rate and safety of approval.

[0012] Optionally, step S1 specifically includes:

[0013] Step S11: Obtain historical transportation data of large-scale transportation companies;

[0014] Step S12: extracting historical transportation violation records, historical transportation accident records and complaint records from the historical transportation data of the large-scale transportation company, thereby obtaining historical transportation violation data, historical transportation accident data and historical transportation complaint data;

[0015] Step S13: evaluating the accident liability of a transport vehicle based on historical transport accident data, thereby obtaining transport vehicle accident liability evaluation data;

[0016] Step S14: performing a transportation vehicle complaint analysis based on historical transportation complaint data, thereby obtaining transportation vehicle complaint data;

[0017] Step S15: Counting the number of violations of transport vehicles based on historical transport violation data, thereby obtaining the number of violations of transport vehicles;

[0018] Step S16: Conduct a credit assessment of the transport vehicle based on the transport vehicle accident liability assessment data, the transport vehicle complaint data, and the transport vehicle violation quantity data, thereby obtaining the transport vehicle credit assessment data.

[0019] The present invention obtains the historical transportation data of the large-scale transportation company, and provides a data basis for subsequent evaluation and analysis. These data include transportation routes, transportation time, cargo types, vehicle information, etc., which provide basic data for evaluating the operation of the transportation company. By extracting historical transportation violation records, accident records and complaint records, the system can analyze the historical safety records of the transportation company. This helps to evaluate the safety management level and service quality of the transportation company, and provides data support for subsequent credit evaluation. By analyzing historical transportation accident data, the system can evaluate the degree of responsibility of the transportation vehicle in the accident. This helps to determine the safety record and risk level of the vehicle, and provides a reference for subsequent credit evaluation. By analyzing historical transportation complaint data, the system can understand the service quality and customer satisfaction of the transportation vehicle. This helps to evaluate the service level and credibility of the transportation company, and provides a reference for credit evaluation. Statistical analysis of historical transportation violation data can understand the violation situation and violation frequency of the vehicle. This helps to evaluate the compliance and safety of the vehicle, and provides a reference for credit evaluation. Taking into account the accident responsibility evaluation, complaint analysis and violation statistics of the transportation vehicle, the system can evaluate the credit of the transportation vehicle. This helps determine which vehicles are more reliable and safer, providing a credit assessment basis for the approval process.

[0020] Optionally, step S13 is specifically:

[0021] Step S131: extracting accident responsibility features from historical transportation accident data, thereby obtaining historical accident responsibility data;

[0022] Step S132: classifying the historical accident responsibility data into primary and secondary responsibility data, thereby obtaining the historical accident primary responsibility data and the historical accident secondary responsibility data;

[0023] Step S133: performing statistics on the main responsible drivers for the historical accident main responsible data, thereby obtaining data on the main responsible drivers for high-frequency accidents and data on the main responsible drivers for low-frequency accidents;

[0024] Step S134: performing statistics on the secondary responsibility driver data for the historical accident secondary responsibility data, thereby obtaining the high-frequency secondary responsibility accident driver data and the low-frequency secondary responsibility accident driver data;

[0025] Step S135: performing a driver data intersection operation on the low-frequency primary accident driver data and the low-frequency secondary accident driver data to obtain low-frequency accident driver data, and performing vehicle data association on the low-frequency accident driver data and the historical transportation data of the heavy-duty transportation company to obtain low-frequency accident vehicle data;

[0026] Step S136: performing a driver data intersection operation on the high-frequency primary accident driver data and the high-frequency secondary accident driver data to obtain high-frequency accident driver data, and performing vehicle data association on the high-frequency accident driver data and the historical transportation data of the large-scale transportation company to obtain high-frequency accident vehicle data;

[0027] Step S137: Merge the data of vehicles with low liability in accidents and vehicles with high liability in accidents to obtain accident liability assessment data of transport vehicles.

[0028] The present invention extracts the responsibility characteristics of historical transportation accident data, and the system can identify the main responsible party and the secondary responsible party in the accident. This helps to more accurately assess the responsibility of the accident and provides a basis for subsequent data processing. Classifying the historical accident responsibility data into primary and secondary responsibilities helps to distinguish the main responsibility from the secondary responsibility and further refine the assessment of the accident responsibility. Such classification can provide more accurate data for subsequent statistics and analysis. By counting the data of the main responsible driver, the system can identify the driver who bears the main responsibility in the accident. This helps to determine the high-frequency and low-frequency main responsible accident drivers, providing an important basis for subsequent evaluation and management. Statistics on the data of the secondary responsible driver help to understand the role and frequency of the secondary responsible party in the accident. This can provide a reference for subsequent driver training and improvement. Through intersection operations and vehicle data association, identifying low-frequency main responsible accident drivers and associating them with vehicle information in the historical transportation data of large-scale transportation companies can help determine the drivers with lower responsibility in the accident and the vehicles they drive. This helps to carry out targeted training and management. Similarly, data association of high-frequency main responsible accident drivers can determine the drivers with higher responsibility in the accident and the vehicles they drive. This provides a basis for taking further management measures. By combining the data of vehicles with low responsibility for accidents and vehicles with high responsibility for accidents, the system can comprehensively evaluate the accident responsibility of transport vehicles. This provides a comprehensive perspective to help decision makers understand the safety status of vehicles and the distribution of responsibility more comprehensively.

[0029] Optionally, step S14 is specifically:

[0030] Step S141: performing frequency statistics of complaint vehicles on historical transportation complaint data, thereby obtaining high-frequency complaint vehicle data and low-frequency complaint vehicle data;

[0031] Step S142: extracting complaint transport features from historical transport data of large-scale transport companies based on the high-frequency complaint vehicle data, thereby obtaining high-frequency complaint transport feature data;

[0032] Step S143: extracting complaint transport features from historical transport data of large-scale transport companies based on low-frequency complaint vehicle data, thereby obtaining low-frequency complaint transport feature data;

[0033] Step S144: constructing a potential complaint behavior detection model based on the high-frequency complaint transportation feature data and the low-frequency complaint transportation feature data;

[0034] Step S145: identifying potential complaint vehicles based on historical transportation complaint data using a potential complaint behavior detection model, thereby obtaining potential complaint vehicle data;

[0035] Step S146: Merge the potential complaint vehicle data, high-frequency complaint transport characteristic data, and low-frequency complaint transport characteristic data to obtain transport vehicle complaint data.

[0036] The present invention can identify vehicles that are frequently complained (high-frequency complaint vehicles) and vehicles that are relatively less complained (low-frequency complaint vehicles) by performing frequency statistics on historical transportation complaint data. This helps to focus attention on vehicles with problems. For high-frequency complaint vehicles and low-frequency complaint vehicles, the complaint transportation features in their historical transportation data are extracted respectively. These features include transportation routes, transportation time, loading conditions, driver information, etc. By extracting these features, it can help identify the key factors that lead to complaints. Combined with high-frequency and low-frequency complaint transportation feature data, a potential complaint behavior detection model is established. This model can identify transportation behavior patterns with potential complaint behaviors based on machine learning or statistical methods. Through such a model, potential problem vehicles can be automatically identified and the detection efficiency of complaint behaviors can be improved. Using the constructed potential complaint behavior detection model, historical transportation complaint data is analyzed and identified to identify vehicles with potential complaint behaviors. This helps to detect vehicles with problems early and take appropriate measures to manage and improve them. The potential complaint vehicle data is merged with the high-frequency and low-frequency complaint transportation feature data to obtain complete transportation vehicle complaint data. Such a dataset can be used for further analysis, such as determining the main factors of complaint behavior and formulating improvement measures.

[0037] Optionally, step S144 is specifically:

[0038] Conduct principal component analysis based on the high-frequency complaint transportation characteristic data to obtain the high-frequency complaint transportation principal component characteristic data;

[0039] Calculate the complaint interval for low-frequency complaint transport characteristic data to obtain complaint time interval data;

[0040] Perform repeated complaint data correction on low-frequency complaint transport characteristic data according to complaint time interval data, thereby obtaining low-frequency complaint transport correction data;

[0041] Conduct principal component analysis on the low-frequency complaint transportation correction data to obtain the principal component characteristic data of low-frequency complaint transportation;

[0042] Performing transportation characteristic intersection operation on the low-frequency complaint transportation principal component characteristic data and the high-frequency complaint transportation principal component characteristic data, thereby obtaining transportation complaint characteristic data;

[0043] A potential complaint behavior detection model is constructed based on transportation complaint feature data and historical transportation complaint data.

[0044] The present invention can convert a high-dimensional feature space into a lower-dimensional principal component space through principal component analysis, thereby reducing the complexity of the data and extracting the principal component feature data that best explains the variance. This helps to reduce the redundancy of the data and better understand the transportation characteristics of vehicles with high frequency complaints. Similarly, principal component analysis can be performed on the low-frequency complaint transportation correction data to identify the transportation characteristics that mainly affect low-frequency complaints and help understand the transportation mode of low-frequency complaint vehicles. By calculating the complaint time interval data of low-frequency complaint vehicles, the time interval between complaint events can be analyzed. This helps to determine the trend and pattern between complaint events and identify whether there is a fixed repeated complaint pattern. By correcting the low-frequency complaint transportation feature data according to the complaint time interval data, the complaint situation can be more accurately reflected and the same complaint event can be avoided from being repeatedly calculated. The common characteristics of transportation complaints can be obtained by performing an intersection operation on the high-frequency complaint transportation principal component feature data and the low-frequency complaint transportation principal component feature data. This helps to identify the common transportation characteristics between vehicles with different complaint frequencies, so as to better understand the root cause of the complaint behavior. A potential complaint behavior detection model is constructed based on the transportation complaint feature data and the historical transportation complaint data, and a machine learning or statistical method can be used to identify the pattern of complaint behavior. This helps to identify potential problems in advance and take appropriate measures to intervene and improve, thereby improving the overall transportation service quality and customer satisfaction.

[0045] Optionally, step S16 is specifically:

[0046] Calculate the primary and secondary liability ratios of transport vehicle accidents based on the transport vehicle accident liability assessment data, thereby obtaining the primary and secondary liability ratios of transport vehicle accidents;

[0047] Complaints are counted based on the transport vehicle complaint data, thereby obtaining the transport vehicle complaint data;

[0048] A credit assessment model for transport vehicles is built based on the data on the primary and secondary liability ratios of transport vehicle accidents, the complaint resolution rate of transport vehicles, and the number of violations of transport vehicles;

[0049] The transport vehicle credit assessment model is used to conduct a transport vehicle credit assessment on the historical transport data of the large-scale transport company, thereby obtaining the transport vehicle credit assessment data.

