A vehicle delay prediction method and system based on data analysis

Abstract

The application provides a vehicle delay prediction method and system based on data analysis, and relates to the technical field of data analysis. The method comprises the following steps: first, obtaining a first driving route, dividing the route into Q nodes, extracting each node, obtaining traffic congestion information between each node through a traffic management platform, obtaining a predicted travel time at this time, then obtaining weather data, performing travel analysis on the weather data, obtaining a predicted travel time at this time, and finally obtaining a final vehicle delay time through the two predicted travel times. The application mainly solves the technical problem that the time and space correlation between multiple trains and multiple routes cannot be comprehensively considered, and accurate prediction cannot be performed when there is a large passenger flow, vehicle failure or extreme weather. By dividing a road section into multiple nodes, predicting the road section between each node, and then analyzing the delay time in combination with weather data, a more accurate vehicle delay prediction result is obtained.

Description

Technical Field

[0001] This invention relates to the technical field of data analysis, specifically to a method and system for predicting vehicle delays based on data analysis. Background Technology

[0002] Currently, vehicle delay prediction technology has begun to be applied in fields such as traffic engineering and intelligent transportation. For example, in intelligent transportation systems, vehicles can obtain real-time traffic information through high-precision maps, sensors, and other technologies, and then use vehicle delay prediction models to derive the optimal driving route to reduce vehicle delays during the journey. However, some challenges and problems still exist in practical applications.

[0003] Existing technology predicts vehicle delays by analyzing historical traffic flow using statistical methods such as historical data and time series analysis.

[0004] Current technologies are primarily based on historical or real-time data, which is costly to acquire and process, and the accuracy of the data may affect the prediction results. There are also technical problems with the inability to accurately predict events such as large passenger flows, vehicle malfunctions, and extreme weather. Summary of the Invention

[0005] This application primarily addresses the technical problem of inaccurate predictions when the temporal and spatial correlations between multiple trains and lines cannot be fully considered, especially during periods of high passenger volume, vehicle malfunctions, and extreme weather. By dividing the road segment into multiple nodes, predicting the road segments between each node separately, and then combining this prediction with weather data analysis, more accurate vehicle delay prediction results are obtained.

[0006] In view of the above problems, embodiments of this application provide a vehicle delay prevention method and system based on data analysis. In a first aspect, embodiments of this application provide a vehicle delay prediction method based on data analysis. The method includes: obtaining a first travel route of a first vehicle, the first travel route comprising Q nodes and Q road segments; extracting the current node, and M-1 nodes and M road segments preceding the current node from the Q nodes; obtaining traffic congestion information of the M road segments through a traffic management platform; performing travel time analysis based on the traffic congestion information to obtain M predicted travel times; sorting the M predicted travel times; obtaining the predicted delay time of the current node based on the sorting results; obtaining current weather data through a meteorological platform, the weather data including at least wind speed, rainfall, visibility, and weather warnings; performing traffic impact analysis based on the weather data to obtain a traffic impact coefficient; and obtaining the vehicle delay time of the current node based on the traffic impact coefficient and the predicted delay time.

[0007] Secondly, embodiments of this application provide a vehicle delay prediction system based on data analysis. The system includes: a first driving route acquisition module, used to acquire a first driving route of a first vehicle, the first driving route comprising Q nodes and Q road segments; a node extraction module, used to extract the current node, as well as M-1 nodes and M road segments preceding the current node from the Q nodes; and a predicted travel time acquisition module, used to acquire traffic congestion information of the M road segments through a traffic management platform, perform travel time analysis based on the traffic congestion information, and acquire M predicted travel times to predict the delay. The system includes a delay time acquisition module, which sorts the M predicted travel times and obtains the predicted delay time for the current node based on the sorting results; a weather data acquisition module, which acquires current weather data through a meteorological platform, including at least wind speed, rainfall, visibility, and weather warnings; a traffic impact coefficient acquisition module, which performs traffic impact analysis based on the weather data and obtains a traffic impact coefficient; and a final delay time acquisition module, which obtains the vehicle delay time for the current node based on the traffic impact coefficient and the predicted delay time.

