An ETA method based on driver's driving proficiency factor

Abstract

The present invention relates to the technical field of online car-hailing, and specifically discloses an ETA method based on a driver's driving proficiency factor, comprising the following steps: S1, calculating the average speed of the driver's driving route on a daily basis; S2, removing abnormal values ​​of the daily data of a city; S3, calculating the average speed of the route according to the starting point, the end point and the departure time; S4, calculating the average speed of the driver on the route; S5, calculating the route proficiency of a single driver and the global route proficiency; S6, calculating the grid proficiency of a single driver; S7, calculating the driver proficiency according to the starting point and the end point; S8, calculating the ETA according to the driver's driving proficiency; the present invention introduces a driver's driving proficiency factor to make the estimated time of the online car-hailing route more accurate, the method calculates the driver's route proficiency and the global proficiency, and by correcting the average time of the existing third-party map platform for driving proficiency, compared with the existing technical methods, the accuracy of the estimated time can be effectively improved.

Description

Technical Field

[0001] The present invention relates to the technical field of online car-hailing, and in particular to an ETA method based on a driver's driving proficiency factor. Background Art

[0002] Currently, in the field of online travel, there are two scenarios that require the use of the ETA method, one is to estimate the time for passengers to be picked up, and the other is to estimate the time it takes for passengers to arrive at the destination from the starting point. Estimating the time for passengers to be picked up will give passengers an accurate estimate of the time the driver will arrive, so that they can arrange things before departure reasonably. Estimating the time it takes for passengers to arrive at the destination from the starting point will provide guidance for passengers to plan things after arriving at the destination. With the development trend of technological intelligence, higher requirements are also put forward for the accuracy of time estimation.

[0003] Currently, online ride-hailing platforms generally use the ETA time provided by third-party map platforms as the ETA time for online ride-hailing. The ETA time provided by third-party map platforms is an average time, taking into account factors such as current road conditions, weather, and traffic accidents. This is a universal time that does not take into account the driver's driving factors. For specific drivers, driving skills, driving habits, and road familiarity will all become factors affecting the ETA of a specific route.

[0004] Based on this, the present invention discloses an ETA method based on the driver's driving proficiency factor to solve the problem of more accurately estimating the driver's ETA time for a certain route, thereby better guiding passengers to arrange their time reasonably. Summary of the invention

[0005] The object of the present invention is to provide an ETA method based on a driver's driving proficiency factor to solve the problems raised in the above background technology.

[0006] To achieve the above object, the present invention provides the following technical solution: an ETA method based on a driver's driving proficiency factor, comprising the following steps:

[0007] S1. Calculate the average speed of the driver's driving route on a daily basis;

[0008] Wherein, step S1 comprises the following steps:

[0009] S11, after the city is divided into grids of uniform size according to longitude and latitude, the starting point a and the end point b of the city's geographical location information route are encoded;

[0010] S12, code the departure time segment t, 24 hours is divided into 10-minute segments, and the codes are 0-143 respectively;

[0011] S13. Driving mileage is coded in kilometers, and driving time is coded in minutes:

[0012] For data within one year, with the same starting point code a, end point code b, departure time code t and date d, the total driving time on date d is st d,a,b,t , the total driving mileage on date d is s1 d,a,b,t , the average speed is v d,a,b,t , where v d,a,b,t =s1 d,a,b,t / st d,a,b,t ;

[0013] S2, remove outliers in the city daily data;

[0014] Reduce interference. Due to traffic accidents and collection errors, some data should be unreliable. Remove low-reliability data. For cities with the same starting point code, end point code, and departure time code, sort the average speed of data within one year.

[0015] Among them, Q1 a,b,t The quarter quantile of the average speed, Q3 a,b,t The IQR is the quartile of the ranking, calculated by the formula: a,b,t =Q3 a,b,t -Q1 a,b,t ;

[0016] Among them, if v on date d d,a,b,t <Q1 a,b,t -1.5IQR a,b,t or v d,a,b,t >Q3 a,b,t +1.5IQR a,b,t , then remove the corresponding data under the date d;

[0017] S3, calculate the average speed of the route according to the starting point, end point and departure time;

[0018] After removing the abnormal data in step S2, the total driving time is calculated as st according to step S1. a,b,t =∑st d,a,b,t , the total driving mileage is s1 a,b,t =∑s1 d,a,b,t , the average speed is v a,b,t =s1 a,b,t / st a,b,t ;

[0019] S4, calculating the average speed of the driver on the route;

[0020] Among them, all drivers are coded and recorded as q. According to the historical driving data of each driver, after filtering invalid data (a, b, t, d) according to step S1, the total driving time of the driver is calculated as st a,b,t,q =∑std,a,b,t,q , the total driving mileage is s1 a,b,t,q =∑s1 d,a,b,t,q , the average speed is v a,b,t,q =s1 a,b,t,q / st a,b,t,q ;

[0021] S5, calculating the route proficiency of a single driver and the global route proficiency;

[0022] Among them, the driver's route proficiency is f(q,a,b)=∑((v a,b,t,q / v a,b,t )) / countf q,a,b , where countf q,a,b The number of records of the starting point a and the end point b of driver q in step S4 is accumulated for these records of the driver;