[0050] The present invention can understand the proportion of responsibility of transport vehicles in accidents by calculating the primary and secondary responsibility ratios of transport vehicle accident responsibility assessment data. This helps to determine which vehicles are more prone to accidents and which responsibilities are more important, thereby guiding the formulation of safety management strategies and risk prevention measures. Based on the transport vehicle complaint data, the historical transport complaint data can be quantitatively counted to analyze the complaints and frequencies of different vehicles. This can help identify which vehicles have problems or potential risks, and the urgency of resolving complaints. By combining the primary and secondary responsibility ratio data of transport vehicle accidents, the complaint resolution rate data and the number of violations data to construct a credit assessment model, the safety, service quality and compliance of the vehicle can be comprehensively considered. Such an assessment model can provide decision support for companies to select suitable transport vehicles and reduce transportation risks. By evaluating historical transport data through the transport vehicle credit assessment model, vehicles and companies with higher credit can be identified. This helps to establish long-term cooperative relationships, improve transportation efficiency and safety, and reduce risks and losses caused by bad vehicles.

[0051] Optionally, step S2 specifically includes:

[0052] Step S21: acquiring real-time transport vehicle data, extracting vehicle sensor data and license scanning data from the real-time transport vehicle data, thereby obtaining real-time vehicle sensor data and real-time license scanning data;

[0053] Step S22: extracting the declared license data from the real-time license scan data, thereby obtaining the real-time declared license data, and converting the real-time declared license data into text, thereby obtaining the declared text data;

[0054] Step S23: extracting the declared load weight feature and the declared driving feature from the declared text data, thereby obtaining the declared load weight data and the declared driving data;

[0055] Step S24: performing a declared critical value overload behavior detection on the real-time vehicle sensor data based on the declared load data, thereby obtaining declared critical value overload behavior detection data;

[0056] Step S25: performing a reported critical value overspeeding behavior detection on the real-time vehicle sensor data based on the reported load data and the reported travel data, thereby obtaining reported critical value overspeeding behavior detection data;

[0057] Step S26: Merge the declared critical value overload behavior detection data and the declared critical value overspeed behavior detection data to obtain the declared critical value behavior detection data.

[0058] The present invention can timely understand the running status and behavior of the vehicle by acquiring real-time transport vehicle data. Extracting vehicle sensor data and license scanning data is helpful to monitor the real-time status and compliance of the vehicle. For example, the sensor data may include information such as the speed, acceleration, and vehicle offset of the vehicle, while the license scanning data may include license information such as the registration certificate and driving license of the vehicle. By extracting the real-time license scanning data and performing text conversion, the real-time license information of the vehicle, such as the registration certificate and driving license, can be obtained. This information is essential for monitoring the legality and compliance of the vehicle. Feature extraction of the declared text data, such as extracting load data and driving data, helps to understand the load situation and driving route of the vehicle. This information is very important for evaluating the rationality and safety of vehicle transportation behavior. By performing overloading behavior detection based on the declared load data and speeding behavior detection based on the load and driving data, the violation of the vehicle can be discovered in time. This helps to prevent traffic accidents and improve transportation safety. By merging the overloading and speeding behavior detection data, the violation of the vehicle can be comprehensively evaluated and necessary measures can be taken in time. This comprehensive analysis can better understand the behavior pattern of the vehicle and provide a reference for transportation management.

[0059] Optionally, step S25 is specifically:

[0060] Step S251: extracting the reported travel characteristics and the reported travel vehicle characteristics from the reported travel data, thereby obtaining the reported travel data and the reported travel vehicle data;

[0061] Step S252: predicting the vehicle speed according to the reported travel data, the reported vehicle data and the reported load data, thereby obtaining the vehicle speed data;

[0062] Step S253: obtaining the trip turning section data, and calculating the turning speed of large-scale transportation according to the declared driving vehicle data and the trip turning section data, thereby obtaining the vehicle transportation turning speed data;

[0063] Step S254: performing a road section vehicle speed critical value statistical analysis on the vehicle travel speed data and the vehicle transportation turning speed data, thereby obtaining estimated road section vehicle speed critical value data;

[0064] Step S255: extracting vehicle speed sensing features from the real-time vehicle sensing data, thereby obtaining vehicle speed sensing data;

[0065] Step S256: classifying the vehicle speed sensing data into travel speeds based on the estimated vehicle speed critical value data, thereby obtaining overspeed travel data and underspeed travel data;

[0066] Step S257: obtaining the speed limit data of the travel section, and making reasonable overspeed correction to the overspeed travel data according to the speed limit data of the travel section, thereby obtaining the overspeed travel correction data; making reasonable underspeed correction to the underspeed travel data according to the speed limit data of the travel section, thereby obtaining the underspeed travel correction data;

[0067] Step S258: Merge the overspeed stroke correction data and the underspeed stroke correction data to report critical value overspeed behavior detection data.

[0068] The present invention can better understand the driving mode and characteristics of the vehicle by extracting the travel characteristics and vehicle characteristics of the declared driving data. This information can be used for speed prediction and turning speed calculation in subsequent steps. The travel speed of the vehicle is predicted by combining the travel data, the driving vehicle data and the load data. This helps to predict the driving speed of the vehicle under different conditions, so as to detect overspeeding or underspeed behavior in advance. The travel turning section data is obtained, and the turning speed of large-scale transportation is calculated according to the vehicle characteristics. This can help to evaluate the safety and stability of the vehicle when turning, so as to reduce the risk of accidents. The vehicle travel speed data and the turning speed data are analyzed to determine the expected road section speed critical value. This helps to identify the existing overspeeding or underspeed behavior and formulate corresponding management and control strategies. The speed sensing characteristics are extracted from the real-time vehicle sensor data to obtain the current speed information of the vehicle. According to the expected road section speed critical value, the real-time speed data of the vehicle is classified to identify overspeeding and underspeed travel. This helps to timely discover and record non-compliant behavior. The travel section speed limit data is obtained, and the overspeeding and underspeed travel data are reasonably corrected according to the speed limit data. This ensures the accuracy and reliability of the trip data. The corrected overspeed and underspeed trip data are combined to form the reporting threshold overspeed behavior detection data. This is a key step in the comprehensive evaluation and monitoring of vehicle driving behavior, which helps managers take necessary measures in a timely manner.

[0069] Optionally, step S3 specifically includes:

[0070] Step S31: extracting driver's license data from the license scan data to obtain driver's license data, and converting the driver's license data into text to obtain driver's license text data;

[0071] Step S32: extracting the driver's license feature and the vehicle license feature from the driver's license text data, thereby obtaining the driver's license text data and the vehicle license text data;

[0072] Step S33: Calculate the name similarity of the driver's license text data and the vehicle license text data, so as to obtain the driver's license name similarity data;

[0073] Step S34: Calculating the similarity of the declared driver's name on the declared text data, the driver's license text data, and the vehicle driving license text data, thereby obtaining the declared driver's name similarity data;

[0074] Step S35: Perform credit evaluation based on the name similarity data of the driver's license and the similarity data of the declared driver's name, so as to obtain the driver's license credit evaluation data.

[0075] The present invention converts the driver's license data into processable text data by extracting and converting the license scan data, which helps to make subsequent data processing and analysis more efficient and accurate. Feature extraction is performed on the driver's license text data and the vehicle driving license text data to obtain key information. These features include name, certificate number, validity period, etc., which are helpful for data matching and analysis in subsequent steps. By calculating the similarity of the names of the driver's license text data and the vehicle driving license text data, the consistency and accuracy of the license information can be evaluated. This helps to identify the existing license information errors or fraudulent use. The declared driver's name similarity calculation is performed on the declared text data, the driver's license text data and the vehicle driving license text data to further verify the consistency of the declared name information with the license information. This can reduce the situation where the driver's identity information does not match. Credit evaluation is performed based on the driver's license name similarity data and the declared driver name similarity data. This step will comprehensively consider the consistency of the license information and the declared information to evaluate the driver's credit level and compliance. This helps to screen out drivers with higher credit and improve the safety and reliability of vehicle transportation.

[0076] Optionally, step S6 specifically includes:

[0077] Step S61: extracting vehicle position features from real-time transport vehicle data, thereby obtaining real-time transport vehicle position data;

[0078] Step S62: analyzing the vehicle driving condition based on the vehicle sensor data, thereby obtaining the vehicle driving condition data;

[0079] Step S63: screening the vehicle stop location based on the vehicle driving road condition data and the real-time transport vehicle location data, thereby obtaining the vehicle stop location data;

[0080] Step S64: performing load optimization analysis on real-time transport vehicle data based on the declared critical value behavior detection data, thereby obtaining vehicle load reduction data;

[0081] Step S65: analyzing the vehicle driving strategy according to the vehicle stop location data and the vehicle load reduction data, thereby obtaining a high-risk mitigation transportation strategy;

[0082] Step S66: Iteratively reduce the risk of the high-risk mitigation transportation strategy through the risk assessment model to obtain a low-risk transportation strategy, and upload the low-risk transportation strategy to the large-scale transportation activity management cloud platform to perform decision support tasks.