[0008] One or more technical solutions provided in the embodiments of this application have at least the following technical effects or advantages:

[0009] This application provides a vehicle delay prediction method and system based on data analysis, relating to the field of data analysis technology. The method includes: firstly, obtaining a first travel route, dividing the route into Q nodes, extracting each node, obtaining traffic congestion information between each node through a traffic management platform, obtaining the predicted travel time at this time, then obtaining weather data, performing traffic analysis on the weather data, obtaining the predicted travel time at this time, and obtaining the final vehicle delay time through the two predicted travel times.

[0010] This application primarily addresses the technical problem of inaccurate predictions when the temporal and spatial correlations between multiple trains and lines cannot be fully considered, especially during periods of high passenger volume, vehicle malfunctions, and extreme weather. By dividing the road segment into multiple nodes, predicting the road segments between each node separately, and then combining this prediction with weather data analysis, more accurate vehicle delay prediction results are obtained.

[0011] The above description is merely an overview of the technical solution of this application. In order to better understand the technical means of this application and to implement it in accordance with the contents of the specification, and to make the above and other objects, features and advantages of this application more obvious and understandable, specific embodiments of this application are given below. Attached Figure Description

[0012] To more clearly illustrate the technical solutions in this disclosure or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are merely exemplary. For those skilled in the art, other drawings can be obtained based on the provided drawings without creative effort.

[0013] Figure 1 This application provides a schematic flowchart of a vehicle delay prediction method based on data analysis.

[0014] Figure 2 This application provides a schematic flowchart of a data analysis-based vehicle delay prediction method that corrects vehicle delay time based on vehicle repair time.

[0015] Figure 3 This application provides a schematic flowchart of a data analysis-based vehicle delay prediction method for compensating for vehicle delay time.

[0016] Figure 4 This application provides a schematic diagram of the structure of a vehicle delay prediction system based on data analysis.

[0017] Explanation of reference numerals in the attached diagram: First travel route acquisition module 10, node extraction module 20, predicted travel time acquisition module 30, predicted delay time acquisition module 40, weather data acquisition module 50, traffic impact coefficient acquisition module 60, and final delay time acquisition module 70. Detailed Implementation

[0018] This application primarily addresses the technical problem of inaccurate predictions when the temporal and spatial correlations between multiple trains and lines cannot be fully considered, especially during periods of high passenger volume, vehicle malfunctions, and extreme weather. By dividing the road segment into multiple nodes, predicting the road segments between each node separately, and then combining this prediction with weather data analysis, more accurate vehicle delay prediction results are obtained.

[0019] The technical solution in this application is to solve the above problems, and the overall solution approach is as follows:

[0020] The first driving route of the first vehicle is obtained, which includes Q nodes and Q road segments. The current node, as well as the M-1 nodes and M road segments preceding the current node, are extracted from the Q nodes. Traffic congestion information of the M road segments is obtained through a traffic management platform. Based on the traffic congestion information, travel time is analyzed to obtain M predicted travel times. The original travel time of the M road segments is obtained, and the difference between the M predicted travel times and the original travel times is calculated. Differences less than a preset time threshold are discarded, and the remaining differences are accumulated to obtain the predicted delay time of the current node. The M predicted travel times are sorted, and the predicted delay time of the current node is obtained based on the sorting results.

[0021] The system acquires current weather data through a meteorological platform, including at least wind speed, rainfall, visibility, and weather warnings. It also acquires historical weather data and historical traffic data, performs correlation analysis on the historical weather data and historical traffic data, obtains the correlation analysis results, calculates the impact of wind speed, rainfall, visibility, and weather warnings on traffic based on the correlation analysis results, weights the impact, and calculates the traffic impact coefficient based on the weighted results for the current wind speed, rainfall, visibility, and weather warnings.

