SaaS-based online car-hailing management method and system for bus hub parking lots

Abstract

The present invention relates to the field of data management, and discloses a SaaS-based online car-hailing management method and system for a bus hub parking lot, which is used to solve the problem that when an online car-hailing vehicle has arrived at a parking spot but the order is canceled by a passenger, the loss of the online car-hailing vehicle will increase, and the user experience of the online car-hailing vehicle on the platform will be reduced. The method includes collecting the driver's willingness through the platform. If the driver chooses to continue to accept orders in the bus hub parking lot, the online car-hailing vehicle information is collected, and an order acceptance priority index is obtained through analysis. The subsequent actions of the online car-hailing vehicle are judged based on the order acceptance priority index and a preset threshold, and the online car-hailing vehicle that is allowed to stay in the bus hub parking lot is screened out. The demand for taxis is evaluated and judged based on the parking lot information, and the online car-hailing vehicle that is allowed to stay in the bus hub parking lot is allocated to the area with the largest taxi demand to continue to accept orders, thereby effectively reducing the loss of the online car-hailing vehicle and improving the user experience of the online car-hailing vehicle driver.

Description

Technical Field

[0001] The present invention relates to the field of data management, and more specifically to a SaaS-based online car-hailing management method and system for a bus hub parking lot. Background Art

[0002] SaaS is an Internet-based software delivery model. In this model, software service providers host applications on cloud servers, and users access and use these applications through the Internet without having to install, maintain or manage the software locally. Bus hub parking lots refer to parking lots located at urban transportation hubs or important nodes of public transportation. They are mainly used to park public transportation vehicles, private vehicles, and vehicles serving public transportation, such as online car-hailing services and taxis. Such parking lots are usually located near transportation hubs such as bus stations, subway stations, and railway stations, providing passengers with transfer, parking and other services to facilitate their public transportation travel.

[0003] The existing online car-hailing management method for bus hub parking lots is that the online car-hailing driver receives the passenger's reservation request on the platform, arranges the trip according to the reservation time and location, and stops at the designated location of the online car-hailing after arriving at the parking lot and waits for the passenger to get on the car. However, this management method may result in the passenger canceling the order when the online car-hailing car has arrived at the parking location and is waiting for the passenger, thereby increasing the loss of the online car-hailing car and reducing the user experience of the online car-hailing car on the platform.

[0004] In view of the above problems, the present invention proposes a solution. Summary of the invention

[0005] In order to overcome the above-mentioned defects of the prior art, the present invention provides a SaaS-based online car-hailing management method and system for bus hub parking lots to solve the problems existing in the above-mentioned background technology.

[0006] To achieve the above object, the present invention provides the following technical solutions:

[0007] The SaaS-based online car-hailing management method for bus hub parking lots includes the following steps:

[0008] S001: The online car-hailing vehicle receives a passenger order and arrives at the bus hub parking lot. If the passenger cancels the order before the departure time, the driver's willingness is collected through the platform. If the driver chooses to leave voluntarily, no data collection and analysis is performed, and the driver is prompted to leave the bus hub parking lot according to the route. If the driver chooses to continue to accept orders in the bus hub parking lot, the online car-hailing vehicle information is collected;

[0009] S002: Analyze and obtain a re-order priority index based on the collected online car-hailing information, and compare the re-order priority index with a preset threshold. If the re-order priority index is less than the preset threshold, it is determined that the cancellation of the order has little impact on the driver, and the driver is instructed to leave the bus hub parking lot and re-accept the order outside the bus hub parking lot. If the re-order priority index is greater than the preset threshold, it is determined that the cancellation of the order has a greater impact on the driver, and the driver is allowed to stay in the bus hub parking lot, and parking lot information is collected;

[0010] S003: Assess and determine the demand for taxis based on parking lot information, and allocate online-hailing taxis that are allowed to stay in bus hub parking lots to areas with the highest demand for taxis to continue accepting orders.

[0011] Preferably, the steps for calculating the re-order priority index are:

[0012] The pick-up distance of the cancelled order is obtained through the GPS positioning device, and the travel time of the cancelled order is calculated, and the pick-up distance and the travel time are multiplied to obtain the cost coefficient of the cancelled order;

[0013] The time accident coefficient is calculated by obtaining the interval between the time when the passenger cancels the order and the time when the trip departs through the online car-hailing platform. The calculation formula is: Among them, AN represents the time accident coefficient, T2 and T1 represent the departure time of the trip and the time when the passenger cancels the order;

[0014] Obtain the frequency of cancellations of online ride-hailing orders by passengers during a period of time through the online ride-hailing platform, and record it as the breach of contract coefficient;

[0015] The priority index for accepting orders again is calculated based on the online car-hailing cost coefficient, time accident coefficient and breach of contract coefficient, and its calculation formula is TA=a1×CM+a2×AN+a3×CF, where TA represents the priority index for accepting orders again, CM represents the online car-hailing cost coefficient, AN represents the time accident coefficient, CF represents the breach of contract coefficient, and a1, a2, and a3 represent the weight coefficients of the online car-hailing cost coefficient, time accident coefficient and breach of contract coefficient.

