Scenic spot vehicle scheduling method, scheduling device and computer equipment
By collecting individual passenger characteristic data and using a logistic regression model to predict the probability of getting off the bus, the problem of low scheduling efficiency in the scenic area vehicle scheduling system was solved, accurate vehicle scheduling was achieved, and the robustness of the system and passenger experience were improved.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- YANGZHOU YAXING MOTOR COACH
- Filing Date
- 2025-12-29
- Publication Date
- 2026-05-01
AI Technical Summary
The existing scenic area vehicle dispatch system suffers from low dispatch efficiency, excessive vehicle congestion or empty vehicles, and lacks accurate identification of passenger numbers, resulting in excessively long waiting times for passengers or waste of resources.
By collecting individual passenger characteristic data inside the vehicle, using a logistic regression model to predict the probability of passengers getting off the vehicle, and combining the number of people on the platform and the number of seats in the vehicle, vehicle dispatch instructions are generated to achieve precise dispatching.
It improves the foresight and accuracy of vehicle dispatching, avoids bias from single-dimensional judgment, enhances the robustness and adaptability of the system to different scenarios, and reduces passenger waiting time and resource waste.
Smart Images
Figure CN121961069A_ABST
Abstract
Description
A vehicle dispatching method, dispatching device, and computer equipment for scenic areas. Technical Field
[0001] This invention belongs to the field of intelligent transportation technology, specifically relating to a vehicle dispatching method, dispatching device, and computer equipment for scenic areas. Background Technology
[0002] Currently, public vehicle dispatching systems or scenic area vehicle dispatching systems typically rely on fixed schedules or driver reports. This leads to low dispatching efficiency, frequently resulting in overcrowded or empty vehicles, impacting passenger experience and operating costs. For example, during peak hours, platforms may be crowded with waiting passengers, but vehicles are already full, causing excessively long waiting times; while during off-peak hours, vehicles may run empty, resulting in wasted resources.
[0003] Some technologies attempt to monitor vehicle location using GPS or sensors, but lack accurate passenger count identification. Other solutions use simple counters or pressure sensors, but cannot distinguish between waiting passengers and pedestrians, resulting in low accuracy. Therefore, a system capable of accurately identifying passenger counts in real time and automatically dispatching passengers is needed. Summary of the Invention
[0004] To address the shortcomings of existing technologies, this paper provides a vehicle dispatching method, dispatching device, and computer equipment for scenic areas, aiming to solve the problem of low vehicle dispatching efficiency in scenic areas.
[0005] On the one hand, the technical solution of the present invention to solve the above-mentioned technical problems is as follows: a vehicle dispatching method for scenic spots, comprising the following steps: S1, collecting the number of passengers in the vehicle and the individual characteristic data of each passenger, wherein the individual characteristic data includes age attribute, weather condition, destination station type attribute, group attribute and travel time attribute; S2, collecting and counting the number of people in the queuing area of the platform; S3, inputting the individual characteristic data of each passenger into the disembarkation probability prediction model, and calculating the conditional probability value of each passenger disembarking at the next station; S4, calculating the conditional probability value of all passengers in the vehicle, and generating the predicted total number of people disembarking at the next station; S5, generating and executing vehicle dispatching instructions based on the comparison result of the predicted total number of people disembarking with a preset dispatching threshold.
[0006] By simultaneously collecting the number of passengers on the platform and inside the train, and collecting the characteristic attributes of each passenger inside the train, the prediction object involves specific passengers. The individual characteristic data of passengers inside the train are used to predict the probability of getting off the train using a prediction model, thereby calculating the total number of passengers getting off the train. Combined with the number of people on the platform and the number of seats in the train, it is determined whether it is necessary to perform advance dispatching of the train, which improves the foresight.
[0007] In one embodiment, the alighting probability prediction model in step S3 is a logistic regression model, and the formula for calculating the conditional probability value P of each passenger alighting is: P_i=1 / [1+exp(-z_i)]; where z=β0+β_age+β_weather+β_scenic+β_group+β_time+β_interaction×(β_age×β_scenic); where i is the passenger number, β0 is the bias term, β_interaction×(β_age×β_scenic) is the interaction term, β_interaction is the interaction term weight coefficient, β_age is the age attribute weight parameter, β_weather is the weather state weight parameter, β_scenic is the station type weight parameter, β_group is the group attribute weight parameter, and β_time is the travel time period weight parameter.
