CN112053116B

CN112053116B - Method and device for identifying carpooling orders

Info

Publication number: CN112053116B
Application number: CN202010946998.XA
Authority: CN
Inventors: 吴文亮; 叶加文
Original assignee: Jiangsu Yunmanman Tongcheng Information Technology Co ltd
Current assignee: Jiangsu Yunmanman Tongcheng Information Technology Co ltd
Priority date: 2020-09-10
Filing date: 2020-09-10
Publication date: 2023-11-03
Anticipated expiration: 2040-09-10
Also published as: CN112053116A

Abstract

The invention discloses a method and a device for identifying a carpooling order, comprising the following steps: acquiring a training sample set, wherein the training sample set comprises a positive sample and a negative sample; training a preset classification model by adopting a positive sample and a negative sample to obtain a target carpooling order identification model; receiving a goods source main order input by a user, and determining a departure point and a destination point of the goods source main order; generating a line space block diagram based on a departure point and a destination point of a main order of a goods source, and taking the order in the line space block diagram as an order to be spliced; constructing a to-be-carpooled sample pair by adopting a goods source main order and a to-be-carpooled order; and inputting the to-be-carpooled sample pair into the target carpooling order identification model, and outputting the carpooling order. Therefore, the technical problems that the order capable of being spliced can not be automatically identified in the prior art, the vehicle owner is difficult to quickly and flexibly acquire the order capable of being spliced, and the effective utilization rate of logistics resources is low are solved, and the vehicle owner is convenient to quickly and flexibly acquire the order capable of being spliced.

Description

Method and device for identifying carpooling orders

Technical Field

The invention relates to the technical field of order recognition, in particular to a method and a device for recognizing a carpooling order.

Background

With the rapid development of logistics industry and electronic information technology, the logistics transportation industry is increasingly closely connected with electronic commerce, a large number of owners and owners carry out carpooling transportation on orders through services provided by a logistics transportation transaction electronic commerce platform, and the problems that the owners of small-scale goods sources have difficulty in searching cars and the owners have difficulty in reasonably distributing the goods are solved to a certain extent.

However, in the existing carpooling scheme, the carpooling information is usually released, the released carpooling information is searched by the vehicle owners according to the own journey, information matching is performed one by one mainly, the carpooling orders can not be automatically identified, the vehicle owners can not easily and flexibly obtain the carpooling orders, and the effective utilization rate of logistics resources is low.

Disclosure of Invention

The invention provides a method and a device for identifying a carpooling order, which solve the technical problems that the carpooling order cannot be automatically identified in the prior art, a vehicle owner is difficult to quickly and flexibly acquire the carpooling order, and the effective utilization rate of logistics resources is low.

The invention provides a method for identifying a carpooling order, which comprises the following steps: acquiring a training sample set, wherein the training sample set comprises a positive sample and a negative sample; training a preset classification model by adopting the positive sample and the negative sample to obtain a target carpoolable order identification model; receiving a goods source main order input by a user, and determining a departure point and a destination point of the goods source main order; generating a line space block diagram based on the starting point and the destination point of the goods source main order, and taking the order in the line space block diagram as an order to be spliced; the line space block diagram comprises space blocks, and the types of the space blocks comprise geohash blocks or google S2 blocks; constructing a to-be-carpooled sample pair by adopting the goods source main order and the to-be-carpooled order; and inputting the to-be-carpooled sample pair into the target carpooling order identification model, and outputting the carpooling order.

Optionally, the step of acquiring a training sample set includes: determining a planning line from a preset map according to the latitude and longitude of a departure point and the latitude and longitude of a destination point of a preset main order; performing line point sampling operation on the planning line to obtain a main line point set; a preset intermediate point searching algorithm is adopted, and a first target intermediate point corresponding to the longitude and latitude of the departure point of each preset sub-order and a second target intermediate point corresponding to the longitude and latitude of the destination point of each preset sub-order are determined from the main line point set; adopting a first connecting line between a starting point of the preset sub-order and a destination point of the preset sub-order, and a second connecting line between the first target intermediate point and the second target intermediate point corresponding to the preset sub-order as a connecting line sample pair; respectively extracting target features in a plurality of connecting line sample pairs through a feature extraction algorithm; taking a connecting line sample pair, which is corresponding to the target characteristics meeting the preset carpooling condition, as the positive sample; and taking the connection line sample pair of which the target characteristic does not meet the preset carpool condition as the negative sample.