[0022] Traffic impact analysis is performed based on the weather data to obtain a traffic impact coefficient. Based on the traffic impact coefficient and the predicted delay time, the vehicle delay time at the current node is obtained. When a vehicle malfunctions, feedback information from a first electronic device is obtained, including the vehicle malfunction type and malfunction level. Repair time analysis is performed based on the vehicle malfunction type and malfunction level to obtain the vehicle repair time. The vehicle delay time is then corrected based on the vehicle repair time. Real-time image acquisition is performed on the M-1 nodes using the image acquisition device to obtain the image acquisition results.

[0023] The image acquisition results are analyzed for passenger numbers, and target detection is performed to obtain real-time personnel count data for each node, including the number of people entering and leaving the station. A preset time window is established, and an average number of people is calculated based on the preset time window and the number of people entering and leaving the station to obtain the number of passengers at each node. The passenger boarding and alighting time is calculated based on the number of passengers, and the vehicle delay time is compensated based on the calculated passenger boarding and alighting time.

[0024] To better understand the above technical solution, the following will provide a detailed description of the solution in conjunction with the accompanying drawings and specific implementation methods:

[0025] Example 1

[0026] like Figure 1 This paper presents a data analysis-based method for predicting vehicle delays. The method is applied to a vehicle delay prediction system, which is communicatively connected to a traffic management platform and a meteorological platform. The method includes:

[0027] A vehicle delay prediction system is a system that uses various technologies and methods to predict potential vehicle delays during travel. It helps traffic management departments and drivers better understand traffic conditions, predict vehicle delays, and thus develop more reasonable traffic management and control strategies to reduce traffic congestion and delays. Simultaneously, the system can also help drivers choose more efficient routes, reducing travel time and delays. Traffic management platforms provide traffic information and management services by using technological means to detect and predict traffic congestion. By utilizing big data technology and artificial intelligence, it provides city management departments and travelers with more comprehensive and accurate traffic information and governance solutions, helping to improve urban traffic management efficiency and alleviate urban traffic congestion. Meteorological platforms can provide real-time weather forecasts for the city, including wind speed, air quality, temperature, and rainfall. The vehicle delay prediction system connects with both traffic management and meteorological platforms to obtain road congestion and weather conditions, using this data as a basis for delay prediction.

[0028] Obtain the first driving route of the first vehicle, which includes Q nodes and Q road segments;

[0029] Specifically, the first route of the first vehicle is obtained. The first vehicle is a vehicle operating in the urban public transportation system, such as a bus. Delay judgment needs to be performed to ensure that the arrival interval of each vehicle at a station is uniform. For buses, the first route is the initial station of the bus. A route contains multiple stations, divided into Q nodes and Q segments, where Q is a positive integer. Each segment is analyzed to facilitate vehicle prediction and management.

[0030] Extract the current node, the M-1 nodes preceding the current node, and the M road segments from the Q nodes;

[0031] Specifically, a data structure needs to be defined to store node and road segment information. An adjacency list can be used to represent the graph, where each node has a list of connected road segments. Extracting the current node from the Q partitioned nodes can be done using a queue to implement a breadth-first search (BFS) algorithm to traverse the graph. Each element in the queue represents a node and its M-1 preceding nodes and M road segments. For the current node, necessary operations can be performed, such as calculating delay time or performing route planning. All adjacent nodes of the current node (and their M-1 preceding nodes and M road segments) are added to the queue. Before adding them, it's crucial to check for duplicate elements in the queue; a hash table can be used to store this information to avoid adding the same nodes and road segments repeatedly. The current node represents the node the vehicle has arrived at in real time, along with the M-1 preceding nodes and M road segments, where M is a positive integer to facilitate further analysis and processing of the current node and the nodes preceding it.

[0032] Traffic congestion information for the M road segments is obtained through a traffic management platform. Based on the traffic congestion information, travel time is analyzed to obtain M predicted travel times.