[0008] By combining information such as age, weather, attraction attributes, and time of day, the system can comprehensively determine the willingness of people to leave each station, avoiding bias from a single dimension and making the system more robust and adaptable to different scenarios.
[0009] In one embodiment, the values of the weight parameters are obtained through training with historical travel data, and their value ranges are as follows: The weight ranges for different age groups in β_age are: first age group (0-18 years old) [-0.5, -0.1], second age group (19-35 years old) [0.5, 1.0], third age group (36-60 years old) [-0.1, 0.3], fourth age group (61 years and above) [0.2, 0.6]; the weight ranges for different weather conditions in β_weather are: sunny [0.5, 0.7], cloudy [0.2, 0.4], light rain [-0.3, -0.1], heavy rain or snow [-0.9, -0.7]; the weight ranges for different station types in β_scenic are: popular tourist attractions [1.0, 1.4], general tourist attractions [0.1, 1.4], and general tourist attractions [0.1, 1.4]. 0.5], less popular tourist attractions [-0.7, -0.3]; the weight ranges for different group identifiers in β_group are: tour group members [1.3, 1.7] and individual tourists [0, 0]; the weight ranges for different travel times in β_time are: morning peak [-0.5, -0.3], daytime [0.5, 0.9], evening peak [-0.4, -0.2], and nighttime [0.1, 0.3].
[0010] In one embodiment, in step S4, the predicted total number of passengers getting off the bus is generated using the following formula: E=ΣP_i; the conditional probability values of all passengers on the bus are summed, and the summation result is used as the predicted total number of passengers getting off the bus E.
[0011] In one embodiment, step S5 includes: based on the number of passengers K in the vehicle and the number of people Q in the platform queuing area; calculating the remaining number of seats S = (CK) + E according to the predicted total number of people getting off the vehicle; comparing the remaining number of seats S with the number of people waiting on the platform Q: if S ≥ Q, then no other vehicle is called; if Q ≥ γ × C, then other vehicles need to be called; where C is the rated seating capacity of the vehicle and γ is a preset coefficient.
[0012] In one embodiment, the specific methods for collecting individual characteristic data in step S1 include: collecting the number of passengers in the vehicle through an in-vehicle camera; obtaining the age attributes of passengers through a scenic area real-name system or an in-vehicle image recognition device; receiving meteorological service data through an in-vehicle communication module to determine the weather conditions; determining the type attribute of the destination station through interaction between the vehicle positioning system and the scenic area heat database; determining the group attribute by identifying uniform objects through a ticketing system, registration information, or an image recognition device; and obtaining the travel time attribute through the system clock.
[0013] In one embodiment, the method further includes step S6: periodically collecting actual disembarkation data, using the actual disembarkation data and prediction results as training samples, and retraining the disembarkation probability prediction model to update the weight parameters.
[0014] Secondly, this embodiment also discloses a vehicle dispatching device for implementing the above-mentioned vehicle dispatching method for scenic areas. The dispatching device includes: an in-vehicle data acquisition module for collecting the number of passengers in the vehicle and the individual characteristic data of each passenger, the individual characteristic data including age attribute, weather condition, destination station type attribute, group attribute, and travel time attribute; a platform data acquisition module for collecting and counting the number of people in the platform queuing area; a probability calculation module for inputting the individual characteristic data of each passenger into a disembarkation probability prediction model to calculate the conditional probability value of each passenger disembarking at the next station; a number prediction module for calculating the conditional probability value of all passengers in the vehicle to generate the predicted total number of people disembarking at the next station; and a dispatching execution module for generating and executing vehicle dispatching instructions based on the comparison result of the predicted total number of people disembarking with a preset dispatching threshold.
[0015] On the other hand, this embodiment also discloses a computer device, including a memory, a processor, and a computer program stored in the memory, characterized in that the processor implements the above-described vehicle scheduling method when executing the computer program. Attached Figure Description
[0016] To more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the specific embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.
[0017] Figure 1 is a schematic flowchart of the method in this invention. Detailed Implementation
[0018] The embodiments of the technical solution of the present invention will now be described in detail with reference to the accompanying drawings. The following embodiments are only used to more clearly illustrate the technical solution of the present invention, and are therefore merely examples and should not be used to limit the scope of protection of the present invention. It should be noted that, unless otherwise stated, the technical or scientific terms used in this application should have the ordinary meaning understood by those skilled in the art to which this invention pertains.