[0016] Preferably, the step of obtaining the pick-up distance of the cancelled order of the online car-hailing service through the GPS positioning device is:

[0017] The GPS positioning device installed on the online car-hailing vehicle is used to collect the latitude and longitude of the starting location of the cancelled order and the current location of the bus hub parking lot, and these are recorded as the starting location and the ending location;

[0018] Convert the latitude and longitude coordinates of the starting and ending positions into radians;

[0019] Use the Haversine formula to calculate the spherical distance, and the calculation formula is: Where SJ is the pick-up distance, Δlat is the difference in latitude between the starting position and the ending position in radians, Δlon is the difference in longitude between the starting position and the ending position in radians, lat1 and lat2 are the latitudes of the starting position and the ending position in radians respectively, and R is the radius of the earth.

[0020] Preferably, the steps of evaluating and judging the taxi demand based on the parking lot information are:

[0021] The parking lot is divided into n parts, recorded as sub-parking lot areas, and the data of passenger boarding and alighting points in each sub-parking lot area are collected. The data includes the density of boarding and alighting points, the density of alighting points, and the correlation of boarding and alighting points. The taxi hotspot is calculated based on the data of passenger boarding and alighting points. The calculation formula is: Where PT represents the taxi hotspot, BD represents the pick-up point density, which is the ratio of the number of repeated pick-up points in the sub-parking lot area to the area of ​​the area, OD represents the drop-off point density, which is the ratio of the number of repeated drop-off points in the sub-parking lot area to the area of ​​the area, and RV represents the relevance of the pick-up and drop-off points;

[0022] Collect passenger flow data, analyze the passenger flow data, and calculate the passenger flow coefficient;

[0023] The calculated passenger flow coefficient is used as a clustering feature to cluster the passenger flow coefficient of each sub-parking lot area, and the passenger flow distribution coefficient is calculated and evaluated based on the clustering results;

[0024] The taxi demand index is calculated based on the parking space occupancy rate, taxi hotspot and passenger flow distribution coefficient. The calculation formula is TX=b1×PT+b2×PD, where TX represents the taxi demand index, PT represents the taxi hotspot, PD represents the passenger flow distribution coefficient, and b1 and b2 represent the weight coefficients of the taxi hotspot and the passenger flow distribution coefficient.

[0025] Preferably, the step of calculating the association between boarding and alighting points is as follows:

[0026] Collect the number of pick-up and drop-off points in all parking lot areas and calculate the average number of pick-up and drop-off points;

[0027] The Pearson correlation coefficient formula is used to calculate the correlation between boarding and alighting points. The calculation formula is: Where XG represents the association between boarding and alighting points, X represents the number of boarding points, and Y represents the number of alighting points. Expressed as the average boarding points, Expressed as the average number of drop-off points.

[0028] Preferably, the passenger flow coefficient calculation step is:

[0029] Count the number of vehicles entering and leaving each sub-parking lot area, that is, the number of times vehicles enter and leave the area during the detection period. The calculation formula is N total =N in +N out , where N total Represents the number of vehicles entering and exiting each sub-parking lot area, N in It is represented by the number of vehicles entering the sub-parking lot area during the detection period, N out It is represented by the number of vehicles that leave the sub-parking lot area during the detection period;

[0030] Counting the total number of vehicles in the area and the time the vehicles stay in the area during the detection time period, and calculating the ratio of the total number of vehicles to the time the vehicles stay in the area to obtain an average stay time, wherein the total number of vehicles includes the number of vehicles that have left and the number of vehicles that have not left;

[0031] The vehicle flow density of each sub-parking lot area is calculated by comparing the number of vehicles entering and exiting each sub-parking lot area with the detection time period;

[0032] The passenger flow coefficient of each sub-parking lot area is obtained by evaluating the number of vehicles entering and exiting each sub-parking lot area, the average stay time, and the traffic density. The calculation formula is: Where PF is the passenger flow coefficient, N total Represented as the number of vehicles entering and exiting each sub-parking lot area, T avg It is expressed as the average stay time, Density is the traffic density, NUM park Expressed as the number of parking spaces in a sub-parking lot area.

[0033] Preferably, the step of clustering the passenger flow coefficient of each sub-parking lot area is:

[0034] Step 1: Take the passenger flow coefficient as the clustering feature and the passenger flow coefficient of each sub-parking lot area as the data point to obtain a set of data sets, and use the elbow method to determine the optimal number of clusters;

[0035] Step 2: Randomly select a certain number of data points as the initial cluster centers, and use the Euclidean distance to calculate the distance between each data point in the data set and each initial cluster center. The calculation formula is: Among them, x ik Represented as x i The value of the k-th dimension feature, c jk Represents the initial cluster center c jThe value of the k-th dimension feature, n represents the number of data points in the data set, and the data point is assigned to the cluster with the nearest cluster center;

[0036] Step 3: Traverse all the data points in the data set to obtain the initial clusters. For each initial cluster, calculate the average value of all data points in the cluster to obtain the new cluster center. The calculation formula is: Where V j Represents the new cluster center of the jth initial cluster, x i Represents cluster c j The data points in c j Represented as the initial clustering cluster, |C j |denoted as cluster c j The number of data points;

[0037] Repeat steps 2 and 3 until the cluster center no longer changes, and obtain the final cluster and the final cluster center.