[0033] Specifically, the traffic management platform obtains traffic congestion information for M road segments. Each segment has different traffic volumes, and the number of passengers getting on and off at each stop varies. The traffic congestion platform refers to a platform that monitors and predicts traffic congestion through technological means, providing traffic information and management services. By utilizing big data technology and artificial intelligence, it provides urban management departments and travelers with more comprehensive and accurate traffic information and governance solutions, helping to improve urban traffic management efficiency and alleviate urban traffic congestion. Communication time analysis is performed based on traffic congestion information, using GIS geographic information systems, GPS positioning systems, simulation technology, etc. Analysis is conducted based on factors such as average formation time, traffic delays, traffic volume, and vehicle speed to obtain M predicted travel times. These predicted travel times are then used to predict travel times based on congestion conditions. Analysis of each road segment facilitates the management of other vehicles and prevents excessive congestion on any particular route.

[0034] The M predicted travel times are sorted, and the predicted delay time of the current node is obtained based on the sorting results;

[0035] Specifically, M predicted travel times are sorted and ranked from smallest to largest. This clearly shows which road segments have short travel times and which have long travel times. If the predicted travel time range is 1-10 minutes, the segments are sorted by length, and a suitable threshold T is selected. Then, all predicted delay times less than or equal to the threshold T are retained, while all predicted delay times greater than the threshold T are discarded. T ≥ 10 minutes. The retained delay times are used as the sorting results. The current node is retrieved from the sorting results to obtain its predicted travel time. Historical travel times are obtained by querying traffic data. The delay time is calculated by subtracting the historical travel time from the predicted travel time. For example, if the historical travel time is 3 minutes and the predicted travel time is 6 minutes, the predicted delay time is 3 minutes. This provides the predicted delay time for the current node based on congestion conditions, which can be used for further management and optimization.

[0036] Current weather data is obtained through a meteorological platform, including at least wind speed, rainfall, visibility, and weather warnings.

[0037] Specifically, current weather data can be obtained through meteorological platforms. Mobile weather applications such as WeatherTong, China Weather, and Fengyun Weather can be used. These applications also provide real-time weather information and can automatically obtain weather information for the user's location. Weather data includes wind speed, rainfall, visibility, weather warnings, and other information to obtain accurate weather conditions and the impact of weather conditions on road conditions.

[0038] Based on the aforementioned weather data, a traffic impact analysis is performed to obtain the traffic impact coefficient;

[0039] Specifically, different influencing factors are identified based on specific weather data and traffic conditions. For example, high temperatures may lead to reduced vehicle speeds and increased congestion, while heavy rain may result in increased traffic accidents and road closures. For each influencing factor, its impact on traffic flow is analyzed. The degree of influence of each factor can be determined using historical data, simulation experiments, and other methods. For instance, in high temperatures, vehicle speeds may decrease by 20%, and congestion may increase by 30%. Finally, based on the degree of influence of each factor, a total traffic impact coefficient is calculated. This coefficient can be a weighted average or a comprehensive evaluation to represent the degree of weather's impact on traffic. The traffic impact coefficient can be used to assess the effectiveness of traffic management strategies or predict traffic conditions under specific weather conditions. Travel time can also be predicted using the impact coefficient.

[0040] Based on the traffic impact coefficient and the predicted delay time, the vehicle delay time at the current node is obtained.

[0041] Specifically, firstly, the predicted delay time is multiplied by the traffic impact coefficient to obtain the total delay time. For example, if the predicted delay time is 10 minutes and the traffic impact coefficient is 0.8 (i.e., 80%), then the total delay time is 10 minutes * 0.8 = 8 minutes. Obtaining the vehicle delay time in this way allows drivers to know it in advance and take action accordingly, avoiding disruptions to traffic order and road safety.

[0042] Furthermore, such as Figure 2 As shown in the present application, the vehicle delay prediction system is also communicatively connected to a first electronic device, and the method further includes:

[0043] When a vehicle malfunctions, feedback information from a first electronic device is obtained, including the vehicle malfunction type and the vehicle malfunction level.

[0044] The repair time is obtained by analyzing the vehicle fault type and fault level.

[0045] The vehicle delay time is compensated based on the vehicle repair time.