[0083] The present invention obtains the real-time location data of the vehicle by extracting the location features in the real-time transport vehicle data. This makes it feasible to monitor the location of the vehicle and provides a data basis for subsequent route planning and risk analysis. Based on the vehicle sensor data, the vehicle driving road condition is analyzed, which helps to understand the driving conditions of the vehicle under different road conditions, including road flatness, congestion, etc., and provides a reference for the subsequent vehicle driving strategy. Combined with the vehicle driving road condition data and the real-time vehicle location data, the vehicle parking location is screened out. This helps to identify suitable parking locations, optimize transportation routes, and improve transportation efficiency. Based on the declared critical value behavior detection data, the real-time transport vehicle data is optimized for load capacity analysis. This helps to determine the actual load capacity of the vehicle and improve transportation efficiency and safety. Combined with the vehicle parking location data and the vehicle load reduction data, the vehicle driving strategy analysis is carried out to find strategies to reduce high-risk transportation, such as selecting a safe parking point or adjusting the load capacity. The high-risk mitigation transportation strategy is evaluated and optimized through the risk assessment model. This helps to reduce potential transportation risks and provide low-risk transportation strategies. After uploading to the large-scale transportation activity management cloud platform, better decision support and coordination management can be achieved. BRIEF DESCRIPTION OF THE DRAWINGS

[0084] Other features, objects and advantages of the present invention will become more apparent from the detailed description of non-limiting embodiments thereof made with reference to the following drawings:

[0085] Figure 1 A schematic flow chart of the steps of the decision support method for the approval of large-scale transportation activities according to the present invention;

[0086] Figure 2 Detailed step flow diagram of step S1 in the present invention;

[0087] Figure 3 Detailed step flow diagram of step S2 in the present invention;

[0088] Figure 4 Detailed step flow diagram of step S3 in the present invention;

[0089] The realization of the purpose, functional features and advantages of the present invention will be further explained in conjunction with embodiments and with reference to the accompanying drawings. DETAILED DESCRIPTION

[0090] The following is a clear and complete description of the technical method of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by technicians in this field without creative work are within the scope of protection of the present invention.

[0091] In addition, the accompanying drawings are only schematic illustrations of the present invention and are not necessarily drawn to scale. The same reference numerals in the figures represent the same or similar parts, and their repeated description will be omitted. Some of the block diagrams shown in the accompanying drawings are functional entities and do not necessarily correspond to physically or logically independent entities. The functional entities can be implemented in software form, or implemented in one or more hardware modules or integrated circuits, or implemented in different networks and / or processor methods and / or microcontroller methods.

[0092] It should be understood that, although the terms "first", "second", etc. may be used herein to describe various units, these units should not be limited by these terms. These terms are used only to distinguish one unit from another unit. For example, without departing from the scope of the exemplary embodiments, the first unit may be referred to as the second unit, and similarly the second unit may be referred to as the first unit. The term "and / or" used herein includes any and all combinations of one or more of the listed associated items.

[0093] To achieve this, please refer to Figures 1 to 4 The present invention provides a decision support method for large-scale transportation activity licensing approval, the method comprising the following steps:

[0094] Step S1: Acquire historical transportation data of large-scale transportation companies, and conduct transportation vehicle credit assessment on the historical transportation data of large-scale transportation companies, thereby obtaining transportation vehicle credit assessment data;

[0095] In this embodiment, the historical transportation data of the large-scale transportation company including the driving record, loading condition, route selection and other information of the transportation vehicle is obtained through the large-scale transportation activity management cloud platform. The historical transportation records of each transportation vehicle are comprehensively analyzed through data mining technology including machine learning algorithms and statistical models. For example, the data on the on-time delivery record, accident report, route selection, violation record and other aspects of the transportation vehicle are comprehensively analyzed. Based on these analysis results, a comprehensive vehicle credit assessment model is established to quantitatively assess the credit level of each transportation vehicle. This vehicle credit assessment model can use scoring cards, logistic regression or decision trees and other technologies to calculate the credit score of the vehicle, thereby obtaining the credit assessment data of the transportation vehicle.

[0096] Step S2: acquiring real-time transport vehicle data, extracting vehicle sensor data and license scanning data from the real-time transport vehicle data, thereby obtaining real-time vehicle sensor data and real-time license scanning data; performing declaration threshold vehicle behavior detection based on the real-time vehicle sensor data and the real-time license scanning data, thereby obtaining declaration threshold behavior detection data;

[0097] In this embodiment, real-time transport vehicle data including license scanning results, vehicle location information, vehicle speed and other data are obtained through equipment such as license scanners, GPS, accelerometers, and vehicle speed sensors deployed on the vehicle, and then the real-time transport vehicle data is transmitted to the large-scale transportation activity management cloud platform through the network for data mining. Then, the sensor data including the cab image and vehicle speed in the real-time transport vehicle data are extracted to obtain the real-time vehicle sensor data. At the same time, OCR technology can be used to extract the driver's name, driver's license number and other information from the vehicle license to obtain real-time license scanning data. Use real-time sensor data to analyze the behavior of the vehicle, such as detecting bad driving behaviors such as sudden acceleration, sudden braking, and speeding. Set a critical value and monitor whether the vehicle behavior exceeds the preset range.

[0098] Step S3: performing a driver's license credit assessment based on the license scan data, thereby obtaining driver's license credit assessment data;

[0099] In this embodiment, after collecting the driver's license scan data, image processing and text analysis technology are used to extract and identify the information of the image obtained after the license scan. Then, the driver's credit assessment is performed using this information. During the assessment process, factors such as whether the driver's information is inconsistent with the driver's information recorded in the driver's license will be considered, and whether there is any behavior of impersonating the driver's license or not carrying the driver's license will be detected. Then, taking into account the driver's driving age, violation record, accident record, and the situation of impersonating the driver's license, a driver credit assessment model can be established using algorithms such as logistic regression or neural networks to assess the driver's credit level and generate corresponding credit assessment data.

[0100] Step S4: Conducting a large-scale transportation risk assessment on the transport vehicle credit assessment data, the declared critical value behavior detection data, and the driver's license credit assessment data, thereby obtaining large-scale transportation risk assessment data, and constructing a risk assessment model based on the large-scale transportation risk assessment data;

[0101] In this embodiment, data mining and statistical analysis techniques are used to integrate and analyze the credit assessment data of transport vehicles, the reporting threshold behavior detection data, and the driver's license credit assessment data. The specific analysis method may include establishing a statistical model, a machine learning model, or an expert system to assess the risk level of large-scale transportation. Based on these analysis results, large-scale transportation risk assessment data can be obtained. Then, through machine learning techniques such as random forests and decision tree algorithms, a large-scale transportation risk assessment model is constructed based on the large-scale transportation risk assessment data to quantitatively assess the risk level of transportation activities.

[0102] Step S5: Based on the large-scale transportation risk assessment data, the real-time transportation vehicle data is classified into high-risk transportation vehicles, thereby obtaining high-risk transportation vehicle data, and uploading the high-risk transportation vehicle data to the large-scale transportation activity management cloud platform to perform high-risk early warning tasks;

[0103] In this embodiment, based on the risk assessment data of large-scale transportation, the real-time transportation vehicle data is divided into high-risk transportation vehicles, and the real-time transportation vehicle data is divided into high-risk, medium-risk and low-risk levels. This can be divided according to the way of setting thresholds or rules. Once the high-risk transportation vehicles are identified, the relevant high-risk transportation vehicle data is uploaded to the large-scale transportation activity management center to perform high-risk early warning tasks, including notifying relevant departments or taking necessary safety measures.

[0104] Step S6: Perform risk mitigation strategy analysis on high-risk transport vehicle data based on real-time transport vehicle data to obtain a high-risk mitigation transport strategy, and iteratively reduce the transport strategy risk of the high-risk mitigation transport strategy through a risk assessment model to obtain a low-risk transport strategy, and upload the low-risk transport strategy to the large-scale transport activity management cloud platform to perform decision support tasks.

[0105] In this embodiment, risk mitigation strategy analysis is performed for high-risk transport vehicle data through data analysis and model optimization technology. This can include measures such as adjusting routes, limiting transport speeds, and increasing monitoring frequency. Subsequently, these strategies are evaluated and iteratively optimized using risk assessment models to reduce the risk level of transportation. Ultimately, low-risk transport strategies are determined and uploaded to the large-scale transport activity management center for decision support and task execution.

[0106] The present invention can evaluate the credit of its transport vehicles by analyzing the historical transport data of large-scale transport companies. This helps to determine which companies' vehicles can transport large-scale goods more safely, thereby improving the efficiency and accuracy of approval. By acquiring real-time transport vehicle data and sensor data, the system can monitor the status and behavior of the vehicle in a timely manner, such as vehicle speed, route, etc. Combined with license scanning data, the driver's identity and qualifications can be verified. This helps to detect abnormal situations and violations in a timely manner, and improve the safety and accuracy of the approval process. By evaluating the driver's license information, his driving record and qualifications can be determined. This helps to screen out qualified drivers, reduce safety hazards caused by driver problems, and improve the safety and reliability of approval. Combined with information such as transport vehicle credit evaluation, behavior detection data, and driver license credit evaluation, the system can perform a comprehensive large-scale transportation risk assessment. This helps to identify potential risks and safety hazards and provide a basis for subsequent decision-making. By identifying and classifying high-risk transport vehicles, the system can issue early warnings to relevant departments in a timely manner, take necessary measures to reduce potential risks and safety issues, and ensure the smooth progress of the approval process. By analyzing high-risk transport vehicle data, the system can formulate corresponding risk mitigation strategies and iterate and optimize through risk assessment models. This helps reduce risks and uncertainties in the approval process and improve the success rate and safety of approval.

[0107] Optionally, step S1 specifically includes:

[0108] Step S11: Obtain historical transportation data of large-scale transportation companies;

[0109] In this embodiment, the permission of the large-scale transport activity management cloud platform is obtained through negotiation with the large-scale transport company, and then the historical transport data of the large-scale transport company including the driving records, loading conditions, route selection and other information of the transport vehicles are obtained through the large-scale transport activity management cloud platform. This information can be extracted from the database of the large-scale transport activity management cloud platform using a database query language (such as SQL) or a data interface. During the data extraction process, attention should be paid to ensuring data privacy and security, and complying with relevant laws and regulations and company policies.

[0110] Step S12: extracting historical transportation violation records, historical transportation accident records and complaint records from the historical transportation data of the large-scale transportation company, thereby obtaining historical transportation violation data, historical transportation accident data and historical transportation complaint data;

[0111] In this embodiment, historical transportation violation records, historical transportation accident records, and historical transportation complaint records are extracted from the historical transportation data of the large-scale transportation company. Relevant information is identified and extracted from the historical transportation data of the large-scale transportation company through data mining techniques, such as text mining and pattern recognition. For example, the accident report or complaint content is analyzed using a natural language processing algorithm to extract keywords or phrases. For violation records, specific violation types and frequencies need to be identified. After data extraction, data cleaning and verification are required to ensure the accuracy and completeness of the extraction.