Optionally, the step of performing a line point sampling operation on the planned line to obtain a main line point set includes: extracting a plurality of line points from the planned line according to the first-level longitude and latitude; the line points comprise a departure point, a destination point and a plurality of intermediate points of the preset main order; converting longitude and latitude corresponding to the circuit points to obtain a plurality of space blocks corresponding to the circuit points respectively; and if the space blocks are connected in pairs, constructing a main line point set by adopting the line points.

Optionally, the method further comprises: if the space blocks which are not connected between every two blocks exist, obtaining line points corresponding to the space blocks which are not connected between every two blocks as line points to be interpolated; interpolation algorithm is adopted to interpolate the area between the line points to be interpolated to obtain at least one interpolation point; and constructing the main line point set by adopting the plurality of line points and the at least one interpolation point.

Optionally, the step of determining, from the main line point set, a first target intermediate point corresponding to the longitude and latitude of the departure point of each preset sub-order and a second target intermediate point corresponding to the longitude and latitude of the destination point of each preset sub-order by using a preset intermediate point search algorithm includes: creating a longitude key value query dictionary and a latitude key value query dictionary by adopting the main line point set according to longitude and latitude respectively; determining a dictionary with the largest key number from the longitude key value dictionary and the latitude key value dictionary as a target query dictionary; determining a first target intermediate point corresponding to the longitude and latitude of the departure point of each preset sub-order from the main line point set by adopting the target query dictionary; and determining a second target intermediate point corresponding to the longitude and latitude of the destination point of each preset sub-order from the main line point set by adopting the target query dictionary.

Optionally, the step of training a preset classification model by using the positive sample and the negative sample to obtain a target carpoolable order identification model includes: sequentially inputting the positive sample and the negative sample into the preset classification model to obtain a plurality of classification results; the classification result comprises meeting the preset carpooling condition and not meeting the preset carpooling condition; judging whether the error rates of the plurality of classification results are smaller than a preset threshold value or not; if yes, outputting a target carpooling order identification model.

Optionally, the method further comprises: and if not, returning to the step of acquiring the training sample set.

Optionally, the step of generating a line space block diagram based on the starting point and the destination point of the goods source main order, and taking the order in the line space block diagram as the order to be assembled includes: determining the longitude and latitude of the departure point of the goods source main order and the longitude and latitude of the destination point of the goods source main order; determining a path to be transported based on the longitude and latitude of the departure point of the goods source main order and the longitude and latitude of the destination point of the goods source main order; extracting a plurality of points to be transported from the path to be transported according to the first-level longitude and latitude; the points to be transported comprise a departure point of the goods source main order, a destination point of the goods source main order and a middle point of the goods source main order; converting longitude and latitude corresponding to the plurality of points to be transported to generate a line space block diagram; the line space block diagram comprises a plurality of space blocks corresponding to the plurality of points to be transported respectively; and determining the child orders of which the longitude and latitude of the departure point and the longitude and latitude of the destination point are in the plurality of space blocks as the to-be-carpooled orders.

Optionally, the step of constructing a pair of samples to be assembled by adopting the goods source main order and the order to be assembled includes: acquiring a departure point of the order to be assembled and a destination point of the order to be assembled; determining a third target intermediate point corresponding to the longitude and latitude of the starting point of the order to be assembled and a fourth target intermediate point corresponding to the longitude and latitude of the destination point of the order to be assembled from the intermediate points of the main order of the goods source by adopting a preset intermediate point searching algorithm; and adopting a first connecting line between the starting point of the to-be-spliced vehicle order and the destination point of the to-be-spliced vehicle order, and a second connecting line between the third target intermediate point and the fourth target intermediate point as a to-be-spliced vehicle sample pair.