[0046] Specifically, the first electronic device is for the driver and can be displayed on a computer terminal, allowing the driver to get the real-time driving status of the vehicle immediately. When the vehicle malfunctions, the first electronic device will provide feedback, which can be in the form of an alarm. Then, it will display the type and level of the vehicle's malfunction. The malfunction types include human-caused malfunctions and natural malfunctions. Human-caused malfunctions refer to malfunctions caused by human factors, such as using unqualified parts or violating assembly technical requirements, the vehicle not meeting the conditions of use, improper transportation and storage, etc. Natural failures refer to malfunctions caused by uncontrollable external and internal natural factors during the lifespan of a vehicle. Fault levels can be categorized as minor, general, serious, and fatal. Minor failures typically do not significantly impact vehicle performance, such as damage to small parts. General failures may cause the vehicle to stop or experience performance degradation, but generally do not damage major components or assemblies and can be repaired quickly with onboard tools, such as a damaged thermostat causing engine overheating. Serious failures damage major components and cannot be repaired quickly, such as an automatic transmission shifting gears. Fatal failures endanger driving safety, cause personal injury or death, render major assemblies unusable, result in significant economic losses, or severely harm the surrounding environment, such as excessive vehicle emissions, severe cylinder block cracks, or engine failure. Time adjustments are made based on repair time or the inability to repair. For example, if repair time is 30 minutes, the predicted delay time is the sum of the predicted and repair times.

[0047] Furthermore, such as Figure 3 As shown in the figure, the vehicle delay prediction system of this application is also communicatively connected to an image acquisition device, and further includes:

[0048] The image acquisition device performs real-time image acquisition on the M-1 nodes to obtain the image acquisition results.

[0049] The passenger count is analyzed based on the image acquisition results to obtain the passenger count at each node;

[0050] The boarding and alighting times are calculated based on the number of passengers, and the vehicle delay time is compensated based on the calculation results.

[0051] Specifically, select one or more suitable nodes from M-1 nodes and install the image acquisition device on the selected nodes. Different image acquisition devices, such as cameras, scanners, and infrared sensors, can be selected according to actual needs, ensuring the stability and reliability of the device. Set the acquisition parameters: Set the parameters of the image acquisition device as needed, such as resolution, exposure, and shooting angle, to ensure that the acquired images meet the requirements. Start the image acquisition device to begin real-time image acquisition from the node. Continuous or timed shooting methods can be used, selecting the appropriate acquisition method according to actual needs. Store the acquired image data in a designated storage device to ensure data security and reliability. During the acquisition process, constantly monitor image quality, such as sharpness and color reproduction. Regularly check the operating status of the image acquisition device during acquisition to ensure it functions properly. Regular maintenance and upkeep are also necessary to extend the device's lifespan. After acquisition is complete, the image acquisition results can be obtained by viewing the image data in the storage device. The acquired images can be processed and analyzed to obtain useful information or for subsequent decision-making. Passenger volume analysis is performed to obtain the number of passengers at each node. Calculations are made based on passenger boarding and alighting times. By comparing passenger numbers at different time points, trends in passenger volume changes can be analyzed, as well as the relationship between passenger volume and influencing factors such as weather, holidays, and public events. Analyzing the passenger volume distribution across M-1 nodes provides insight into passenger flow at each node and allows for analysis of passenger travel characteristics, such as travel time, mode of transport, and travel purpose. This analysis can predict passenger flow. If passenger flow is high and boarding and alighting times are long, waiting times at stations are increased by adding the predicted delay time to these waiting times.

[0052] Furthermore, the method of this application, which analyzes the number of passengers in the image acquisition results, also includes:

[0053] Target detection is performed on the image acquisition results to obtain real-time personnel count data for the nodes, including the number of people entering and leaving the station;

[0054] Specifically, object detection in image acquisition results requires algorithms to extract features representing various target objects from the image, such as color, shape, and texture. Then, a classifier is used to classify these features, determine the location and size of various target objects, and obtain real-time personnel count data for the nodes, including the number of people entering and leaving the station.