[0112] Step S13: evaluating the accident liability of a transport vehicle based on historical transport accident data, thereby obtaining transport vehicle accident liability evaluation data;

[0113] In this embodiment, based on historical transportation accident data, statistical analysis and machine learning models are used to evaluate the degree of accident responsibility of each transportation vehicle. For example, indicators such as the accident rate, the number of serious accidents, and the identification of responsibility for each vehicle can be calculated. Through regression analysis or classification algorithms, an accident responsibility assessment model is established to quantify the accident risk of each vehicle. The model should take into account the type of accident, the frequency of the accident, and other relevant factors.

[0114] Step S14: performing a transportation vehicle complaint analysis based on historical transportation complaint data, thereby obtaining transportation vehicle complaint data;

[0115] In this embodiment, historical transportation complaint data is used for complaint analysis. Text mining technology and sentiment analysis algorithms are used to identify key issues and trends in complaint content. For example, the main reasons for complaints, the frequency of complaints, and the results of handling can be analyzed. Through statistical methods and machine learning models, a complaint analysis model is established to evaluate the complaint risk and customer satisfaction of each vehicle.

[0116] Step S15: Counting the number of violations of transport vehicles based on historical transport violation data, thereby obtaining the number of violations of transport vehicles;

[0117] In this embodiment, historical transport violation data is statistically analyzed. Through data aggregation and summary, the number and type of violations for each transport vehicle are calculated. For example, the frequency of specific violations such as speeding violations and road rights issues can be counted. These data are helpful in assessing the compliance and risk level of vehicle drivers. After statistical analysis, data verification and quality control are performed to ensure the accuracy and reliability of the statistical results.

[0118] Step S16: Conduct a credit assessment of the transport vehicle based on the transport vehicle accident liability assessment data, the transport vehicle complaint data, and the transport vehicle violation quantity data, thereby obtaining the transport vehicle credit assessment data.

[0119] In this embodiment, the accident liability assessment data, complaint data and violation data of transport vehicles are comprehensively considered to evaluate the credit of transport vehicles. A weighted scoring method or a decision tree model can be used to comprehensively evaluate different data indicators to generate a credit assessment score for each vehicle. These assessment scores can help large-scale transport companies formulate strategies to optimize transportation plans, manage risks and improve service quality.

[0120] The present invention obtains the historical transportation data of the large-scale transportation company, and provides a data basis for subsequent evaluation and analysis. These data include transportation routes, transportation time, cargo types, vehicle information, etc., which provide basic data for evaluating the operation of the transportation company. By extracting historical transportation violation records, accident records and complaint records, the system can analyze the historical safety records of the transportation company. This helps to evaluate the safety management level and service quality of the transportation company, and provides data support for subsequent credit evaluation. By analyzing historical transportation accident data, the system can evaluate the degree of responsibility of the transportation vehicle in the accident. This helps to determine the safety record and risk level of the vehicle, and provides a reference for subsequent credit evaluation. By analyzing historical transportation complaint data, the system can understand the service quality and customer satisfaction of the transportation vehicle. This helps to evaluate the service level and credibility of the transportation company, and provides a reference for credit evaluation. Statistical analysis of historical transportation violation data can understand the violation situation and violation frequency of the vehicle. This helps to evaluate the compliance and safety of the vehicle, and provides a reference for credit evaluation. Taking into account the accident responsibility evaluation, complaint analysis and violation statistics of the transportation vehicle, the system can evaluate the credit of the transportation vehicle. This helps determine which vehicles are more reliable and safer, providing a credit assessment basis for the approval process.

[0121] Optionally, step S13 is specifically:

[0122] Step S131: extracting accident responsibility features from historical transportation accident data, thereby obtaining historical accident responsibility data;

[0123] In this embodiment, the responsibility characteristics of each accident case are extracted from the historical transportation accident data. These characteristics include accident type, accident location, vehicle type, weather conditions, road conditions, responsibility judgment, etc. Through the analysis and processing of the data, a historical accident responsibility data set with rich information is established to provide a basis for subsequent analysis and evaluation.

[0124] Step S132: classifying the historical accident responsibility data into primary and secondary responsibility data, thereby obtaining the historical accident primary responsibility data and the historical accident secondary responsibility data;

[0125] In this embodiment, the historical accident responsibility data is classified according to the accident penalty results in the historical accident responsibility data to distinguish between primary responsibility and secondary responsibility. Primary responsibility may include illegal driving, speeding, fatigue driving, etc., while secondary responsibility may involve traffic signal violations, failure to maintain vehicle distance, etc. Through this classification, the distribution and characteristics of accidents at different responsibility levels can be more clearly understood.

[0126] Step S133: performing statistics on the main responsible drivers for the historical accident main responsible data, thereby obtaining data on the main responsible drivers for high-frequency accidents and data on the main responsible drivers for low-frequency accidents;

[0127] In this embodiment, when the main responsible driver data of historical accident main responsible data is counted, it is necessary to analyze the frequency distribution and related characteristics of the main responsible driver, such as driver age, driving experience, vehicle type, etc. Technically, data mining and statistical analysis techniques, such as cluster analysis and association rule mining, can be used to identify high-frequency and low-frequency main responsible accident drivers. These analyses can be performed based on the driver's behavior pattern and the pattern of historical accidents.

[0128] Step S134: performing statistics on the secondary responsibility driver data for the historical accident secondary responsibility data, thereby obtaining the high-frequency secondary responsibility accident driver data and the low-frequency secondary responsibility accident driver data;

[0129] In this embodiment, the secondary responsibility data of historical accidents are statistically analyzed to identify drivers with higher and lower frequencies of secondary responsibility accidents. The secondary responsibility drivers are analyzed including their driving habits, accident types, frequency, etc., to distinguish between high-frequency secondary responsibility accident drivers and low-frequency secondary responsibility accident drivers. By analyzing the data of high-frequency secondary responsibility accident drivers and low-frequency secondary responsibility accident drivers, the performance of different driver groups in accidents can be more comprehensively understood.

[0130] Step S135: performing a driver data intersection operation on the low-frequency primary accident driver data and the low-frequency secondary accident driver data to obtain low-frequency accident driver data, and performing vehicle data association on the low-frequency accident driver data and the historical transportation data of the heavy-duty transportation company to obtain low-frequency accident vehicle data;

[0131] In this embodiment, the data of low-frequency main-responsibility accident drivers and low-frequency secondary-responsibility accident drivers are intersected to identify drivers who are both low-frequency main-responsibility and low-frequency secondary-responsibility. Then, the data of these drivers are associated with the historical transportation data of the large-scale transportation company to obtain the vehicle information associated with these drivers. This can help identify low-responsibility accident drivers and the vehicles they drive, and provide data support for subsequent responsibility assessment. Although low-frequency main-responsibility accident drivers bear the main responsibility in some accidents, due to the low frequency of such occurrence, their behavior is related to specific situations and does not necessarily represent daily driving behavior. Although low-frequency secondary-responsibility accident drivers participate in accidents at a low frequency, if they bear responsibility in some accidents, it is due to specific circumstances or momentary negligence, and does not represent their overall driving behavior. Therefore, it is more appropriate to classify drivers who are both low-frequency main-responsibility and low-frequency secondary-responsibility as low-responsibility drivers, because they participate in accidents at a relatively low frequency and have a small impact on accidents.

[0132] Step S136: performing a driver data intersection operation on the high-frequency primary accident driver data and the high-frequency secondary accident driver data to obtain high-frequency accident driver data, and performing vehicle data association on the high-frequency accident driver data and the historical transportation data of the large-scale transportation company to obtain high-frequency accident vehicle data;

[0133] In this embodiment, the data of drivers with high frequency of main responsibility accidents and the data of drivers with high frequency of secondary responsibility accidents are intersected to identify drivers with high frequency of main responsibility and high frequency of secondary responsibility. Then, the data of these drivers are associated with the historical transportation data of the large-scale transportation company to obtain the vehicle information associated with these drivers. This can help identify high-responsibility accident drivers and the vehicles they drive, and provide data support for subsequent responsibility assessments. Drivers with high frequency of main responsibility accidents bear the main responsibility in the accident, that is, their behavior or decision-making led to the accident. Even if the responsibility of drivers with high frequency of secondary responsibility accidents is not the main one in some accidents, their frequent involvement in accidents indicates that there are certain problems or risks in their driving behavior. Therefore, drivers who are both high frequency of main responsibility and high frequency of secondary responsibility are classified as high-responsibility drivers because their driving behavior frequently leads to accidents.

[0134] Step S137: Merge the data of vehicles with low liability in accidents and vehicles with high liability in accidents to obtain accident liability assessment data of transport vehicles.

[0135] In this embodiment, the data of low-liability accident vehicles and high-liability accident vehicles are merged to obtain a complete transportation vehicle accident responsibility assessment data set. This data set will contain the accident responsibility information of all vehicles, which can be used to assess the degree of responsibility of different vehicles in accidents and provide a basis for risk management and the formulation of safety measures.

[0136] The present invention extracts the responsibility characteristics of historical transportation accident data, and the system can identify the main responsible party and the secondary responsible party in the accident. This helps to more accurately assess the responsibility of the accident and provides a basis for subsequent data processing. Classifying the historical accident responsibility data into primary and secondary responsibilities helps to distinguish the main responsibility from the secondary responsibility and further refine the assessment of the accident responsibility. Such classification can provide more accurate data for subsequent statistics and analysis. By counting the data of the main responsible driver, the system can identify the driver who bears the main responsibility in the accident. This helps to determine the high-frequency and low-frequency main responsible accident drivers, providing an important basis for subsequent evaluation and management. Statistics on the data of the secondary responsible driver help to understand the role and frequency of the secondary responsible party in the accident. This can provide a reference for subsequent driver training and improvement. Through intersection operations and vehicle data association, identifying low-frequency main responsible accident drivers and associating them with vehicle information in the historical transportation data of large-scale transportation companies can help determine the drivers with lower responsibility in the accident and the vehicles they drive. This helps to carry out targeted training and management. Similarly, data association of high-frequency main responsible accident drivers can determine the drivers with higher responsibility in the accident and the vehicles they drive. This provides a basis for taking further management measures. By combining the data of vehicles with low responsibility for accidents and vehicles with high responsibility for accidents, the system can comprehensively evaluate the accident responsibility of transport vehicles. This provides a comprehensive perspective to help decision makers understand the safety status of vehicles and the distribution of responsibility more comprehensively.