[0038] Preferably, the step of evaluating and calculating the passenger flow distribution coefficient based on the clustering result is:

[0039] According to the total number of data points and the number of data points in each final cluster, the weight of each final cluster is calculated, and the calculation formula is: Among them WG j Indicates the weight of the jth final cluster, NUM j represents the number of data points in the jth final cluster, and n represents the number of data points in the data set;

[0040] The final cluster center of each final cluster is collected, and the passenger flow distribution coefficient is calculated based on the weight of the final cluster center and the final cluster. The calculation formula is PD j =WG j ×ZX j , where PD j It is expressed as the passenger flow distribution coefficient in the jth final cluster, then the passenger flow distribution coefficient is PD, WG j represents the weight of the jth final cluster, ZX j Denotes the final cluster center of the jth final cluster.

[0041] Preferably, the online car-hailing management system for bus hub parking lots based on SaaS includes:

[0042] The driver intention collection module is used to collect the intentions of online car-hailing drivers whose orders have been cancelled, including whether to leave the parking lot or stay in the parking lot to continue accepting orders;

[0043] The online car-hailing data collection module is used to collect relevant data information of online car-hailing vehicles that choose to stay in the parking lot and continue to receive orders;

[0044] The online car-hailing status judgment module is used to conduct a comprehensive evaluation based on the relevant data information of the online car-hailing vehicle collected by the online car-hailing data collection module, and determine whether to keep or remove the online car-hailing vehicle based on the evaluation results;

[0045] The parking lot status evaluation module is used to collect the current data information of the parking lot and evaluate the parking lot status based on the current data information of the parking lot;

[0046] The online car-hailing allocation module is used to allocate the online car-hailing vehicles that are determined to be left in the parking lot to the area with the greatest demand for taxis based on the parking lot status.

[0047] Technical effects and advantages of the present invention:

[0048] The platform collects drivers' intentions. If the driver chooses to continue accepting orders in the bus hub parking lot, the platform collects online car-hailing information and analyzes it to obtain a priority index for accepting orders again. The platform then determines the subsequent actions of the online car-hailing vehicle based on the priority index for accepting orders again and the preset threshold. The online car-hailing vehicle that is allowed to stay in the bus hub parking lot is screened out, and the demand for taxis is evaluated and judged based on the parking lot information. The online car-hailing vehicle that is allowed to stay in the bus hub parking lot is allocated to the area with the highest demand for taxis to continue accepting orders, effectively reducing the losses of online car-hailing vehicles and improving the experience of online car-hailing drivers. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] Figure 1 It is the overall flow chart of the present invention.

[0050] Figure 2 This is a flow chart of calculating the order priority index for the present invention. DETAILED DESCRIPTION

[0051] The technical solution of the present invention will be clearly and completely described below in conjunction with the accompanying drawings in the present invention. In addition, the forms of the various structures recorded in the following implementation modes are merely illustrative. The online car-hailing management method and system for bus hub parking lots based on SaaS involved in the present invention are not limited to the various structures recorded in the following implementation modes. All other implementation modes obtained by ordinary technicians in this field without making creative work belong to the scope of protection of the present invention.

[0052] The present invention provides a SaaS-based online car-hailing management method for a bus hub parking lot, comprising the following steps:

[0053] S001: The online car-hailing vehicle receives a passenger order and arrives at the bus hub parking lot. If the passenger cancels the order before the departure time, the driver's willingness is collected through the platform. If the driver chooses to leave voluntarily, no data collection and analysis is performed, and the driver is prompted to leave the bus hub parking lot according to the route. If the driver chooses to continue to accept orders in the bus hub parking lot, the online car-hailing vehicle information is collected;

[0054] S002: Analyze and obtain a re-order priority index based on the collected online car-hailing information, and compare the re-order priority index with a preset threshold. If the re-order priority index is less than the preset threshold, it is determined that the cancellation of the order has little impact on the driver, and the driver is instructed to leave the bus hub parking lot and re-accept the order outside the bus hub parking lot. If the re-order priority index is greater than the preset threshold, it is determined that the cancellation of the order has a greater impact on the driver, and the driver is allowed to stay in the bus hub parking lot, and parking lot information is collected;

[0055] S003: Assess and determine the demand for taxis based on parking lot information, and allocate online-hailing taxis that are allowed to stay in bus hub parking lots to areas with the highest demand for taxis to continue accepting orders.

[0056] Bus hub parking lots are usually key nodes in urban transportation. Managing online ride-hailing vehicles can maintain traffic order in parking lots and avoid chaos and congestion caused by a large number of online ride-hailing vehicles waiting or parked. By managing the number, service level and parking locations of online ride-hailing vehicles, the quality and reliability of online ride-hailing services can be improved, improving passengers' travel experience while ensuring the safety of passengers and drivers. By managing the number and distribution of online ride-hailing vehicles, parking lot resources can be better utilized, the utilization rate of parking spaces can be improved, and resource waste and space occupation problems can be avoided.