[0019] In the description of this application, it should be understood that the terms "upper", "lower", "front", "rear", "left", "right", "top", "bottom", "inner", "outer", etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are only for the convenience of describing the present invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limiting the present invention.
[0020] Furthermore, the terms "first," "second," etc., are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. In the description of this invention, "a plurality of" means two or more, unless otherwise explicitly defined.
[0021] In this application, unless otherwise expressly specified and limited, the terms "installation," "connection," "linking," and "fixing," etc., should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral part; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; they can refer to the internal communication of two components or the interaction between two components. Those skilled in the art can understand the specific meaning of the above terms in this invention according to the specific circumstances.
[0022] Example 1, as shown in Figure 1, provides a vehicle dispatching method for scenic areas, comprising the following steps: Step S1, collecting the number of passengers in the vehicle and the individual characteristic data of each passenger, wherein the individual characteristic data includes age attribute, weather condition, destination station type attribute, group attribute, and travel time attribute; specifically, the number of passengers in the vehicle can be collected and counted through an in-vehicle camera, the age attribute of passengers can be obtained through the scenic area real-name system or vehicle-mounted image recognition device, and the age information of each passenger can be determined by comparing the facial images collected by the camera with the scenic area real-name ticketing system, and the passengers can be numbered and their characteristic attributes matched; specifically, through The vehicle-mounted communication module receives meteorological service data to determine weather conditions; through interaction between the vehicle positioning system and the scenic area popularity database, it determines the type and attribute of the destination station, i.e., it judges the popularity of attractions near the next station; it identifies group attributes by recognizing uniform objects through the ticketing system, registration information, or image recognition devices. For example, passengers who purchase tickets or register as a group are marked as having a group attribute. At the same time, it can combine image recognition analysis of camera footage to determine whether there are uniform objects, such as uniform tourist hats or vests. If more than 3 identical objects are found, it is marked as having a group attribute. The specific number of identical objects can be adjusted; the time period attribute can be obtained through the system clock.
[0023] Step S2: Collect and count the number of people in the platform queuing area; The number of people in the platform queuing area can be counted by taking real-time images of the platform through cameras installed on the platform, and the number of people in the queuing area can be identified by image recognition algorithms.
[0024] In the queuing area, physical barriers (such as railings) can help separate people waiting for a bus from pedestrians passing by.
[0025] In conjunction with time attributes, the data collection unit records the dwell time of people, and includes people whose dwell time exceeds a preset value (e.g., 30 seconds) in the number of people waiting for the train, thereby calculating the final number of people on the platform.
[0026] Step S3: Input the individual characteristic data of each passenger into the disembarkation probability prediction model to calculate the conditional probability value of each passenger getting off at the next stop. To calculate the disembarkation probability of each passenger more accurately, a prediction model is introduced, and a feature vector is introduced for each passenger. The disembarkation probability prediction model is a logistic regression model, and its formula for calculating the conditional probability value P of a passenger getting off is: P_i=1 / [1+exp(-z_i)]; where i is the passenger number. The passenger number can be assigned by combining the boarding order with the date or by using the ID card number, so that each passenger corresponds to a different number.
[0027] Where z = β0 + β_age + β_weather + β_scenic + β_group + β_time + β_interaction × (β_age × β_scenic); where β0 is the bias term, β_interaction × (β_age × β_scenic) is the interaction term, β_interaction is the interaction term weight coefficient; β_age is the age attribute weight parameter, β_weather is the weather state weight parameter, β_scenic is the station type weight parameter, β_group is the group attribute weight parameter, and β_time is the travel time period weight parameter.
[0028] By assigning the above weighted parameters to each passenger to form a combined value z, and combining information such as age, weather, attraction attributes, and time of day, the system can comprehensively determine the passengers' willingness to get off at each station, avoiding bias in judgment based on a single dimension, and making the system more robust and adaptable to different scenarios.
[0029] The values of the weight parameters are obtained through training with historical travel data, and their range is as follows: In this embodiment, the bias term β0 is pre-set to -2.8, and the bias term is the basic probability tendency when there is no feature information.
[0030] The weight ranges for different age groups in β_age are as follows: first age group [-0.5, -0.1], second age group [0.5, 1.0], third age group [-0.1, 0.3], and fourth age group [0.2, 0.6]. The age ranges from the first to the fourth age group can be divided into (0-18 years old), (19-35 years old), (36-60 years old), and (61 years old and above), which are used to divide the group into youth, young adult, middle-aged, and elderly groups in sequence. Corresponding to different age groups, teenagers usually follow fixed routes, young adults are usually more exploratory, middle-aged adults are more regular, and the elderly have more free time but limited physical strength. Therefore, the positive correlation from strong to weak is youth, elderly, middle-aged, and teenager.