[0137] Optionally, step S14 is specifically:

[0138] Step S141: performing frequency statistics of complaint vehicles on historical transportation complaint data, thereby obtaining high-frequency complaint vehicle data and low-frequency complaint vehicle data;

[0139] In this embodiment, the frequency of complaint vehicles is counted for historical transportation complaint data, and the Pandas library in Python can be used to process the data. First, the historical transportation complaint data is imported, and then Pandas is used to clean and preprocess the data to remove duplicates and missing values. Next, the groupby function of Pandas is used to group by vehicle number and calculate the complaint frequency of each vehicle. Finally, the vehicles are divided into two groups according to the frequency: high-frequency complaint vehicles and low-frequency complaint vehicles.

[0140] Step S142: extracting complaint transport features from historical transport data of large-scale transport companies based on the high-frequency complaint vehicle data, thereby obtaining high-frequency complaint transport feature data;

[0141] In this embodiment, for the high-frequency complaint vehicle data, the transportation features related to the complaints can be extracted from the historical transportation data of the large-scale transportation company. For example, the transportation route, transportation time, transportation distance, loading weight and other features of each vehicle are extracted. The Python Scikit-learn library can be used for feature extraction and data conversion, such as One-Hot encoding classification features, standardized numerical features, etc.

[0142] Step S143: extracting complaint transport features from historical transport data of large-scale transport companies based on low-frequency complaint vehicle data, thereby obtaining low-frequency complaint transport feature data;

[0143] In this embodiment, for low-frequency complaint vehicle data, transportation features related to complaints can be extracted from the historical transportation data of large-scale transportation companies. For example, the transportation route, transportation time, transportation distance, loading weight and other features of each vehicle are extracted. Python's Scikit-learn library can be used for feature extraction and data conversion, such as One-Hot encoding classification features, standardized numerical features, etc.

[0144] Step S144: constructing a potential complaint behavior detection model based on the high-frequency complaint transportation feature data and the low-frequency complaint transportation feature data;

[0145] In this embodiment, a potential complaint behavior detection model can be constructed based on the high-frequency complaint transportation feature data and the low-frequency complaint transportation feature data. This model can be a supervised learning model, such as logistic regression, support vector machine, random forest, etc., or an unsupervised learning model, such as clustering algorithm, anomaly detection algorithm, etc. When constructing the model, it is necessary to consider the selection of features, model parameter tuning, and model evaluation methods.

[0146] Step S145: identifying potential complaint vehicles based on historical transportation complaint data using a potential complaint behavior detection model, thereby obtaining potential complaint vehicle data;

[0147] In this embodiment, the potential complaint behavior detection model is used to identify potential complaint vehicles based on historical transportation complaint data. The previously constructed potential complaint behavior detection model is used to predict historical transportation complaint data and identify vehicles with potential complaint behaviors.

[0148] Step S146: Merge the potential complaint vehicle data, high-frequency complaint transport characteristic data, and low-frequency complaint transport characteristic data to obtain transport vehicle complaint data.

[0149] In this embodiment, the potential complaint vehicle data, high-frequency complaint transportation feature data, and low-frequency complaint transportation feature data are merged, and the merge function in the Pandas library can be used to merge data from different data sources according to vehicle numbers. The final transportation vehicle complaint data can be used for further analysis and decision making.

[0150] The present invention can identify vehicles that are frequently complained (high-frequency complaint vehicles) and vehicles that are relatively less complained (low-frequency complaint vehicles) by performing frequency statistics on historical transportation complaint data. This helps to focus attention on vehicles with problems. For high-frequency complaint vehicles and low-frequency complaint vehicles, the complaint transportation features in their historical transportation data are extracted respectively. These features include transportation routes, transportation time, loading conditions, driver information, etc. By extracting these features, it can help identify the key factors that lead to complaints. Combined with high-frequency and low-frequency complaint transportation feature data, a potential complaint behavior detection model is established. This model can identify transportation behavior patterns with potential complaint behaviors based on machine learning or statistical methods. Through such a model, potential problem vehicles can be automatically identified and the detection efficiency of complaint behaviors can be improved. Using the constructed potential complaint behavior detection model, historical transportation complaint data is analyzed and identified to identify vehicles with potential complaint behaviors. This helps to detect vehicles with problems early and take appropriate measures to manage and improve them. The potential complaint vehicle data is merged with the high-frequency and low-frequency complaint transportation feature data to obtain complete transportation vehicle complaint data. Such a dataset can be used for further analysis, such as determining the main factors of complaint behavior and formulating improvement measures.

[0151] Optionally, step S144 is specifically:

[0152] Conduct principal component analysis based on the high-frequency complaint transportation characteristic data to obtain the high-frequency complaint transportation principal component characteristic data;

[0153] In this embodiment, principal component analysis (PCA) is performed on the high-frequency complaint transportation feature data to reduce the dimension and retain the most important features. The PCA class in the Scikit-learn library in Python is used to specify the number of principal components or the proportion of explained variance to be retained. Then, the original data is projected into the selected principal component space to obtain the high-frequency complaint transportation principal component feature data.

[0154] Calculate the complaint interval for low-frequency complaint transport characteristic data to obtain complaint time interval data;

[0155] In this embodiment, the complaint interval calculation is performed on the low-frequency complaint transportation characteristic data, and the time interval between complaints of each vehicle can be calculated. First, the data is sorted according to the vehicle number and the complaint time, and then the time difference between adjacent complaint records is calculated. This can be achieved through Python's Pandas library.

[0156] Perform repeated complaint data correction on low-frequency complaint transport characteristic data according to complaint time interval data, thereby obtaining low-frequency complaint transport correction data;

[0157] In this embodiment, repeated complaint data correction is performed on low-frequency complaint transport characteristic data according to complaint time interval data, and true repeated complaint records can be screened out according to the complaint interval. For example, a time threshold can be set, and complaint intervals less than the threshold are regarded as repeated complaints. Then, the data is adjusted according to the screening results.

[0158] Conduct principal component analysis on the low-frequency complaint transportation correction data to obtain the principal component characteristic data of low-frequency complaint transportation;

[0159] In this embodiment, principal component analysis is performed on the low-frequency complaint transportation correction data, and PCA dimension reduction is also used to obtain the principal component characteristic data of low-frequency complaint transportation.

[0160] Performing transportation characteristic intersection operation on the low-frequency complaint transportation principal component characteristic data and the high-frequency complaint transportation principal component characteristic data, thereby obtaining transportation complaint characteristic data;

[0161] In this embodiment, the transportation feature intersection operation is performed on the low-frequency complaint transportation principal component feature data and the high-frequency complaint transportation principal component feature data, and the feature data of the two can be merged and the intersection can be taken. For example, the merge function or concatenate function of the Pandas library is used to merge the two sets of data, and then the duplicates are removed. In this way, the transportation complaint feature data can be obtained.

[0162] A potential complaint behavior detection model is constructed based on transportation complaint feature data and historical transportation complaint data.

[0163] In this embodiment, a potential complaint behavior detection model is constructed based on the transportation complaint feature data and historical transportation complaint data. A machine learning algorithm can be selected for modeling, such as logistic regression, random forest, etc. The Scikit-learn library in Python can be used to implement model construction and training. Then, the model is used to predict new data and identify potential complaint behaviors.

[0164] The present invention can convert a high-dimensional feature space into a lower-dimensional principal component space through principal component analysis, thereby reducing the complexity of the data and extracting the principal component feature data that best explains the variance. This helps to reduce the redundancy of the data and better understand the transportation characteristics of vehicles with high frequency complaints. Similarly, principal component analysis can be performed on the low-frequency complaint transportation correction data to identify the transportation characteristics that mainly affect low-frequency complaints and help understand the transportation mode of low-frequency complaint vehicles. By calculating the complaint time interval data of low-frequency complaint vehicles, the time interval between complaint events can be analyzed. This helps to determine the trend and pattern between complaint events and identify whether there is a fixed repeated complaint pattern. By correcting the low-frequency complaint transportation feature data according to the complaint time interval data, the complaint situation can be more accurately reflected and the same complaint event can be avoided from being repeatedly calculated. The common characteristics of transportation complaints can be obtained by performing an intersection operation on the high-frequency complaint transportation principal component feature data and the low-frequency complaint transportation principal component feature data. This helps to identify the common transportation characteristics between vehicles with different complaint frequencies, so as to better understand the root cause of the complaint behavior. A potential complaint behavior detection model is constructed based on the transportation complaint feature data and the historical transportation complaint data, and a machine learning or statistical method can be used to identify the pattern of complaint behavior. This helps to identify potential problems in advance and take appropriate measures to intervene and improve, thereby improving the overall transportation service quality and customer satisfaction.

[0165] Optionally, step S16 is specifically:

[0166] Calculate the primary and secondary liability ratios of transport vehicle accidents based on the transport vehicle accident liability assessment data, thereby obtaining the primary and secondary liability ratios of transport vehicle accidents;

[0167] In this embodiment, the ratio of the transport vehicle's primary liability accidents and the transport vehicle's secondary liability accidents contained in the transport vehicle accident liability assessment data is calculated to obtain the transport vehicle accident primary and secondary liability ratio data.

[0168] Complaints are counted based on the transport vehicle complaint data, thereby obtaining the transport vehicle complaint data;

[0169] In this embodiment, a data analysis tool is used to perform statistical analysis on these data to determine the trend and distribution of complaints in the transportation vehicle complaint data. For example, data visualization techniques (such as bar charts, line charts, etc.) can be used to show the change in the number of complaints over time and to classify and count different types of complaints.

[0170] A credit assessment model for transport vehicles is built based on the data on the primary and secondary liability ratios of transport vehicle accidents, the complaint resolution rate of transport vehicles, and the number of violations of transport vehicles;

[0171] In this embodiment, the data of the primary and secondary responsibility ratio of transport vehicle accidents, the complaint resolution rate data and the number of violations are used as features to build a comprehensive evaluation model to evaluate the credit of transport vehicles. Machine learning algorithms (such as logistic regression, decision trees, etc.) can be used to train this model, and it can be tuned and verified based on historical data. The output of the model will be a comprehensive credit score that reflects the overall credit level of the transport vehicle.