[0057] Bus hub parking lots are usually located near urban transportation hubs. Online ride-hailing can provide convenient transportation connection services and provide passengers with a convenient travel experience. Through the reasonable management of online ride-hailing, the transportation connection efficiency of bus hubs can be improved and the connectivity of urban transportation networks can be enhanced. By managing the number, parking locations, and service quality of online ride-hailing, the quality and reliability of online ride-hailing services can be improved, and the travel satisfaction of passengers can be improved. It is helpful to improve the service image of bus hubs and attract more passengers to choose to use bus hubs for transfers and travel. Effective management of the online ride-hailing industry can improve the operational efficiency and market competitiveness of the online ride-hailing industry and promote the healthy development of the online ride-hailing industry. At the same time, the provision of online ride-hailing services has also brought more business opportunities and passenger flow to the commercial and service industries around bus hubs, which helps promote economic development.

[0058] In this embodiment, it should be specifically explained that Figure 2 As shown, the steps for calculating the re-order priority index are:

[0059] The pick-up distance of the cancelled order is obtained through the GPS positioning device, and the travel time of the cancelled order is counted. The cost coefficient of the online car-hailing service is calculated based on the pick-up distance and the travel time. The calculation formula is CM=SJ×JT, where CM represents the cost coefficient of the online car-hailing service, SJ represents the pick-up distance, and JT represents the travel time.

[0060] The time accident coefficient is calculated by obtaining the interval between the time when the passenger cancels the order and the time when the trip departs through the online car-hailing platform. The calculation formula is: Among them, AN represents the time accident coefficient, T2 and T1 represent the departure time of the trip and the time when the passenger cancels the order;

[0061] The frequency of cancellation of the order by passengers during the time period is obtained through the online car-hailing platform, which is recorded as the breach of contract coefficient. The calculation formula is: Where CF is the default coefficient, NUM C It is represented by the number of times the online car-hailing service is cancelled by passengers within a time period, and T′ is represented by a time period, which can be one month or half a year;

[0062] The order-taking priority index is calculated based on the online car-hailing cost coefficient, time accident coefficient and breach of contract coefficient. The calculation formula is TA=a1×CM+a2×AN+a3×CF, where TA represents the order-taking priority index and CM represents the online car-hailing cost coefficient. If the online car-hailing cost coefficient increases, it means that the cost or time consumption of the online car-hailing service increases, and drivers may be more willing to accept orders again because they may hope to make up for the increase in costs by accepting more orders. On the contrary, if the cost coefficient decreases, drivers may be more cautious about accepting orders again because they may be more willing to wait for high-value orders. Therefore, the higher the online car-hailing cost coefficient, the higher the order-taking priority index. AN represents the time accident coefficient. If the time accident coefficient increases, it means that the passenger is more likely to have a higher risk of being late. If an order is cancelled shortly before the start of the journey, the online car-hailing vehicle may have to wait for a long time. The cancelled order will cause the driver to lose time and income. Therefore, the higher the time accident coefficient, the higher the priority index for accepting the order again. CF is represented by the breach of contract coefficient. An increase in the breach of contract coefficient means that the driver encounters more passengers canceling their orders within a certain period of time, and the orders accepted by the driver are more unstable, resulting in increased losses for the driver. In this case, the driver may be more inclined to accept the order again to make up for the idle time and losses caused by the cancellation of the order. Therefore, the higher the breach of contract coefficient, the higher the priority index for accepting the order again. a1, a2, and a3 are represented by the weight coefficients of the online car-hailing vehicle consumption coefficient, the time accident coefficient, and the breach of contract coefficient, and this embodiment does not specifically calculate the specific values ​​of a1, a2, and a3.

[0063] The impact of order cancellations on drivers can be used as one of the important indicators for evaluating the benefits of drivers staying in the parking lot to pick up passengers. Specifically, if order cancellations have little impact on drivers, that is, canceling orders does not cause too much loss or inconvenience to drivers, canceling orders may affect drivers' work efficiency, especially if frequent order cancellations lead to a decrease in the driver's order completion rate, which may affect the driver's work enthusiasm and income level. Frequent order cancellations may have a negative impact on drivers' psychology, reducing their enthusiasm and involvement in order-taking tasks.

[0064] In this embodiment, it should be specifically explained that the steps of obtaining the pick-up distance of the cancelled order of the online car-hailing vehicle through the GPS positioning device are:

[0065] The latitude and longitude of the starting position of the cancelled order and the current location of the bus hub parking lot are collected through the GPS positioning device installed on the online car-hailing vehicle, and recorded as the starting position and the end position. The GPS positioning device is a device for obtaining global positioning system signals and calculating location information. The GPS positioning device usually includes a GPS receiver and related hardware components, which are used to receive positioning signals from satellites and calculate the precise location coordinates of the device;

[0066] Convert the latitude and longitude coordinates of the starting and ending positions into radians;

[0067] Use the Haversine formula to calculate the spherical distance, and the calculation formula is: Wherein, SJ represents the pick-up distance, Δlat represents the radian expression of the latitude difference between the starting position and the end position, Δlon represents the radian expression of the longitude difference between the starting position and the end position, lat1 and lat2 represent the latitudes of the starting position and the end position in radians, respectively, and R represents the radius of the earth, and the average radius of the earth is usually taken as about 6371 kilometers. The Haversine formula is a mathematical formula for calculating the distance between two points on two spheres, and is usually used to calculate the distance between two longitude and latitude coordinate points on the earth.