[0055] A preset time window is established. Based on the preset time window and the number of people entering and leaving the station, the average number of people is calculated to obtain the number of passengers at each node.

[0056] Specifically, a preset time window is established. Based on actual needs, a time window is set, such as every 10 minutes, every hour, or every day. This time window should be long enough to avoid excessive data noise due to its short duration, but also short enough to respond promptly to sudden traffic situations. The number of people entering and leaving the station at each node within the corresponding time period is obtained from the traffic data collection system. Within the preset time window, the number of people entering and leaving the station is added together to obtain the total number of people within this time window. Dividing the total number of people by the length of the time window yields the average number of people. Multiplying the average number of people within each time window by the length of the time window gives the number of passengers at each node within that time window.

[0057] Furthermore, the method of this application, which sorts the M predicted travel times and obtains the predicted delay time of the current node based on the sorting results, further includes:

[0058] Obtain the original travel time for the M road segments;

[0059] Calculate the difference between the M predicted travel times and the original travel times, and discard the differences that are less than a preset time threshold;

[0060] The remaining differences are summed to obtain the predicted delay time for the current node.

[0061] Specifically, the process involves obtaining the original travel times for M road segments, determining the predicted travel time for the current node, and retrieving historical travel times (e.g., 4 minutes per segment) from traffic data. The time difference between these two times is then calculated. This difference can be either the predicted travel time minus the original travel time, or the original travel time minus the predicted travel time, depending on the magnitude of the predicted and original times. A time threshold, such as X seconds or X minutes, is set based on actual needs. This threshold can be an empirical value or determined according to actual traffic conditions. The calculated time differences are then sorted, and differences smaller than the preset time threshold are discarded. These discarded differences may be small errors caused by inaccurate predictions or traffic fluctuations, and their impact on the predicted delay time can be considered minimal. The remaining differences are then summed up, i.e., all differences greater than or equal to the preset time threshold are added together. This summation is the predicted delay time for the current node.

[0062] Furthermore, the method of this application, which performs traffic impact analysis based on the weather data to obtain a traffic impact coefficient, also includes:

[0063] Obtain historical weather data and historical traffic data;

[0064] Specifically, historical weather and traffic data are collected from relevant data sources (such as meteorological bureaus and traffic management departments) or using publicly available APIs (such as OpenWeatherMap). The collected data is then cleaned to remove invalid, erroneous, or duplicate data, ensuring the quality and accuracy of the data.

[0065] Perform a correlation analysis on the historical weather data and the historical traffic data, and obtain the correlation analysis results;

[0066] Specifically, data preprocessing involves performing actions such as scaling, standardization, and outlier removal to make the data easier to analyze and compare. Statistical correlation analysis methods (such as Pearson correlation coefficient and Spearman rank correlation coefficient) are used to analyze the degree of correlation between weather and traffic data. The results of the correlation analysis are then used to determine the relationship between the two.

[0067] Based on the correlation analysis results, the impact of wind speed, rainfall, visibility, and weather warnings on traffic is calculated.

[0068] The impact levels are weighted, and based on the weighting results, the current wind speed, rainfall, visibility, and weather warnings are calculated to obtain the traffic impact coefficient.

[0069] Specifically, the impact of wind speed, rainfall, visibility, and weather warnings on traffic is calculated. Based on the correlation analysis results, a model is constructed to describe the impact of weather factors (wind speed, rainfall, visibility, etc.) on traffic. Linear regression models and neural network models can be used. The model is trained using historical data, and its performance is evaluated. Methods such as cross-validation are used to assess the model's predictive accuracy. Based on the model's prediction results and actual traffic data, the degree of impact of each weather factor on traffic is calculated.

[0070] Example 2

[0071] Based on the same inventive concept as the data analysis-based vehicle delay prediction method described in the foregoing embodiments, such as... Figure 4 As shown, this application provides a vehicle delay prediction system based on data analysis, the system comprising:

[0072] The first driving route acquisition module is used to acquire the first driving route of the first vehicle, which includes Q nodes and Q road segments.