[0023] The driver's global proficiency is g(q) = ∑f(q,a,b) / countg q , where countg q The total number of routes taken by the driver;

[0024] S6, calculating the grid proficiency of a single driver;

[0025] Here, e is defined as any grid number, and h(q,e) is the driver’s proficiency in a certain grid;

[0026] Select q, a, b in step S4 and calculate f(q, a, b):

[0027] Where h(q,e)=∑f(q,a,b) / counth q,a,b ,

[0028] counth q,a,b For all historical routes of q, add up the number of times they pass through grid e and calculate h(q,e);

[0029] S7. Calculate the driver's proficiency based on the starting point and the end point;

[0030] Wherein, step S7 comprises the following steps:

[0031] S71. If f(q,a,b) exists, it means the driver has walked the route, then r(q,a,b)=f(q,a,b);

[0032] S72. If the driver has not taken this route, find all the grids e along the route from a to b, and their set is S a,b ,r(q,a,b)=∑h(q,e) / countr a,b , where countr a,b For Sa,b Medium grid;

[0033] S8. Calculate ETA based on the driver's driving proficiency;

[0034] Among them, ETA(a,b) is the pick-up time returned by the third-party map platform. After calculation, META(q,a,b)=ETA(a,b) / r(q,a,b).

[0035] Preferably, in step S6, if the driver has never been to the grid e, then h(q,e)=g(q).

[0036] Compared with the prior art, the beneficial effects of the present invention are: the present invention introduces the driver's driving proficiency factor to make the estimated time of the online car-hailing route more accurate. The method calculates the driver's route proficiency and global proficiency, and corrects the average time of the existing third-party map platform according to the driving proficiency. Compared with the existing technical methods, it can effectively improve the accuracy of the estimated time, thereby bringing greater value to the passenger experience; it solves the problem of large deviations in the existing online car-hailing order estimates and the estimates of the time from the passenger getting on the car to the destination due to failure to take into account the actual driver's driving factors; improving the accuracy of time estimates, while bringing a better experience to passengers, can also provide value to the order dispatching strategy of the online car-hailing platform. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] Figure 1 It is a structural flowchart of the overall steps of the present invention. DETAILED DESCRIPTION

[0038] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0039] See also Figure 1 The present invention provides a technical solution: an ETA method based on a driver's driving proficiency factor, comprising the following steps:

[0040] S1. Calculate the average speed of the driver's driving route on a daily basis;

[0041] Wherein, step S1 comprises the following steps:

[0042] S11, after the city is divided into grids of uniform size according to longitude and latitude, the starting point a and the end point b of the city's geographical location information route are encoded;

[0043] S12, code the departure time segment t, 24 hours is divided into 10-minute segments, and the codes are 0-143 respectively;

[0044] S13. Driving mileage is coded in kilometers, and driving time is coded in minutes:

[0045] For data within one year, with the same starting point code a, end point code b, departure time code t and date d, the total driving time on date d is st d,a,b,t , the total driving mileage on date d is s1 d,a,b,t , the average speed is v d,a,b,t , where v d,a,b,t =s1 d,a,b,t / st d,a,b,t ;

[0046] S2, remove outliers in the city daily data;

[0047] Reduce interference. Due to traffic accidents and collection errors, some data should be unreliable. Remove low-reliability data. For cities with the same starting point code, end point code, and departure time code, sort the average speed of data within one year.

[0048] Among them, Q1 a,b,t The quarter quantile of the average speed, Q3 a,b,t The IQR is the quartile of the ranking, calculated by the formula: a,b,t =Q3 a,b,t -Q1 a,b,t ;

[0049] Among them, if v on date d d,a,b,t <Q1 a,b,t -1.5IQR a,b,t or v d,a,b,t >Q3 a,b,t +1.5IQR a,b,t , then remove the corresponding data under the date d;

[0050] S3, calculate the average speed of the route according to the starting point, end point and departure time;

[0051] After removing the abnormal data in step S2, the total driving time is calculated as st according to step S1. a,b,t =∑st d,a,b,t , the total driving mileage is s1 a,b,t =∑s1 d,a,b,t , the average speed is v a,b,t =s1 a,b,t / st a,b,t ;

[0052] S4, calculating the average speed of the driver on the route;

[0053] Among them, all drivers are coded and recorded as q. According to the historical driving data of each driver, after filtering invalid data (a, b, t, d) according to step S1, the total driving time of the driver is calculated as st a,b,t,q =∑st d,a,b,t,q , the total driving mileage is s1 a,b,t,q =∑s1 d,a,b,t,q , the average speed is v a,b,t,q =s1 a,b,t,q / st a,b,t,q ;

[0054] S5, calculating the route proficiency of a single driver and the global route proficiency;

[0055] Among them, the driver's route proficiency is f(q,a,b)=∑((v a,b,t,q / v a,b,t )) / countf q,a,b , where countf q,a,b The number of records of the starting point a and the end point b of driver q in step S4 is accumulated for these records of the driver;

[0056] The driver's global proficiency is g(q) = ∑f(q,a,b) / countg q , where countg q The total number of routes taken by the driver;