[0068] The steps of converting the longitude and latitude coordinates of the starting position and the end position into radians are as follows: the range of longitude is from 180 degrees west longitude to 180 degrees east longitude, and the range of latitude is from 90 degrees south latitude to 90 degrees north latitude, so the conversion formula of longitude and latitude is

[0069] Allocating canceled ride-hailing vehicles to areas with high demand can maximize the use of ride-hailing vehicle resources, meet the travel needs of passengers in high-demand areas, and improve resource utilization efficiency. Concentrating ride-hailing vehicles in areas with high demand can shorten the time passengers spend waiting for vehicles, improve service response speed, and enhance passenger satisfaction with ride-hailing services. Allocating ride-hailing vehicles to areas with high demand means that drivers have more orders to take, thereby increasing drivers' income levels and improving their work enthusiasm and satisfaction.

[0070] In this embodiment, it should be specifically explained that the steps of evaluating and judging the taxi demand based on the parking lot information are as follows:

[0071] The parking lot is divided into n parts, recorded as sub-parking lot areas, and the data of passenger boarding and alighting points in each sub-parking lot area are collected. The data includes the density of boarding and alighting points, the density of alighting points, and the correlation of boarding and alighting points. The taxi hotspot is calculated based on the data of passenger boarding and alighting points. The calculation formula is: PT is the taxi hotspot, BD is the pick-up point density, which is the ratio of the number of multiple pick-up points in the sub-parking lot area to the area of ​​the area. The number or ratio of pick-up points in the area reflects the activeness of passengers waiting for taxis in the area. There are more passenger pick-up points in the hotspot area, which means that the demand for passengers waiting for taxis is large. OD is the drop-off point density, which is the ratio of the number of multiple drop-off points in the sub-parking lot area to the area of ​​the area. The number or ratio of drop-off points in the area reflects the distribution of passenger destinations in the area. There are more passenger drop-off points in the hotspot area, which means that the area is a popular destination and there may be more passengers who need to take a taxi to leave. RV is the pick-up and drop-off point correlation, which considers the spatial distribution correlation of passenger pick-up and drop-off points, which can help identify possible hotspot areas.

[0072] Collect passenger flow data, analyze the passenger flow data, and calculate the passenger flow coefficient;

[0073] The calculated passenger flow coefficient is used as a clustering feature to cluster the passenger flow coefficient of each sub-parking lot area, and the passenger flow distribution coefficient is calculated and evaluated based on the clustering results;

[0074] The taxi demand index is calculated based on the parking space occupancy rate, taxi hotspot and passenger flow distribution coefficient, and the calculation formula is TX=b1×PT+b2×PD, where TX represents the taxi demand index, PT represents the taxi hotspot. If the taxi hotspot of an area is high, it means that there are more passengers who need taxis in the area, and the taxi demand index of the area will also be high, because more passengers need to use online car-hailing services. PD represents the passenger flow distribution coefficient. If the passenger flow distribution coefficient of an area is high, it means that the passenger flow in the area is relatively concentrated, and it may be an area with dense traffic. In this case, the taxi demand index of the area is likely to be high, because an increase in passenger flow is usually accompanied by an increase in demand for transportation services. b1 and b2 represent the weight coefficients of the taxi hotspot and the passenger flow distribution coefficient, and this embodiment does not perform specific calculations on the specific values ​​of b1 and b2.

[0075] In this embodiment, it should be specifically explained that the steps of calculating the association between boarding and alighting points are as follows:

[0076] Collect the number of pick-up and drop-off points in all parking lot areas and calculate the average number of pick-up and drop-off points;

[0077] The Pearson correlation coefficient formula is used to calculate the correlation between boarding and alighting points. The calculation formula is: Where XG represents the association between boarding and alighting points, X represents the number of boarding points, and Y represents the number of alighting points. Expressed as the average boarding points, It is expressed as the average number of alighting points, and the value range of XG is between -1 and 1. If the XG value is 1, it means that the boarding point and the alighting point increase or decrease completely synchronously. If the XG value is -1, it means that the boarding point and the alighting point change completely in the opposite direction. If the XG value is 0, it means that there is no linear relationship between the boarding point and the alighting point. The Pearson correlation coefficient is a statistic used to measure the strength and direction of the linear relationship between two variables, and measures the degree of linear correlation between the two variables.