[0073] The node extraction module extracts the current node, as well as the M-1 nodes and M road segments preceding the current node from the Q nodes.

[0074] The predicted travel time acquisition module is used to acquire traffic congestion information of the M road segments through the traffic management platform, analyze the travel time based on the traffic congestion information, and acquire M predicted travel times.

[0075] The predicted delay time acquisition module sorts the M predicted travel times and obtains the predicted delay time of the current node based on the sorting results.

[0076] The weather data acquisition module is used to acquire current weather data through a meteorological platform. The weather data includes at least wind speed, rainfall, visibility, and weather warnings.

[0077] A traffic impact coefficient acquisition module is used to perform traffic impact analysis based on the weather data and acquire a traffic impact coefficient.

[0078] The final delay time acquisition module is used to acquire the vehicle delay time of the current node based on the traffic impact coefficient and the predicted delay time.

[0079] Furthermore, the system also includes:

[0080] The feedback information acquisition module acquires feedback information from the first electronic device when a vehicle malfunctions. The feedback information includes the vehicle malfunction type and the vehicle malfunction level.

[0081] The repair time acquisition module analyzes repair time based on vehicle fault type and fault level to obtain vehicle repair time.

[0082] The time correction module corrects the vehicle delay time based on the vehicle repair time.

[0083] Furthermore, the system also includes:

[0084] The image acquisition result acquisition module performs real-time image acquisition on the M-1 nodes through the image acquisition device and acquires the image acquisition results.

[0085] The passenger count acquisition module analyzes the passenger count of the image acquisition results to obtain the passenger count of each node;

[0086] The time calculation module calculates the boarding and alighting times based on the number of passengers and compensates for the vehicle delay time based on the calculation results.

[0087] Furthermore, the system also includes:

[0088] The target detection module performs target detection on the image acquisition results and obtains real-time personnel count data for the nodes, including the number of people entering the station and the number of people leaving the station.

[0089] The passenger number prediction module establishes a preset time window, and calculates the average number of passengers at each node based on the preset time window and the number of people entering and leaving the station.

[0090] Furthermore, the system also includes:

[0091] The passage time acquisition module acquires the original passage time of the M road segments;

[0092] The difference elimination module calculates the difference between the M predicted travel times and the original travel time, and eliminates the differences that are less than a preset time threshold.

[0093] The delay time acquisition module accumulates the remaining differences to obtain the predicted delay time for the current node.

[0094] Furthermore, the system also includes:

[0095] The historical weather data acquisition module acquires historical weather data and historical traffic data.

[0096] The correlation analysis result acquisition module performs correlation analysis on the historical weather data and the historical traffic data, and obtains the correlation analysis result.

[0097] The impact degree acquisition module calculates the impact degree of wind speed, rainfall, visibility, and weather warnings on traffic based on the correlation degree analysis results.

[0098] The impact coefficient acquisition module weights the degree of impact and calculates the current wind speed, rainfall, visibility, and weather warnings based on the weighting results to obtain the traffic impact coefficient.

[0099] Through the detailed description of the data analysis-based vehicle delay prediction method described above, those skilled in the art can clearly understand the data analysis-based vehicle delay prediction system in this embodiment. As the system disclosed in the embodiment corresponds to the device disclosed in the embodiment, the description is relatively simple, and relevant parts can be referred to the method section.