The invention also provides a device for identifying the carpooling order, which comprises the following steps:

the training sample set acquisition module is used for acquiring a training sample set, wherein the training sample set comprises a positive sample and a negative sample; the target carpooling order recognition model generation module is used for training a preset classification model by adopting the positive sample and the negative sample to obtain a target carpooling order recognition model; the goods source main order receiving module is used for receiving the goods source main order input by a user and determining the departure point and the destination point of the goods source main order; the to-be-carpooled order determining module is used for generating a line space block diagram based on the starting point and the destination point of the goods source main order, and taking the order in the line space block diagram as the to-be-carpooled order; the line space block diagram comprises space blocks, and the types of the space blocks comprise geohash blocks or google S2 blocks; the to-be-carpooled sample pair construction module is used for constructing a to-be-carpooled sample pair by adopting the goods source main order and the to-be-carpooled order; and the carpooling order output module is used for inputting the pair of the samples to be carpooled into the target carpooling order identification model and outputting a carpooling order.

From the above technical scheme, the invention has the following advantages:

according to the method, the device and the system, the target carpooling order identification model is obtained based on training of the training data set on the preset classification model, the goods source main order input by a user is received, the starting point and the destination point of the goods source main order are determined, the route space block diagram is generated based on the starting point and the destination point, the order in the route space block diagram is used as a carpooling sub-order, the goods source main order and the carpooling order are adopted to construct a carpooling sample pair, the carpooling sample pair is finally input into the target carpooling identification model, whether the carpooling order can be carpooled or not is judged, if yes, the carpooling order is displayed, and therefore the technical problems that the carpooling order cannot be automatically identified in the prior art, the vehicle owner is difficult to rapidly and flexibly acquire the carpooling order, and the effective utilization of the logistics resource is low are solved, and the vehicle owner can rapidly and flexibly acquire the carpooling order conveniently.

Drawings

In order to more clearly illustrate the embodiments of the invention or the technical solutions of the prior art, the drawings which are used in the description of the embodiments or the prior art will be briefly described, it being obvious that the drawings in the description below are only some embodiments of the invention, and that other drawings can be obtained from these drawings without inventive faculty for a person skilled in the art.

FIG. 1 is a flow chart of steps of a method for identifying a poolable car order according to an embodiment of the present invention;

FIG. 2 is a flowchart illustrating steps of a method for identifying a collable order according to an alternative embodiment of the present invention;

FIG. 3 shows a schematic longitude and latitude diagram of a main line point set according to an embodiment of the present invention;

FIG. 4a is a schematic diagram of a path formed by circuit points according to an embodiment of the present invention;

FIG. 4b is a diagram illustrating the interpolation of circuit points according to an embodiment of the present invention;

FIG. 5 is a schematic diagram of a pair of connection samples according to an embodiment of the present invention;

FIG. 6 is a schematic diagram of target features in an embodiment of the invention;

FIG. 7a is a schematic diagram of a comparison of a connection sample pair with a positive sample according to an embodiment of the present invention;

FIG. 7b is a schematic diagram of alignment of a connection sample with a positive sample according to an alternative embodiment of the present invention;

FIG. 8 is a schematic diagram of an alignment of a wired sample pair with a negative sample in another embodiment of the present invention;

FIG. 9 is a flowchart of a training process of a preset classification model in an embodiment of the invention;

fig. 10 is a block diagram of a device for identifying a carpooling order according to an embodiment of the present invention.

Detailed Description

The embodiment of the invention provides a method and a device for identifying a carpooling order, which are used for solving the technical problems that the carpooling order cannot be automatically identified in the prior art, a vehicle owner is difficult to quickly and flexibly acquire the carpooling order, and the effective utilization rate of logistics resources is low.

In order to make the objects, features and advantages of the present invention more comprehensible, the technical solutions in the embodiments of the present invention are described in detail below with reference to the accompanying drawings, and it is apparent that the embodiments described below are only some embodiments of the present invention, but not all embodiments of the present invention. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.

Referring to fig. 1, fig. 1 is a flowchart illustrating steps of a method for identifying a carpooling order according to an embodiment of the invention.

The invention provides a method for identifying a carpooling order, which comprises the following steps:

step 101, obtaining a training sample set, wherein the training sample set comprises a positive sample and a negative sample;

in the embodiment of the invention, in order to train the preset classification model so as to improve the accuracy of judging whether the order can be carpooled, a training sample set preset by a user is required to be acquired at the moment, and the training sample set comprises a positive sample and a negative sample.

It should be noted that, a positive sample indicates that the corresponding order can be carpooled, and a negative sample indicates that the corresponding order cannot be carpooled.

Step 102, training a preset classification model by adopting the positive sample and the negative sample to obtain a target carpoolable order identification model;

in a specific implementation, the positive sample and the negative sample are sequentially input into a preset classification model for classification training to obtain a classification result, and when the classification result meets the requirement set by a user, the target carpooling order identification model can be obtained.