[0078] In this embodiment, it should be specifically explained that the steps for calculating the passenger flow coefficient are:

[0079] Count the number of vehicles entering and leaving each sub-parking lot area, that is, the number of times vehicles enter and leave the area during the detection period. The calculation formula is N total =N in +N out , where N total Represents the number of vehicles entering and exiting each sub-parking lot area, N in It is represented by the number of vehicles entering the sub-parking lot area during the detection period, N out It is represented by the number of vehicles that leave the sub-parking lot area during the detection period;

[0080] The average residence time is calculated by counting the total number of vehicles in the area and the residence time of the vehicles in the area during the detection period. The calculation formula is: Where T avg Expressed as the average residence time, T total It is expressed as the sum of the residence time of all vehicles in the area, NUM stay It is expressed as the total number of vehicles, where the total number of vehicles includes the number of vehicles that have left and the number of vehicles that have not left;

[0081] The traffic density of each sub-parking lot area is calculated based on the number of vehicles entering and exiting each sub-parking lot area and the detection time period. The calculation formula is: Density is the traffic density, N total It is represented by the number of vehicles entering and exiting each sub-parking lot area, and ΔT is represented by the detection time period;

[0082] The passenger flow coefficient of each sub-parking lot area is obtained by evaluating the number of vehicles entering and exiting each sub-parking lot area, the average stay time, and the traffic density. The calculation formula is: Where PF is the passenger flow coefficient, N total Represented as the number of vehicles entering and exiting each sub-parking lot area, T avg It is expressed as the average stay time, Density is the traffic density, NUM park Expressed as the number of parking spaces in a sub-parking lot area.

[0083] In this embodiment, it should be specifically explained that the steps of clustering the passenger flow coefficient of each sub-parking lot area are:

[0084] The passenger flow coefficient is used as a clustering feature, and the passenger flow coefficient of each sub-parking lot area is used as a data point to obtain a set of data sets. The elbow method is used to determine the optimal number of clusters. The elbow method is a method based on the sum of squared clustering errors. By plotting a curve graph of the number of clusters and the corresponding SSE value, the "elbow" point in the graph is found, that is, the point where the curve begins to flatten. This point is usually considered to be the optimal number of clusters. The idea of ​​the elbow method is that as the number of clusters increases, the SSE value will gradually decrease, but when the optimal number of clusters is reached, the rate of decrease of the SSE value will slow down, forming an "elbow". The horizontal axis in the curve graph is the number of clusters (K), and the vertical axis is the corresponding SSE value. When the number of clusters is small, the SSE value will be higher because the distance from each data point to the centroid of the cluster to which it belongs will be larger. As the number of clusters increases, the SSE value will gradually decrease because more clusters will make each data point closer to the centroid of the cluster to which it belongs. However, as the number of clusters continues to increase, the rate of decrease of the SSE value will gradually slow down. At the elbow point, the curve has an obvious inflection point, which means that adding more clusters will not significantly reduce the SSE value. Therefore, the number of clusters corresponding to the elbow point is usually selected as the optimal number of clusters.

[0085] Randomly select a corresponding number of data points as the initial cluster centers, and use the Euclidean distance to calculate the distance between each data point in the data set and each initial cluster center. The calculation formula is: Among them, x ik Represented as x i The value of the k-th dimension feature, c jk Represents the initial cluster center c j The value of the k-th dimension feature, n represents the number of data points in the data set, and the data points are assigned to the cluster with the nearest cluster center. In clustering, Euclidean distance is a commonly used distance measurement method to measure the similarity or difference between two samples. The initial cluster center is a set of initial points randomly selected when starting the clustering process. These points are used as the centers of the clusters;

[0086] Traverse all the data points in the data set to obtain the initial clusters. For each initial cluster, calculate the average value of all data points in the cluster to obtain the new cluster center. The calculation formula is: Where V j Represents the new cluster center of the jth initial cluster, x i Represents cluster C j The data points in C j Represented as the initial clustering cluster, |C j |Represented as cluster C jThe number of data points, the initial cluster refers to the initial state in the clustering process, usually all data points are regarded as a separate cluster, and then iterative optimization is performed according to the rules and methods of the clustering algorithm to gradually divide the data points into appropriate clusters;

[0087] Repeat the above clustering steps until the cluster center no longer changes, and obtain the final cluster and the final cluster center.

[0088] In this embodiment, it should be specifically explained that the steps of evaluating and calculating the passenger flow distribution coefficient according to the clustering result are as follows:

[0089] According to the total number of data points and the number of data points in each final cluster, the weight of each final cluster is calculated, and the calculation formula is: Among them WG j Indicates the weight of the jth final cluster, NUM j represents the number of data points in the jth final cluster, and n represents the number of data points in the data set;

[0090] The final cluster center of each final cluster is collected, and the passenger flow distribution coefficient is calculated based on the weight of the final cluster center and the final cluster. The calculation formula is PD j =WG j ×ZX j , where PD j It is expressed as the passenger flow distribution coefficient in the jth final cluster, then the passenger flow distribution coefficient is PD, WG j represents the weight of the jth final cluster, ZX j Denotes the final cluster center of the jth final cluster.

[0091] The passenger flow distribution takes into account the weight distribution of clusters and can be used to better understand the clustering results because it reflects the importance and influence of different clusters in the entire data set.