[0100] The above description of the disclosed embodiments enables those skilled in the art to make or use this application. Various modifications to these embodiments will be readily 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 this application. Therefore, this application is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A vehicle delay prediction method based on data analysis, characterized in that, The method is applied to a vehicle delay prediction system, which is communicatively connected to a traffic management platform and a meteorological platform. The method includes: Obtain the first driving route of the first vehicle, which includes Q nodes and Q road segments; Extract the current node, the M-1 nodes preceding the current node, and the M road segments from the Q nodes; Traffic congestion information for the M road segments is obtained through a traffic management platform. Based on the traffic congestion information, travel time is analyzed to obtain M predicted travel times. The M predicted travel times are sorted, and the predicted delay time of the current node is obtained based on the sorting results; Current weather data is obtained through a meteorological platform, including at least wind speed, rainfall, visibility, and weather warnings. Based on the aforementioned weather data, a traffic impact analysis is performed to obtain the traffic impact coefficient; Based on the traffic impact coefficient and the predicted delay time, obtain the vehicle delay time at the current node; The vehicle delay prediction system is also communicatively connected to a first electronic device, and the method further includes: When a vehicle malfunctions, feedback information from a first electronic device is obtained, including the vehicle malfunction type and the vehicle malfunction level. The repair time is obtained by analyzing the vehicle fault type and fault level. The vehicle delay time is compensated based on the vehicle repair time; The process of sorting the M predicted travel times and obtaining the predicted delay time of the current node based on the sorting results includes: Obtain the original travel time for the M road segments; Calculate the difference between the M predicted travel times and the original travel times, and discard the differences that are less than a preset time threshold; The remaining differences are summed to obtain the predicted delay time for the current node.

2. The method as described in claim 1, characterized in that, The vehicle delay prediction system is also communicatively connected to an image acquisition device, and the method further includes: The image acquisition device performs real-time image acquisition on the M-1 nodes to obtain the image acquisition results. The passenger count is analyzed based on the image acquisition results to obtain the passenger count at each node; The boarding and alighting times are calculated based on the number of passengers, and the vehicle delay time is compensated based on the calculation results.

3. The method as described in claim 2, characterized in that, The passenger number analysis is performed on the image acquisition results, including: Target detection is performed on the image acquisition results to obtain real-time personnel count data for the nodes, including the number of people entering and leaving the station; A preset time window is established. Based on the preset time window and the number of people entering and leaving the station, the average number of people is calculated to obtain the number of passengers at each node.

4. The method as described in claim 1, characterized in that, Based on the aforementioned weather data, a traffic impact analysis is performed to obtain the traffic impact coefficient, including: Obtain historical weather data and historical traffic data; Perform a correlation analysis on the historical weather data and the historical traffic data, and obtain the correlation analysis results; Based on the correlation analysis results, the impact of wind speed, rainfall, visibility, and weather warnings on traffic is calculated. The impact levels are weighted, and based on the weighting results, the current wind speed, rainfall, visibility, and weather warnings are calculated to obtain the traffic impact coefficient.

5. A vehicle delay prediction system based on data analysis, characterized in that, The system is used to perform the method according to any one of claims 1 to 4, the system comprising: The first driving route acquisition module is used to acquire the first driving route of the first vehicle, which includes Q nodes and Q road segments. The node extraction module extracts the current node, as well as the M-1 nodes and M road segments preceding the current node from the Q nodes. The predicted travel time acquisition module is used to acquire traffic congestion information of the M road segments through the traffic management platform, analyze the travel time based on the traffic congestion information, and acquire M predicted travel times. The predicted delay time acquisition module sorts the M predicted travel times and obtains the predicted delay time of the current node based on the sorting results. The weather data acquisition module is used to acquire current weather data through a meteorological platform. The weather data includes at least wind speed, rainfall, visibility, and weather warnings. A traffic impact coefficient acquisition module is used to perform traffic impact analysis based on the weather data and acquire a traffic impact coefficient. The final delay time acquisition module is used to acquire the vehicle delay time of the current node based on the traffic impact coefficient and the predicted delay time. The feedback information acquisition module acquires feedback information from the first electronic device when a vehicle malfunctions. The feedback information includes the vehicle malfunction type and the vehicle malfunction level. The repair time acquisition module analyzes repair time based on vehicle fault type and fault level to obtain vehicle repair time. The time correction module corrects the vehicle delay time based on the vehicle repair time; The passage time acquisition module acquires the original passage time of the M road segments; The difference elimination module calculates the difference between the M predicted travel times and the original travel time, and eliminates the differences that are less than a preset time threshold. The delay time acquisition module accumulates the remaining differences to obtain the predicted delay time for the current node.

Images