Step 103, receiving a goods source main order input by a user, and determining a departure point and a destination point of the goods source main order;

after the target carpooling order recognition model is obtained, the goods source main order input by the user can be received, and then the departure point and the destination point of the goods source main order are determined according to the record in the goods source main order, so that the longitude and latitude of the departure point and the longitude and latitude of the destination point of the goods source main order are obtained.

Step 104, generating a line space block diagram based on the departure point and the destination point of the goods source main order, and taking the order in the line space block diagram as an order to be spliced;

in an embodiment of the present invention, the line space block diagram includes space blocks, and types of the space blocks may include, but are not limited to: in order to retrieve an order to be assembled, which accords with the target order identification model capable of being input into the target order identification model capable of being assembled, a line space block diagram can be generated according to the starting point and the destination point of the goods source main order, if an order requiring assembly exists in the line space block diagram, the order is determined to be the order to be assembled, and then the order to be assembled is judged to be capable of being identified based on the target order identification model capable of being assembled.

Step 105, constructing a to-be-carpooled sample pair by adopting the goods source main order and the to-be-carpooled order;

in the embodiment of the application, after one or more to-be-carpooled orders are detected, in order to further detect whether the to-be-carpooled orders can be carpooled, a pair of to-be-carpooled samples can be constructed by adopting a goods source main order and the to-be-carpooled orders, so that the judgment process of a subsequent target carpooled order identification model is facilitated.

And 106, inputting the pair of to-be-carpooled samples into the target carpooling order identification model, and outputting a carpooling order.

In the specific implementation, after the to-be-carpooled sample pair is constructed, the to-be-carpooled sample pair is input into a target carpoolable order identification model, whether the to-be-carpooled order in the target carpoolable order identification model can be further judged, if so, the to-be-carpooled order can be displayed, and whether to carpoole is determined by a vehicle owner.

Referring to fig. 2, fig. 2 is a flowchart illustrating steps of a method for identifying a carpooling order according to an alternative embodiment of the present invention, which may include steps 201 to 208 as follows:

step 201, a training sample set is obtained, wherein the training sample set comprises a positive sample and a negative sample;

in an embodiment of the present invention, the step 201 may include the following sub-steps S1-S7:

step S1, determining a planning line from a preset map according to the latitude and longitude of a departure point and the latitude and longitude of a destination point of a preset main order;

in a specific implementation, the departure point longitude and latitude and the destination point longitude and latitude of a preset main order can be input in navigation software, and a planning line is generated in a preset map.

Optionally, the planned route may also be determined according to user-defined rules.

Referring to fig. 3, fig. 3 shows a schematic view of the longitude and latitude of the main line point set according to an embodiment of the present invention, where the longitude and latitude of some points in the main line point set are included.

S2, performing line point sampling operation on the planned line to obtain a main line point set;

referring to fig. 4a, fig. 4a shows a schematic path diagram formed by line points according to an embodiment of the present invention, wherein the path diagram includes a departure point, a destination point and a plurality of main order path intermediate sampling points, and the sub-step S2 may include the following sub-steps:

Extracting a plurality of line points from the planned line according to the first-level longitude and latitude; the line points comprise a departure point, a destination point and a plurality of intermediate points of the preset main order;

converting longitude and latitude corresponding to the circuit points to obtain a plurality of space blocks corresponding to the circuit points respectively;

and if the space blocks are connected in pairs, constructing a main line point set by adopting the line points.

It should be noted that the Geohash block may be obtained by a Geohash algorithm, the Google S2 block may be obtained by a Google S2 algorithm, and the Geohash algorithm and the Google S2 algorithm are both algorithms that encode longitude and latitude, change two dimensions into one dimension, and partition address locations.

The first-level longitude and latitude refers to longitude and latitude with longitude or latitude being a preset interval.

In the embodiment of the invention, after the planned line is obtained, line point extraction is needed to be carried out on the planned line, and in order to prevent the excessive calculated amount in the subsequent sub-order matching process, line points of a plurality of planned lines can be extracted according to the first-level longitude and latitude; and converting the longitude and latitude of the line point through a preset geohash algorithm to obtain a plurality of corresponding space blocks. When the space blocks are connected with each other, a plurality of line points are directly adopted to form a point set, and a main line point set is obtained.