[0092] In this embodiment, it is necessary to specifically explain that the online car-hailing management system for the bus hub parking lot based on SaaS includes:

[0093] The driver intention collection module is used to collect the intentions of online car-hailing drivers whose orders have been cancelled, including whether to leave the parking lot or stay in the parking lot to continue accepting orders;

[0094] The online car-hailing data collection module is used to collect relevant data information of online car-hailing vehicles that choose to stay in the parking lot and continue to receive orders;

[0095] The online car-hailing status judgment module is used to conduct a comprehensive evaluation based on the relevant data information of the online car-hailing vehicle collected by the online car-hailing data collection module, and determine whether to keep or remove the online car-hailing vehicle based on the evaluation results;

[0096] The parking lot status evaluation module is used to collect the current data information of the parking lot and evaluate the parking lot status based on the current data information of the parking lot;

[0097] The online car-hailing allocation module is used to allocate the online car-hailing vehicles that are determined to be left in the parking lot to the area with the greatest demand for taxis based on the parking lot status.

[0098] Finally: The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

[0099] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art who is familiar with the present technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

Claims

1. A SaaS-based online car-hailing management method for a bus hub parking lot, characterized in that: The following steps are involved: S001: The online car-hailing vehicle receives a passenger order and arrives at the bus hub parking lot. If the passenger cancels the order before the departure time, the driver's willingness is collected through the platform. If the driver chooses to leave voluntarily, no data collection and analysis is performed, and the driver is prompted to leave the bus hub parking lot according to the route. If the driver chooses to continue to accept orders in the bus hub parking lot, the online car-hailing vehicle information is collected; S002: Analyze and obtain a re-order priority index based on the collected online car-hailing information, and compare the re-order priority index with a preset threshold. If the re-order priority index is less than the preset threshold, it is determined that the cancellation of the order has little impact on the driver, and the driver is instructed to leave the bus hub parking lot and re-accept the order outside the bus hub parking lot. If the re-order priority index is greater than the preset threshold, it is determined that the cancellation of the order has a greater impact on the driver, and the driver is allowed to stay in the bus hub parking lot, and parking lot information is collected; S003: Evaluate and judge the demand for taxis based on parking lot information, and allocate online-hailing taxis that are allowed to stay in the bus hub parking lot to the area with the highest demand for taxis to continue to receive orders; The steps of evaluating and judging the taxi demand based on parking lot information are as follows: The parking lot is divided into n parts, recorded as sub-parking lot areas, and the data of passenger boarding and alighting points in each sub-parking lot area are collected. The data includes the density of boarding and alighting points, the density of alighting points, and the correlation of boarding and alighting points. The taxi hotspot is calculated based on the data of passenger boarding and alighting points. The calculation formula is: ,in It is represented as the taxi hotspot degree. It is expressed as the pick-up point density, which is the ratio of the number of re-boarding points in the sub-parking lot area to the area of ​​the area. It is expressed as the drop-off point density, which is the ratio of the number of drop-off points in the sub-parking lot area to the area of ​​the area. It is expressed as the association between boarding and alighting points; Collect passenger flow data, analyze the passenger flow data, and calculate the passenger flow coefficient; The calculated passenger flow coefficient is used as a clustering feature to cluster the passenger flow coefficient of each sub-parking lot area, and the passenger flow distribution coefficient is calculated and evaluated based on the clustering results; The taxi demand index is calculated based on the parking space occupancy rate, taxi hotspot and passenger flow distribution coefficient, and its calculation formula is: ,in It is expressed as the taxi demand index, It is represented as the taxi hotspot degree. It is represented as the passenger flow distribution coefficient, b1 and b2 are the weight coefficients of taxi hotspot and passenger flow distribution coefficient; The steps for calculating the association between boarding and alighting points are as follows: Collect the number of pick-up and drop-off points in all parking lot areas and calculate the average number of pick-up and drop-off points; The Pearson correlation coefficient formula is used to calculate the correlation between boarding and alighting points. The calculation formula is: ,in It is expressed as the association between boarding and alighting points. Expressed as boarding points, It is expressed as the number of drop-off points. Expressed as the average boarding points, It is expressed as the average number of drop-off points; The steps for calculating the passenger flow coefficient are as follows: The number of vehicles entering and leaving each sub-parking lot area is counted, that is, the number of times vehicles enter and leave the area during the detection period. The calculation formula is: ,in It is represented by the number of vehicles entering and exiting each sub-parking lot area. It is represented by the number of vehicles entering the sub-parking lot area during the detection period. It is represented by the number of vehicles leaving the sub-parking lot area during the detection period; Counting the total number of vehicles in the area and the time the vehicles stay in the area during the detection time period, and calculating the ratio of the total number of vehicles to the time the vehicles stay in the area to obtain an average stay time, wherein the total number of vehicles includes the number of vehicles that have left and the number of vehicles that have not left; The vehicle flow density of each sub-parking lot area is calculated by comparing the number of vehicles entering and exiting each sub-parking lot area with the detection time period; The passenger flow coefficient of each sub-parking lot area is obtained by evaluating the number of vehicles entering and exiting each sub-parking lot area, the average stay time, and the traffic density. The calculation formula is: ,in Expressed as the passenger flow coefficient, It is represented by the number of vehicles entering and exiting each sub-parking lot area. is the average residence time, Expressed as traffic density, The number of parking spaces expressed as a sub-parking lot area; The steps of clustering the passenger flow coefficient of each sub-parking lot area are as follows: Step 1: Take the passenger flow coefficient as the clustering feature and the passenger flow coefficient of each sub-parking lot area as the data point to obtain a set of data sets, and use the elbow method to determine the optimal number of clusters; Step 2: Randomly select a certain number of data points as the initial cluster centers, and use the Euclidean distance to calculate the distance between each data point in the data set and each initial cluster center. The calculation formula is: ,in, Expressed as In the The value of the dimension feature, Represents the initial cluster center In the The value of the dimension feature, It is represented as the number of data points in the data set, and the data points are assigned to the cluster with the nearest cluster center; Step 3: Traverse all the data points in the data set to obtain the initial clusters. For each initial cluster, calculate the average value of all data points in the cluster to obtain the new cluster center. The calculation formula is: ,in Expressed as The new cluster centers of the initial clusters, Representation Cluster The data points in Represents the initial clustering cluster, Represented as a cluster The number of data points; Repeat steps 2 and 3 until the cluster center does not change, and obtain the final cluster and the final cluster center; The steps of evaluating and calculating the passenger flow distribution coefficient based on the clustering results are as follows: According to the total number of data points and the number of data points in each final cluster, the weight of each final cluster is calculated, and the calculation formula is: ,in Indicates The weights of the final clusters, Indicates The number of data points in the final clusters, Expressed as the number of data points in the dataset; The final cluster center of each final cluster is collected, and the passenger flow distribution coefficient is calculated based on the weight of the final cluster center and the final cluster. The calculation formula is: ,in Expressed as The passenger flow distribution coefficient in the final cluster is , Indicates The weights of the final clusters, Expressed as The final cluster centers of the final clusters.