[0057] S6, calculating the grid proficiency of a single driver;

[0058] Here, e is defined as any grid number, and h(q,e) is the driver’s proficiency in a certain grid;

[0059] Select q, a, b in step S4 and calculate f(q, a, b):

[0060] Where h(q,e)=∑f(q,a,b) / counth q,a,b ,

[0061] counth q,a,b For all historical routes of q, add up the number of times they pass through grid e and calculate h(q,e);

[0062] S7. Calculate the driver's proficiency based on the starting point and the end point;

[0063] Wherein, step S7 comprises the following steps:

[0064] S71. If f(q,a,b) exists, it means the driver has walked the route, then r(q,a,b)=f(q,a,b);

[0065] S72. If the driver has not taken this route, find all the grids e along the route from a to b, and their set is S a,b ,r(q,a,b)=∑h(q,e) / countr a,b , where countr a,b For S a,b Medium grid;

[0066] S8. Calculate ETA based on the driver's driving proficiency;

[0067] Among them, ETA(a,b) is the pick-up time returned by the third-party map platform. After calculation, META(q,a,b)=ETA(a,b) / r(q,a,b).

[0068] Furthermore, in step S6, if the driver has never been to the grid e, then h(q,e)=g(q).

[0069] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. An ETA method based on a driver's driving proficiency factor, characterized in that: The following steps are involved: S1. Calculate the average speed of the driver's driving route on a daily basis; Wherein, step S1 comprises the following steps: S11, after the city is divided into grids of uniform size according to longitude and latitude, the starting point a and the end point b of the city's geographical location information route are encoded; S12, code the departure time segment t, 24 hours is divided into 10-minute segments, and the codes are 0-143 respectively; S13. Driving mileage is coded in kilometers, and driving time is coded in minutes: For data within one year, with the same starting point code a, end point code b, departure time code t and date d, the total driving time on date d is st d,a,b,t , the total driving mileage on date d is s1 d,a,b,t , the average speed is v d,a,b,t , where v d,a,b,t =s1 d,a,b,t / st d,a,b,t ; S2. Remove outliers from the daily data of cities to reduce interference. Due to traffic accidents and collection errors, some data should be unreliable. Remove data with low credibility. For cities with the same starting point code, end point code, and departure time code, sort the average speed of data within 1 year. Among them, Q1 a,b,t The quarter quantile of the average speed, Q3 a,b,t The IQR is the quartile of the ranking, calculated by the formula: a,b,t =Q3 a,b,t -Q1 a,b,t ; Among them, if v on date d d,a,b,t <Q1 a,b,t -1.5IQR a,b,t or v d,a,b,t >Q3 a,b,t +1.5IQR a,b,t , then remove the corresponding data under the date d; S3, calculate the average speed of the route according to the starting point, end point and departure time; After removing the abnormal data in step S2, the total driving time is calculated as st according to step S1. a,b,t =∑st d,a,b,t , the total driving mileage is s1 a,b,t =∑s1 d,a,b,t , the average speed is v a,b,t =s1 a,b,t / st a,b,t ; S4, calculating the average speed of the driver on the route; Among them, all drivers are coded and recorded as q. According to the historical driving data of each driver, after filtering invalid data (a, b, t, d) according to step S1, the total driving time of the driver is calculated as st a,b,t,q =∑st d,a,b,t,q , the total driving mileage is s1 a,b,t,q =∑s1 d,a,b,t,q , the average speed is v a,b,t,q =s1 a,b,t,q / st a,b,t,q ; S5, calculating the route proficiency of a single driver and the global route proficiency; Among them, the driver's route proficiency is f(q,a,b)=∑((v a,b,t,q / v a,b,t )) / countf q,a,b , where countf q,a,b The number of records of the starting point a and the end point b of driver q in step S4 is accumulated for these records of the driver; The driver's global proficiency is g(q) = ∑f(q,a,b) / countg q , where countg q The total number of routes taken by the driver; S6, calculating the grid proficiency of a single driver; Here, e is defined as any grid number, and h(q,e) is the driver’s proficiency in a certain grid; Select q, a, b in step S4 and calculate f(q, a, b): Where h(q,e)=∑f(q,a,b) / counth q,a,b , counth q,a,b For all historical routes of q, add up the number of times they pass through grid e and calculate h(q,e); S7. Calculate the driver's proficiency based on the starting point and the end point; Wherein, step S7 comprises the following steps: S71. If f(q,a,b) exists, it means the driver has walked the route, then r(q,a,b)=f(q,a,b); S72. If the driver has not taken this route, find all the grids e along the route from a to b, and their set is S a,b ,r(q,a,b)=∑h(q,e) / countr a,b , where countr a,b For S a,b Medium grid; S8. Calculate ETA based on the driver's driving proficiency; Among them, ETA(a,b) is the pick-up time returned by the third-party map platform. After calculation, META(q,a,b)=ETA(a,b) / r(q,a,b).

2. The ETA method based on the driver's driving proficiency factor according to claim 1, characterized in that: In step S6, if the driver has never been to grid e, then h(q,e)=g(q).

Images