[0172] The transport vehicle credit assessment model is used to conduct a transport vehicle credit assessment on the historical transport data of the large-scale transport company, thereby obtaining the transport vehicle credit assessment data.

[0173] In this embodiment, the constructed transport vehicle credit assessment model is applied to the historical transport data of large-scale transport companies to assess the credit level of each transport vehicle. By inputting the historical transport data of large-scale transport companies into the model, the credit score of each vehicle can be obtained, thereby helping company managers to better understand the operating status and risk of their vehicles. These assessment data can serve as an important basis for decision support and help companies formulate more effective transportation management strategies and measures.

[0174] The present invention can understand the proportion of responsibility of transport vehicles in accidents by calculating the primary and secondary responsibility ratios of transport vehicle accident responsibility assessment data. This helps to determine which vehicles are more prone to accidents and which responsibilities are more important, thereby guiding the formulation of safety management strategies and risk prevention measures. Based on the transport vehicle complaint data, the historical transport complaint data can be quantitatively counted to analyze the complaints and frequencies of different vehicles. This can help identify which vehicles have problems or potential risks, and the urgency of resolving complaints. By combining the primary and secondary responsibility ratio data of transport vehicle accidents, the complaint resolution rate data and the number of violations data to construct a credit assessment model, the safety, service quality and compliance of the vehicle can be comprehensively considered. Such an assessment model can provide decision support for companies to select suitable transport vehicles and reduce transportation risks. By evaluating historical transport data through the transport vehicle credit assessment model, vehicles and companies with higher credit can be identified. This helps to establish long-term cooperative relationships, improve transportation efficiency and safety, and reduce risks and losses caused by bad vehicles.

[0175] Optionally, step S2 specifically includes:

[0176] Step S21: acquiring real-time transport vehicle data, extracting vehicle sensor data and license scanning data from the real-time transport vehicle data, thereby obtaining real-time vehicle sensor data and real-time license scanning data;

[0177] In this embodiment, real-time transport vehicle data including license scanning results, vehicle location information, vehicle speed and other data are obtained through devices deployed on the vehicle, such as document scanners, GPS, accelerometers, and vehicle speed sensors. The real-time transport vehicle data is then transmitted through the network to the large-scale transport activity management cloud platform for data mining.

[0178] Step S22: extracting the declared license data from the real-time license scan data, thereby obtaining the real-time declared license data, and converting the real-time declared license data into text, thereby obtaining the declared text data;

[0179] In this embodiment, after obtaining the real-time license scan data, the declared license data including the vehicle registration information, vehicle owner information and other images are extracted from the real-time license scan data. Subsequently, the extracted data needs to be converted into text to convert the text information in the image into processable text data. This can be achieved using technologies such as OCR technology, image processing, text recognition and natural language processing.

[0180] Step S23: extracting the declared load weight feature and the declared driving feature from the declared text data, thereby obtaining the declared load weight data and the declared driving data;

[0181] In this embodiment, the declared load characteristics and driving declaration characteristics need to be extracted from the declared text data. The declared load characteristics extraction involves regular expressions or keyword matching techniques to extract load information from the text. Driving declaration characteristics extraction requires parsing the driving route or vehicle driving plan to obtain relevant driving data, such as the estimated driving time, driving route, etc.

[0182] Step S24: performing a declared critical value overload behavior detection on the real-time vehicle sensor data based on the declared load data, thereby obtaining declared critical value overload behavior detection data;

[0183] In this embodiment, based on the declared load data, the real-time vehicle sensor data can be used to detect the declared critical value overload behavior. This involves data analysis and logical judgment. For example, a reasonable load threshold is set based on the declared load data. When the load in the vehicle sensor data exceeds the threshold, the vehicle data is classified as overload behavior data, otherwise it is classified as normal load behavior data.

[0184] Step S25: performing a reported critical value overspeeding behavior detection on the real-time vehicle sensor data based on the reported load data and the reported travel data, thereby obtaining reported critical value overspeeding behavior detection data;

[0185] In this embodiment, based on the declared load data and the declared driving data, the real-time vehicle sensor data can be used to detect the declared critical value overspeeding behavior. This requires comparing the actual driving speed of the vehicle with the declared data. If the actual speed of the vehicle exceeds the declared speed, the vehicle data is classified as overspeeding behavior data, otherwise it is classified as low-speed behavior data or normal behavior data.

[0186] Step S26: Merge the declared critical value overload behavior detection data and the declared critical value overspeed behavior detection data to obtain the declared critical value behavior detection data.

[0187] In this embodiment, the declared critical value overload behavior detection data and the declared critical value overspeed behavior detection data are merged to obtain complete declared critical value behavior detection data. This involves data integration and statistical analysis technology to ensure the integrity and accuracy of the data and provide a reliable basis for subsequent decision-making.

[0188] The present invention can timely understand the running status and behavior of the vehicle by acquiring real-time transport vehicle data. Extracting vehicle sensor data and license scanning data is helpful to monitor the real-time status and compliance of the vehicle. For example, the sensor data may include information such as the speed, acceleration, and vehicle offset of the vehicle, while the license scanning data may include license information such as the registration certificate and driving license of the vehicle. By extracting the real-time license scanning data and performing text conversion, the real-time license information of the vehicle, such as the registration certificate and driving license, can be obtained. This information is essential for monitoring the legality and compliance of the vehicle. Feature extraction of the declared text data, such as extracting load data and driving data, helps to understand the load situation and driving route of the vehicle. This information is very important for evaluating the rationality and safety of vehicle transportation behavior. By performing overloading behavior detection based on the declared load data and speeding behavior detection based on the load and driving data, the violation of the vehicle can be discovered in time. This helps to prevent traffic accidents and improve transportation safety. By merging the overloading and speeding behavior detection data, the violation of the vehicle can be comprehensively evaluated and necessary measures can be taken in time. This comprehensive analysis can better understand the behavior pattern of the vehicle and provide a reference for transportation management.

[0189] Optionally, step S25 is specifically:

[0190] Step S251: extracting the reported travel characteristics and the reported travel vehicle characteristics from the reported travel data, thereby obtaining the reported travel data and the reported travel vehicle data;

[0191] In this embodiment, information on the declared trip features and the declared vehicle features are extracted from the declared travel data through feature engineering. The declared trip features may include extraction of information such as the trip distance, trip time, starting point, and destination. The declared vehicle features include identification of information such as vehicle type, vehicle length, and vehicle height.

[0192] Step S252: predicting the vehicle speed according to the reported travel data, the reported vehicle data and the reported load data, thereby obtaining the vehicle speed data;

[0193] In this embodiment, the vehicle speed is predicted based on the declared trip data, the declared vehicle data, and the declared load data. This can be achieved through a machine learning model, such as a regression analysis of historical driving data contained in the historical transportation data of the large-scale transportation company or a time series prediction model. The training of the model will take into account the influencing factors of different vehicle types, loads, and driving sections to improve the accuracy of the prediction.

[0194] Step S253: obtaining the trip turning section data, and calculating the turning speed of large-scale transportation according to the declared driving vehicle data and the trip turning section data, thereby obtaining the vehicle transportation turning speed data;

[0195] In this embodiment, the data of the turning section of the trip is obtained, and the turning speed of the large-scale transportation is calculated using the declared driving vehicle data. This requires the calculation of the turning speed of the vehicle transportation based on mathematical modeling and physical dynamics calculation methods, taking into account the geometric characteristics of the turning section and the mechanical characteristics of the vehicle.

[0196] Step S254: performing a road section vehicle speed critical value statistical analysis on the vehicle travel speed data and the vehicle transportation turning speed data, thereby obtaining estimated road section vehicle speed critical value data;

[0197] In this embodiment, after statistical analysis of the vehicle speed data and the vehicle transportation turning speed data, the estimated road section vehicle speed critical value data is obtained. This can be used to identify and analyze the vehicle speed critical value on a specific road section and the safe speed limit on the turning road section using data mining and statistical methods. For example, using statistical methods, the vehicle speed data is analyzed including statistical indicators such as the calculation of the average speed, the standard deviation of the speed distribution, the skewness and kurtosis of the speed distribution. Based on the results of the speed distribution analysis, the normal driving speed range of each road section is predicted. This can be estimated by determining the average speed plus or minus one standard deviation.

[0198] Step S255: extracting vehicle speed sensing features from the real-time vehicle sensing data, thereby obtaining vehicle speed sensing data;

[0199] In this embodiment, vehicle speed sensing features are extracted from real-time vehicle sensor data through feature engineering to obtain vehicle speed sensor data for subsequent analysis and comparison.

[0200] Step S256: classifying the vehicle speed sensing data into travel speeds based on the estimated vehicle speed critical value data, thereby obtaining overspeed travel data and underspeed travel data;

[0201] In this embodiment, based on the estimated vehicle speed critical value data, the vehicle speed sensor data is classified into travel speed data, which may include using a data classification algorithm such as a decision tree or a support vector machine to classify the vehicle speed data into overspeed travel data and underspeed travel data, so as to identify overspeeding behavior and underspeed behavior.

[0202] Step S257: obtaining the speed limit data of the travel section, and making reasonable overspeed correction to the overspeed travel data according to the speed limit data of the travel section, thereby obtaining the overspeed travel correction data; making reasonable underspeed correction to the underspeed travel data according to the speed limit data of the travel section, thereby obtaining the underspeed travel correction data;

[0203] In this embodiment, after obtaining the speed limit data of the travel section through the database of the traffic management department, the overspeeding travel data is reasonably corrected for overspeeding to obtain the overspeeding travel correction data; the low-speed travel data is reasonably corrected for low speed to obtain the low-speed travel correction data. The actual vehicle speed is compared with the section speed limit data, and appropriate corrections are made to accurately reflect the actual driving conditions of the vehicle. For example, the overspeeding travel data and the low-speed travel data are matched with the section speed limit data to determine the section speed limit corresponding to each travel. The overspeeding travel data is corrected, and the overspeeding travel data that exceeds the predetermined speed due to the section speed limit is adjusted and marked as normal travel data. The low-speed travel data is corrected, and the low-speed travel data that is lower than the predetermined speed due to the section speed limit is adjusted and marked as normal travel data. The corrected overspeeding travel data and low-speed travel data are merged to form the final declared driving data for subsequent analysis and application.

[0204] Step S258: Merge the overspeed stroke correction data and the underspeed stroke correction data to report critical value overspeed behavior detection data.