2. The SaaS-based online car-hailing management method for a bus hub parking lot according to claim 1 is characterized by: The steps for calculating the re-order priority index are as follows: The pick-up distance of the cancelled order is obtained through the GPS positioning device, and the travel time of the cancelled order is calculated, and the pick-up distance and the travel time are multiplied to obtain the cost coefficient of the cancelled order; The time accident coefficient is calculated by obtaining the interval between the time when the passenger cancels the order and the time when the trip departs through the online car-hailing platform. The calculation formula is: ,in Expressed as the time contingency coefficient, and It is expressed as the departure time of the trip and the time when the passenger cancels the order; Obtain the frequency of cancellations of online ride-hailing orders by passengers during a period of time through the online ride-hailing platform, and record it as the breach of contract coefficient; The priority index for accepting an order again is calculated based on the online car-hailing cost coefficient, time accident coefficient, and breach of contract coefficient. The calculation formula is: ,in It is expressed as the priority index for accepting orders again. Expressed as the cost coefficient of online car-hailing, Expressed as the time contingency coefficient, It is represented as the breach coefficient, and a1, a2, and a3 are the weight coefficients of the online car-hailing cost coefficient, time accident coefficient, and breach coefficient.

3. The SaaS-based online car-hailing management method for a bus hub parking lot according to claim 2 is characterized by: The steps of obtaining the pick-up distance of the cancelled order of the online car-hailing service through the GPS positioning device are as follows: The GPS positioning device installed on the online car-hailing vehicle is used to collect the latitude and longitude of the starting location of the cancelled order and the current location of the bus hub parking lot, and these are recorded as the starting location and the ending location; Convert the latitude and longitude coordinates of the starting and ending positions into radians; Use the Haversine formula to calculate the spherical distance, and the calculation formula is: ,in Expressed as the pick-up distance, It is expressed as the difference in latitude between the starting position and the ending position in radians. It is expressed as the difference in longitude between the starting position and the end position in radians. and They are respectively expressed as the latitude of the starting position and the end position in arcs, Expressed as the radius of the Earth.

4. A system for implementing the SaaS-based online car-hailing management method for a bus hub parking lot according to claim 1, characterized in that: The system comprises: The driver intention collection module is used to collect the intentions of online car-hailing drivers whose orders have been cancelled, including whether to leave the parking lot or stay in the parking lot to continue accepting orders; The online car-hailing data collection module is used to collect relevant data information of online car-hailing vehicles that choose to stay in the parking lot and continue to receive orders; The online car-hailing status judgment module is used to conduct a comprehensive evaluation based on the relevant data information of the online car-hailing vehicle collected by the online car-hailing data collection module, and determine whether to keep or remove the online car-hailing vehicle based on the evaluation results; The parking lot status evaluation module is used to collect the current data information of the parking lot and evaluate the parking lot status based on the current data information of the parking lot; The online car-hailing allocation module is used to allocate the online car-hailing vehicles that are determined to be left in the parking lot to the area with the greatest demand for taxis based on the parking lot status.

Images