[0031] The weight ranges for different weather conditions in β_weather are as follows: sunny [0.5, 0.7], cloudy [0.2, 0.4], light rain [-0.3, -0.1], and heavy rain or snow [-0.9, -0.7]. By classifying the weather into different types, the willingness to get off the bus can be greatly correlated. When the weather is sunny, the willingness to get off the bus for sightseeing is high. Secondly, the weights of the progressively worsening weather conditions, such as cloudy, light rain, and heavy rain, decrease accordingly. This means that the worse the weather, the lower the willingness to get off the bus. The specific weather classification can be based on the degree of sunshine, and is not limited to sunny, cloudy, light rain, and heavy rain.
[0032] The weight ranges for different site types in β_scenic are as follows: popular attractions [1.0, 1.4], general attractions [0.1, 0.5], and less popular attractions [-0.7, -0.3]. Similarly, classifying attractions into different types can greatly correlate with disembarkation intentions. When an attraction is popular, passengers are more willing to get off. As the popularity of an attraction gradually decreases to general and less popular, the willingness to get off gradually decreases, and the weights also gradually decrease.
[0033] The weight ranges for different team identifiers in β_group are: tour group members [1.3, 1.7] and individual travelers [0, 0]; while the group attribute is assigned a weight of 0 because the itinerary is fixed and the group will get off at the destination. Individual travelers are only used as a distinction and do not have an additional weight.
[0034] The weighting ranges for different travel periods in β_time are as follows: morning peak [-0.5, -0.3], daytime [0.5, 0.9], evening peak [-0.4, -0.2], and nighttime [0.1, 0.3]. Considering the time period, during the morning and evening peaks, commuting is the main activity, with less tourism. Daytime is the peak tourism period, hence the higher weighting. Currently, there are more nighttime tourism activities, so the weighting of nighttime is higher than that of morning and evening peaks, but lower than that of daytime.
[0035] At the same time, interactive items are introduced first, and the combined effect of age and attraction type is considered. Specifically, the weight coefficient β_interaction of interactive items can be increased by 50%.
[0036] For example, when the pre-selected values for each weight vector are as follows: Age weight (teenager, youth, middle-aged, elderly); β_age=[-0.3,0.8,0.1,0.4]; Weather weight (sunny, cloudy, light rain, heavy rain / snow); β_weather=[0.6,0.3,-0.2,-0.8]; Scenic spot weight (popular, average, less popular); β_scenic=[1.2,0.3,-0.5]; Tour group weight (is a tour group, individual traveler); β_group=[1.5,0.0]; Time of day weight (morning peak, daytime, evening). (Peak hours, nighttime); β_time=[-0.4,0.7,-0.3,0.2]; At this time, when a group of young tourists goes to a popular attraction in good weather, the probability of them getting off the bus is calculated as: z=-2.8+0.8+0.6+1.2+1.5+0.7+0.5×(0.8×1.2)=2.48; the probability of them getting off the bus is P_i=1 / [1+exp(-2.48)]=1 / 1.084≈0.923; then it can be predicted that the probability of the members of this group of young tourists getting off the bus at the next stop is 92.3% when they go to a popular attraction in good weather during the day.
[0037] For example, if a middle-aged individual traveler is heading to a less popular tourist attraction during the morning rush hour on a rainy day, the probability of them getting off the bus is calculated as: z = -2.8 + 0.1 + (-0.2) + (-0.5) + 0 + (-0.4) + 0.5 × (0.1 × (-0.5)) = -3.825; their probability of getting off the bus is P_i = 1 / [1 + exp(-2.48)] = 1 / 46.86 ≈ 0.021; therefore, it can be predicted that the probability of this middle-aged individual traveler getting off at the next stop during the morning rush hour on a light rainy day is only 2.1%.
[0038] For example, if an elderly tour group travels to a general tourist attraction during the day on a cloudy day, the probability of them getting off the bus is calculated as: z = -2.8 + 0.4 + 0.3 + 0.3 + 1.5 + 0.7 + 0.5 × (0.4 × 0.3) = 0.46; their probability of getting off the bus is P_i = 1 / [1 + exp(-2.48)] = 1 / 1.631 ≈ 0.613; therefore, it can be predicted that the probability of the elderly tour group getting off at the next stop on a cloudy day is 61.3%. Step S4: Calculate the conditional probability values of all passengers on the bus and generate the predicted total number of people getting off at the next stop; specifically, the predicted total number of people getting off the bus is generated using the following formula: E = ΣP_i; sum the conditional probability values of all passengers on the bus, and use the summation result as the predicted total number of people getting off the bus E.