[0205] In this embodiment, the overspeed travel correction data and the underspeed travel correction data are merged to obtain the reporting threshold overspeed behavior detection data. This involves data integration and statistical analysis to generate the final overspeed behavior detection results for subsequent behavior evaluation and monitoring.

[0206] The present invention can better understand the driving mode and characteristics of the vehicle by extracting the travel characteristics and vehicle characteristics of the declared driving data. This information can be used for speed prediction and turning speed calculation in subsequent steps. The travel speed of the vehicle is predicted by combining the travel data, the driving vehicle data and the load data. This helps to predict the driving speed of the vehicle under different conditions, so as to detect overspeeding or underspeed behavior in advance. The travel turning section data is obtained, and the turning speed of large-scale transportation is calculated according to the vehicle characteristics. This can help to evaluate the safety and stability of the vehicle when turning, so as to reduce the risk of accidents. The vehicle travel speed data and the turning speed data are analyzed to determine the expected road section speed critical value. This helps to identify the existing overspeeding or underspeed behavior and formulate corresponding management and control strategies. The speed sensing characteristics are extracted from the real-time vehicle sensor data to obtain the current speed information of the vehicle. According to the expected road section speed critical value, the real-time speed data of the vehicle is classified to identify overspeeding and underspeed travel. This helps to timely discover and record non-compliant behavior. The travel section speed limit data is obtained, and the overspeeding and underspeed travel data are reasonably corrected according to the speed limit data. This ensures the accuracy and reliability of the trip data. The corrected overspeed and underspeed trip data are combined to form the reporting threshold overspeed behavior detection data. This is a key step in the comprehensive evaluation and monitoring of vehicle driving behavior, which helps managers take necessary measures in a timely manner.

[0207] Optionally, step S3 specifically includes:

[0208] Step S31: extracting driver's license data from the license scan data to obtain driver's license data, and converting the driver's license data into text to obtain driver's license text data;

[0209] In this embodiment, optical character recognition (OCR) technology is used to process the scanned certificate image and extract the text content in the image. For example, the scanned certificate image is preprocessed, including noise removal, image enhancement, edge detection and other operations to improve the accuracy of subsequent character recognition. Image processing technology is used to detect the text area in the certificate image to determine the location and size of the text. Character recognition is performed on the text area, and the text content in the image is converted into recognizable text data using OCR technology. The extracted text data is formatted for subsequent data analysis and processing. This may involve the extraction and organization of information such as name, certificate number, and expiration date.

[0210] Step S32: extracting the driver's license feature and the vehicle license feature from the driver's license text data, thereby obtaining the driver's license text data and the vehicle license text data;

[0211] In this embodiment, key information such as name, certificate number, driver's license type, validity period, etc. is extracted from the driver's license text data. This can be achieved through name segmentation, regular expression matching and other technologies. Similarly, key information such as owner's name, vehicle registration number, vehicle brand, engine number, etc. is extracted from the vehicle driving license text data. This can be achieved using regular expressions.

[0212] Step S33: Calculate the name similarity of the driver's license text data and the vehicle license text data, so as to obtain the driver's license name similarity data;

[0213] In this embodiment, a string similarity algorithm (such as Levenshtein distance, Jaccard similarity, etc.) is used to compare the similarity between the driver's name in the driver's license and the owner's name in the vehicle registration certificate. A similarity threshold is set to determine whether two names are similar. According to actual conditions and needs, the threshold can be set according to business logic.

[0214] Step S34: Calculating the similarity of the declared driver's name on the declared text data, the driver's license text data, and the vehicle driving license text data, thereby obtaining the declared driver's name similarity data;

[0215] In this embodiment, similarly, a string similarity algorithm is used to compare the similarity between the declared driver's name, the vehicle name in the vehicle license, and the driver's name in the driver's license. A similarity threshold is set to determine whether the declared name is similar to the driver's name in the driver's license.

[0216] Step S35: Perform credit evaluation based on the name similarity data of the driver's license and the similarity data of the declared driver's name, so as to obtain the driver's license credit evaluation data.

[0217] In this embodiment, a credit assessment model is established based on the similarity data of the names of the driver's license and the similarity data of the driver's name, using machine learning algorithms, statistical models and other methods. The credit assessment model is then used to assess the driver's license credit, and the driver's credit assessment results including credit rating, credit score and other information are output for subsequent business decision-making and application use.

[0218] The present invention converts the driver's license data into processable text data by extracting and converting the license scan data, which helps to make subsequent data processing and analysis more efficient and accurate. Feature extraction is performed on the driver's license text data and the vehicle driving license text data to obtain key information. These features include name, certificate number, validity period, etc., which are helpful for data matching and analysis in subsequent steps. By calculating the similarity of the names of the driver's license text data and the vehicle driving license text data, the consistency and accuracy of the license information can be evaluated. This helps to identify the existing license information errors or fraudulent use. The declared driver's name similarity calculation is performed on the declared text data, the driver's license text data and the vehicle driving license text data to further verify the consistency of the declared name information with the license information. This can reduce the situation where the driver's identity information does not match. Credit evaluation is performed based on the driver's license name similarity data and the declared driver name similarity data. This step will comprehensively consider the consistency of the license information and the declared information to evaluate the driver's credit level and compliance. This helps to screen out drivers with higher credit and improve the safety and reliability of vehicle transportation.

[0219] Optionally, step S6 specifically includes:

[0220] Step S61: extracting vehicle position features from real-time transport vehicle data, thereby obtaining real-time transport vehicle position data;

[0221] In this embodiment, location information including longitude and latitude data is extracted from real-time transport vehicle data. Feature engineering is performed on the extracted location information to convert longitude and latitude into specific locations, calculate vehicle travel speed, etc.

[0222] Step S62: analyzing the vehicle driving condition based on the vehicle sensor data, thereby obtaining the vehicle driving condition data;

[0223] In this embodiment, data mining or machine learning technology is used to analyze the road surface smoothness, turning conditions, and traffic flow based on vehicle sensor data to obtain road condition information for the vehicle.

[0224] Step S63: screening the vehicle stop location based on the vehicle driving road condition data and the real-time transport vehicle location data, thereby obtaining the vehicle stop location data;

[0225] In this embodiment, the vehicle driving road condition data and the real-time vehicle location data are integrated to facilitate subsequent analysis. A clustering algorithm, such as a k-means algorithm, is used to analyze the vehicle driving road condition data to identify the location where the vehicle stops. Based on the distance relationship between the stop point and the vehicle, a closer stop location is identified.

[0226] Step S64: performing load optimization analysis on real-time transport vehicle data based on the declared critical value behavior detection data, thereby obtaining vehicle load reduction data;

[0227] In this embodiment, based on the declared critical value behavior detection data, the real-time transport vehicle data is subjected to load optimization analysis, the vehicle loading condition is evaluated, the load deviation is analyzed, etc. The vehicle load reduction, that is, the load that the vehicle needs to reduce, is calculated.

[0228] Step S65: analyzing the vehicle driving strategy according to the vehicle stop location data and the vehicle load reduction data, thereby obtaining a high-risk mitigation transportation strategy;

[0229] In this embodiment, the vehicle parking location data is associated with the vehicle load reduction data to determine the load reduction corresponding to the parking location. According to the parking location and load reduction data, a vehicle driving strategy is formulated, such as selecting the optimal parking point, adjusting the load, etc.

[0230] Step S66: Iteratively reduce the risk of the high-risk mitigation transportation strategy through the risk assessment model to obtain a low-risk transportation strategy, and upload the low-risk transportation strategy to the large-scale transportation activity management cloud platform to perform decision support tasks.

[0231] In this embodiment, the high-risk mitigation transportation strategy is iteratively optimized through the risk assessment model, and the transportation parameters of the transportation strategy are adjusted to minimize the risk. The optimized low-risk transportation strategy is uploaded to the large-scale transportation activity management cloud platform for the execution and monitoring of decision support tasks, and the large-scale transportation target declaration task is optimized according to the low-risk transportation strategy to improve the declaration pass rate and transportation safety rate.

[0232] The present invention obtains the real-time location data of the vehicle by extracting the location features in the real-time transport vehicle data. This makes it feasible to monitor the location of the vehicle and provides a data basis for subsequent route planning and risk analysis. Based on the vehicle sensor data, the vehicle driving road condition is analyzed, which helps to understand the driving conditions of the vehicle under different road conditions, including road flatness, congestion, etc., and provides a reference for the subsequent vehicle driving strategy. Combined with the vehicle driving road condition data and the real-time vehicle location data, the vehicle parking location is screened out. This helps to identify suitable parking locations, optimize transportation routes, and improve transportation efficiency. Based on the declared critical value behavior detection data, the real-time transport vehicle data is optimized for load capacity analysis. This helps to determine the actual load capacity of the vehicle and improve transportation efficiency and safety. Combined with the vehicle parking location data and the vehicle load reduction data, the vehicle driving strategy analysis is carried out to find strategies to reduce high-risk transportation, such as selecting a safe parking point or adjusting the load capacity. The high-risk mitigation transportation strategy is evaluated and optimized through the risk assessment model. This helps to reduce potential transportation risks and provide low-risk transportation strategies. After uploading to the large-scale transportation activity management cloud platform, better decision support and coordination management can be achieved.

[0233] Therefore, the embodiments should be regarded as illustrative and non-restrictive from all points, and the scope of the present invention is limited by the appended claims rather than the above description, and it is therefore intended that all changes falling within the meaning and range of equivalent elements of the application documents are included in the present invention.

[0234] The above description is only a specific embodiment of the present invention, so that those skilled in the art can understand or implement the present invention. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to the embodiments shown herein, but should conform to the widest scope consistent with the principles and novel features invented herein.

Claims

1. A decision support method for approval of large-scale transportation activities, characterized in that: The following steps are involved: Step S1: Obtain historical transportation data of large-scale transportation companies, and conduct transportation vehicle credit assessment on the historical transportation data of large-scale transportation companies, so as to obtain transportation vehicle credit assessment data; Step S1 is specifically as follows: Step S11: Obtain historical transportation data of large-scale transportation companies; Step S12: extracting historical transportation violation records, historical transportation accident records and complaint records from the historical transportation data of the large-scale transportation company, thereby obtaining historical transportation violation data, historical transportation accident data and historical transportation complaint data; Step S13: Performing a transport vehicle accident liability assessment based on historical transport accident data, thereby obtaining transport vehicle accident liability assessment data; Step S13 is specifically as follows: Step S131: extracting accident responsibility features from historical transportation accident data, thereby obtaining historical accident responsibility data; Step S132: classifying the historical accident responsibility data into primary and secondary responsibility data, thereby obtaining the historical accident primary responsibility data and the historical accident secondary responsibility data; Step S133: performing statistics on the main responsible drivers for the historical accident main responsible data, thereby obtaining data on the main responsible drivers for high-frequency accidents and data on the main responsible drivers for low-frequency accidents; Step S134: performing statistics on the secondary responsibility driver data for the historical accident secondary responsibility data, thereby obtaining the high-frequency secondary responsibility accident driver data and the low-frequency secondary responsibility accident driver data; Step S135: performing a driver data intersection operation on the low-frequency primary accident driver data and the low-frequency secondary accident driver data to obtain low-frequency accident driver data, and performing vehicle data association on the low-frequency accident driver data and the historical transportation data of the heavy-duty transportation company to obtain low-frequency accident vehicle data; Step S136: performing a driver data intersection operation on the high-frequency primary accident driver data and the high-frequency secondary accident driver data to obtain high-frequency accident driver data, and performing vehicle data association on the high-frequency accident driver data and the historical transportation data of the large-scale transportation company to obtain high-frequency accident vehicle data; Step S137: merging the data of vehicles with low liability in accidents and vehicles with high liability in accidents, thereby obtaining accident liability assessment data of transport vehicles; Step S14: Performing a transportation vehicle complaint analysis based on historical transportation complaint data, thereby obtaining transportation vehicle complaint data; Step S14 is specifically as follows: Step S141: performing frequency statistics of complaint vehicles on historical transportation complaint data, thereby obtaining high-frequency complaint vehicle data and low-frequency complaint vehicle data; Step S142: extracting complaint transport features from historical transport data of large-scale transport companies based on the high-frequency complaint vehicle data, thereby obtaining high-frequency complaint transport feature data; Step S143: extracting complaint transport features from historical transport data of large-scale transport companies based on low-frequency complaint vehicle data, thereby obtaining low-frequency complaint transport feature data; Step S144: construct a potential complaint behavior detection model based on the high-frequency complaint transportation feature data and the low-frequency complaint transportation feature data; Step S144 is specifically as follows: Conduct principal component analysis based on the high-frequency complaint transportation characteristic data to obtain the high-frequency complaint transportation principal component characteristic data; Calculate the complaint interval for low-frequency complaint transport characteristic data to obtain complaint time interval data; Perform repeated complaint data correction on low-frequency complaint transport characteristic data according to complaint time interval data, thereby obtaining low-frequency complaint transport correction data; Conduct principal component analysis on the low-frequency complaint transportation correction data to obtain the principal component characteristic data of low-frequency complaint transportation; Performing transportation characteristic intersection operation on the low-frequency complaint transportation principal component characteristic data and the high-frequency complaint transportation principal component characteristic data, thereby obtaining transportation complaint characteristic data; Construct a potential complaint behavior detection model based on transportation complaint feature data and historical transportation complaint data; Step S145: identifying potential complaint vehicles based on historical transportation complaint data using a potential complaint behavior detection model, thereby obtaining potential complaint vehicle data; Step S146: merging the potential complaint vehicle data, the high-frequency complaint transportation characteristic data, and the low-frequency complaint transportation characteristic data to obtain the transportation vehicle complaint data; Step S15: Counting the number of violations of transport vehicles based on historical transport violation data, thereby obtaining the number of violations of transport vehicles; Step S16: performing a credit assessment of the transport vehicle based on the transport vehicle accident liability assessment data, the transport vehicle complaint data, and the transport vehicle violation quantity data, thereby obtaining the transport vehicle credit assessment data; Step S2: Acquire real-time transport vehicle data, extract vehicle sensor data and license scan data from the real-time transport vehicle data, thereby obtaining real-time vehicle sensor data and real-time license scan data; perform declaration threshold vehicle behavior detection based on the real-time vehicle sensor data and real-time license scan data, thereby obtaining declaration threshold behavior detection data. Step S2 is specifically as follows: Step S21: acquiring real-time transport vehicle data, extracting vehicle sensor data and license scanning data from the real-time transport vehicle data, thereby obtaining real-time vehicle sensor data and real-time license scanning data; Step S22: extracting the declared license data from the real-time license scan data, thereby obtaining the real-time declared license data, and converting the real-time declared license data into text, thereby obtaining the declared text data; Step S23: extracting the declared load weight feature and the declared driving feature from the declared text data, thereby obtaining the declared load weight data and the declared driving data; Step S24: performing a declared critical value overload behavior detection on the real-time vehicle sensor data based on the declared load data, thereby obtaining declared critical value overload behavior detection data; Step S25: Based on the declared load data and the declared driving data, the real-time vehicle sensor data is subjected to a declared critical value overspeeding behavior detection, thereby obtaining declared critical value overspeeding behavior detection data. Step S25 is specifically as follows: Step S251: extracting the reported travel characteristics and the reported travel vehicle characteristics from the reported travel data, thereby obtaining the reported travel data and the reported travel vehicle data; Step S252: predicting the vehicle speed according to the reported travel data, the reported vehicle data and the reported load data, thereby obtaining the vehicle speed data; Step S253: obtaining the trip turning section data, and calculating the turning speed of large-scale transportation according to the declared driving vehicle data and the trip turning section data, thereby obtaining the vehicle transportation turning speed data; Step S254: performing a road section vehicle speed critical value statistical analysis on the vehicle travel speed data and the vehicle transportation turning speed data, thereby obtaining estimated road section vehicle speed critical value data; Step S255: extracting vehicle speed sensing features from the real-time vehicle sensing data, thereby obtaining vehicle speed sensing data; Step S256: classifying the vehicle speed sensing data into travel speeds based on the estimated vehicle speed critical value data, thereby obtaining overspeed travel data and underspeed travel data; Step S257: obtaining the speed limit data of the travel section, and making reasonable overspeed correction to the overspeed travel data according to the speed limit data of the travel section, thereby obtaining the overspeed travel correction data; making reasonable underspeed correction to the underspeed travel data according to the speed limit data of the travel section, thereby obtaining the underspeed travel correction data; Step S258: merging the overspeed stroke correction data and the underspeed stroke correction data, thereby reporting the critical value overspeed behavior detection data; Step S26: merging the declared critical value overload behavior detection data and the declared critical value overspeed behavior detection data to obtain the declared critical value behavior detection data; Step S3: performing a driver's license credit assessment based on the license scan data, thereby obtaining driver's license credit assessment data; Step S4: Conducting a large-scale transportation risk assessment on the transport vehicle credit assessment data, the declared critical value behavior detection data, and the driver's license credit assessment data, thereby obtaining large-scale transportation risk assessment data, and constructing a risk assessment model based on the large-scale transportation risk assessment data; Step S5: Based on the large-scale transportation risk assessment data, the real-time transportation vehicle data is classified into high-risk transportation vehicles, thereby obtaining high-risk transportation vehicle data, and uploading the high-risk transportation vehicle data to the large-scale transportation activity management cloud platform to perform high-risk early warning tasks; Step S6: Perform risk mitigation strategy analysis on high-risk transport vehicle data based on real-time transport vehicle data to obtain a high-risk mitigation transport strategy, and iteratively reduce the transport strategy risk of the high-risk mitigation transport strategy through a risk assessment model to obtain a low-risk transport strategy, and upload the low-risk transport strategy to the large-scale transport activity management cloud platform to perform decision support tasks.

2. The decision support method for large-scale transportation activity licensing according to claim 1 is characterized in that: Step S16 is specifically as follows: Calculate the primary and secondary liability ratios of transport vehicle accidents based on the transport vehicle accident liability assessment data, thereby obtaining the primary and secondary liability ratios of transport vehicle accidents; Complaints are counted based on the transport vehicle complaint data, thereby obtaining the transport vehicle complaint data; A credit assessment model for transport vehicles is built based on the data on the primary and secondary liability ratios of transport vehicle accidents, the complaint resolution rate of transport vehicles, and the number of violations of transport vehicles; The transport vehicle credit assessment model is used to conduct a transport vehicle credit assessment on the historical transport data of the large-scale transport company, thereby obtaining the transport vehicle credit assessment data.

3. The decision support method for large-scale transportation activity licensing according to claim 1 is characterized in that: Step S3 is specifically as follows: Step S31: extracting driver's license data from the license scan data to obtain driver's license data, and converting the driver's license data into text to obtain driver's license text data; Step S32: extracting the driver's license feature and the vehicle license feature from the driver's license text data, thereby obtaining the driver's license text data and the vehicle license text data; Step S33: Calculate the name similarity of the driver's license text data and the vehicle license text data, so as to obtain the driver's license name similarity data; Step S34: Calculating the similarity of the declared driver's name on the declared text data, the driver's license text data, and the vehicle driving license text data, thereby obtaining the declared driver's name similarity data; Step S35: Perform credit evaluation based on the name similarity data of the driver's license and the similarity data of the declared driver's name, so as to obtain the driver's license credit evaluation data.

4. The decision support method for large-scale transportation activity licensing according to claim 1 is characterized in that: Step S6 is specifically as follows: Step S61: extracting vehicle position features from real-time transport vehicle data, thereby obtaining real-time transport vehicle position data; Step S62: analyzing the vehicle driving condition based on the vehicle sensor data, thereby obtaining the vehicle driving condition data; Step S63: screening the vehicle stop location based on the vehicle driving road condition data and the real-time transport vehicle location data, thereby obtaining the vehicle stop location data; Step S64: performing load optimization analysis on real-time transport vehicle data based on the declared critical value behavior detection data, thereby obtaining vehicle load reduction data; Step S65: analyzing the vehicle driving strategy according to the vehicle stop location data and the vehicle load reduction data, thereby obtaining a high-risk mitigation transportation strategy; Step S66: Iteratively reduce the risk of the high-risk mitigation transportation strategy through the risk assessment model to obtain a low-risk transportation strategy, and upload the low-risk transportation strategy to the large-scale transportation activity management cloud platform to perform decision support tasks.

Images