[0039] Step S5: Based on the comparison result between the predicted total number of passengers getting off the bus and the preset scheduling threshold, generate and execute vehicle scheduling instructions.
[0040] Specifically, by simultaneously collecting the number of passengers on the platform and inside the vehicle, and collecting the characteristic attributes of each passenger inside the vehicle, the prediction object involves specific passengers. The individual characteristic data of passengers inside the vehicle are used to predict the probability of getting off the vehicle using a prediction model, thereby calculating the total number of passengers getting off the vehicle. Combined with the number of people on the platform and the number of seats in the vehicle, it is determined whether it is necessary to perform the operation of pre-scheduling the vehicle, thus improving the foresight.
[0041] Specifically, step S5 includes: based on the number of passengers K inside the vehicle and the number of people Q in the platform queuing area; wherein the number of passengers inside the vehicle is collected and counted by the in-vehicle camera and recorded as the number K, and the number of people in the platform queuing area is collected and counted by the platform camera and recorded as the number Q.
[0042] Based on the predicted total number of passengers disembarking, E, calculate the remaining number of seats: S = (CK) + E; compare the remaining number of seats S with the number of passengers waiting on the platform, Q: if S ≥ Q, it means there are enough seats in the vehicle to accommodate the platform passengers, and no other vehicle needs to be called; if Q ≥ γ × C, where C is the vehicle's rated seating capacity and γ is a preset coefficient, the preset coefficient γ takes a value of 1.4-1.6, which can be optimized based on historical data to form the preset coefficient. When Q ≥ γ × C, it means there are too many passengers on the platform, and additional vehicles need to be called.
[0043] When additional vehicles are needed, an alert can be sent to the dispatch center via a wireless communication module, including the platform location, the number of waiting passengers, and the suggested number of vehicles to be dispatched. The dispatch center can then manually or automatically dispatch backup vehicles.
[0044] It also includes step S6, periodically collecting actual disembarkation data, using the actual disembarkation data and prediction results as training samples to retrain the disembarkation probability prediction model in order to update the weight parameters.
[0045] Specifically, Example 2 also discloses a vehicle dispatching device for implementing the above-mentioned vehicle dispatching method for scenic areas. The dispatching device includes: an in-vehicle data acquisition module for collecting the number of passengers in the vehicle and the individual characteristic data of each passenger, including age attributes, weather conditions, destination station type attributes, group attributes, and travel time attributes; a platform data acquisition module for collecting and counting the number of people in the platform queuing area; a probability calculation module for inputting the individual characteristic data of each passenger into a disembarkation probability prediction model to calculate the conditional probability value of each passenger disembarking at the next station; a number prediction module for calculating the conditional probability value of all passengers in the vehicle to generate the predicted total number of people disembarking at the next station; and a dispatching execution module for generating and executing vehicle dispatching instructions based on the comparison result of the predicted total number of people disembarking with a preset dispatching threshold.
[0046] Example 3 Those skilled in the art will understand that all or part of the steps in the above embodiments can be implemented using a computer program flow. This embodiment also discloses a computer device, including a memory, a processor, and a computer program stored in the memory. When the processor executes the computer program, it implements the above-described vehicle scheduling method. During execution, it includes one or a combination of the steps of the method embodiments.
[0047] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.
Claims
1. A vehicle dispatching method for scenic areas, characterized in that, Includes the following steps: S1. Collect the number of passengers in the vehicle and the individual characteristic data of each passenger. The individual characteristic data includes age attribute, weather condition, destination station type attribute, group attribute and travel time attribute. S2. Collect and count the number of people in the queuing area of the platform; S3. Input the individual characteristic data of each passenger into the disembarkation probability prediction model to calculate the conditional probability value of each passenger getting off at the next station. S4. Calculate the conditional probability values of all passengers on the train and generate the predicted total number of passengers getting off at the next stop. S5. Based on the comparison result between the predicted total number of passengers getting off the vehicle and the preset scheduling threshold, generate and execute vehicle scheduling instructions.
2. The vehicle dispatching method for scenic areas according to claim 1, characterized in that, The disembarkation probability prediction model in step S3 is a logistic regression model. The formula for calculating the conditional probability value P of each passenger disembarking is: P_i=1 / [1+exp(-z_i)]; where z=β0+β_age+β_weather+β_scenic+β_group+β_time+β_interaction×(β_age×β_scenic); where i is the passenger number, β0 is the bias term, β_interaction×(β_age×β_scenic) is the interaction term, β_interaction is the interaction term weight coefficient, β_age is the age attribute weight parameter, β_weather is the weather state weight parameter, β_scenic is the station type weight parameter, β_group is the group attribute weight parameter, and β_time is the travel time period weight parameter.
3. The vehicle dispatching method for scenic areas according to claim 2, characterized in that, The values of the weight parameters are obtained through training on historical travel data, and their ranges are as follows: The weight ranges for different age groups in β_age are: first age group [-0.5, -0.1], second age group [0.5, 1.0], third age group [-0.1, 0.3], fourth age group [0.2, 0.6]; the weight ranges for different weather conditions in β_weather are: sunny [0.5, 0.7], cloudy [0.2, 0.4], light rain [-0.3, -0.1], heavy rain or snow [-0.9, -0.7]; the weight ranges for different station types in β_scenic are: popular tourist attractions [1.0, 1.4], general tourist attractions [0.1, 0.5], less popular tourist attractions [-0.7, 1.4], and less popular tourist attractions [-0.7, 1.4]. -0.3]; The weight ranges for different team identifiers in β_group are: tour group members [1.3, 1.7] and individual travelers [0, 0]; The weight ranges for different travel times in β_time are: morning peak [-0.5, -0.3], daytime [0.5, 0.9], evening peak [-0.4, -0.2], and nighttime [0.1, 0.3].
4. The vehicle dispatching method for scenic areas according to claim 2, characterized in that, In step S4, the predicted total number of passengers getting off the bus is generated using the following formula: E=ΣP_i; the conditional probability values of all passengers on the bus are summed, and the summation result is used as the predicted total number of passengers getting off the bus E.
5. The vehicle dispatching method for scenic areas according to claim 4, characterized in that, Step S5 includes: based on the number of passengers K in the vehicle and the number of people Q in the platform queuing area; calculating the remaining number of seats S = (CK) + E according to the predicted total number of people getting off the vehicle; comparing the remaining number of seats S with the number of people waiting on the platform Q: if S ≥ Q, then no other vehicles are called; if Q ≥ γ × C, then other vehicles need to be called; where C is the rated seating capacity of the vehicle and γ is a preset coefficient.
6. The vehicle dispatching method for scenic areas according to claim 1, characterized in that, In step S1, the specific methods for collecting individual characteristic data include: collecting the number of passengers in the vehicle through an in-vehicle camera; obtaining the age attributes of passengers through the scenic area's real-name system or in-vehicle image recognition device; receiving meteorological service data through the in-vehicle communication module to determine the weather conditions; determining the type attribute of the destination station through the interaction between the vehicle positioning system and the scenic area's heat database; determining the group attribute by identifying uniform objects through the ticketing system, registration information, or image recognition device; and obtaining the travel time attribute through the system clock.
7. The vehicle dispatching method for scenic areas according to claim 1, characterized in that, The method further includes step S6: periodically collecting actual disembarkation data, using the actual disembarkation data and prediction results as training samples, and retraining the disembarkation probability prediction model to update the weight parameters.
8. A vehicle dispatching device, characterized in that, The vehicle dispatching method for implementing the scenic area according to any one of claims 1-7, the dispatching device comprising: an in-vehicle data acquisition module for collecting the number of passengers in the vehicle and individual characteristic data of each passenger, the individual characteristic data including age attribute, weather condition, destination station type attribute, group attribute, and travel time attribute; a platform data acquisition module for collecting and counting the number of people in the platform queuing area; a probability calculation module for inputting the individual characteristic data of each passenger into a disembarkation probability prediction model to calculate the conditional probability value of each passenger disembarking at the next station; a number prediction module for calculating the conditional probability value of all passengers in the vehicle to generate the predicted total number of people disembarking at the next station; and a dispatching execution module for generating and executing vehicle dispatching instructions based on the comparison result of the predicted total number of people disembarking with a preset dispatching threshold.
9. A computer device, comprising a memory, a processor, and a computer program stored in the memory, characterized in that, When the processor executes the computer program, it implements the vehicle scheduling method as described in any one of claims 1-7.