Referring to fig. 4b, fig. 4b shows a line point interpolation process diagram according to an embodiment of the present invention, in which a plurality of spatial blocks converted by the longitude and latitude of a line point of a preset master order are included.

Further, the substep S2 may further include the substeps of:

if the space blocks which are not connected between every two blocks exist, obtaining line points corresponding to the space blocks which are not connected between every two blocks as line points to be interpolated;

interpolation algorithm is adopted to interpolate the area between the line points to be interpolated to obtain at least one interpolation point;

and constructing the main line point set by adopting the plurality of line points and the at least one interpolation point.

In the embodiment of the invention, as the distance between the first-level longitudes and latitudes may be far apart, the situation that space blocks which are not connected between every two adjacent space blocks may exist may occur, at this time, line points corresponding to the space blocks which are not connected between every two adjacent space blocks are obtained as line points to be interpolated, and an interpolation algorithm is adopted to interpolate the area between the line points to be interpolated so as to obtain a plurality of interpolation points; the point set formed by the plurality of line points and the at least one interpolation point is used as a main line point set, so that the space blocks are kept from being disconnected, continuity is kept, and the situation that the departure point and the destination point of a sub order fall in the area between the line points to be interpolated and the sub order cannot be detected is further prevented.

The interpolation algorithm may be an equidistant interpolation algorithm or other non-equidistant interpolation algorithms, which is not limited in the present application.

Step S3, a preset intermediate point searching algorithm is adopted, and a first target intermediate point corresponding to the longitude and latitude of the departure point of each preset sub-order and a second target intermediate point corresponding to the longitude and latitude of the destination point of each preset sub-order are determined from the main line point set;

in one example of the present application, the substep S3 may include the substeps of:

creating a longitude key value query dictionary and a latitude key value query dictionary by adopting the main line point set according to longitude and latitude respectively;

determining a dictionary with the largest key number from the longitude key value dictionary and the latitude key value dictionary as a target query dictionary;

determining a first target intermediate point corresponding to the longitude and latitude of the departure point of each preset sub-order from the main line point set by adopting the target query dictionary;

and determining a second target intermediate point corresponding to the longitude and latitude of the destination point of each preset sub-order from the main line point set by adopting the target query dictionary.

In the embodiment of the application, after the main line point set is acquired, a corresponding key value query dictionary, namely a longitude key value query dictionary and a latitude key value query dictionary, can be respectively created according to the longitude and the latitude.

In a specific implementation, the process of constructing the key-value query dictionary may be as follows:

the formula for establishing the longitude and latitude dictionary key is as follows:

lonkey_i＝|lon_key_x|，min_lon<＝x<＝max_lon

latkey_i＝|lat_key_x|，min_lat<＝x<＝max_lat

wherein: min_lon is the minimum longitude of the current master order line; max_lon is the maximum longitude of the current master order line; min_lat is the minimum latitude of the current main order line; max_lat is the maximum latitude of the current master order line.

The key value query dictionary is as follows:

lonkey_i:

(lon_lonkey_i_x1,lat_latkey_n1_y1)

(lon_lonkey_i_x2,lat_latkey_n2_y2)

(lon_lonkey_i_x3,lat_latkey_n3_y3)

(lon_lonkey_i_x4,lat_latkey_n2_y4)

taking longitude and latitude as (103, 25) and (104, 26) as examples, the obtained longitude and latitude key dictionaries are as follows:

longitude key dictionary:

lonkey:103

[103.82381，25.565736]

[103.875265，25.654019]

lonkey:104

[104.355421，25.674809]

[104.907274，25.801758]

[104.988451，25.780516]

latitude key value dictionary:

latkey:25

[103.82381，25.565736]

[103.875265，25.654019]

[104.355421，25.674809]

[104.907274，25.801758]

[104.988451，25.780516]

latkey:26

[105.704427，26.030069]

[105.90691，26.20362]

[106.368016，26.498424]

[106.432995，26.526775]

selecting a dictionary with the largest key value number from the longitude key value dictionary and the latitude key value dictionary as a target query dictionary to preset the latitude and longitude of a main order to be (103.82381,25.565736), wherein the latitude and longitude of a destination is as follows: (111.9253,33.03928) for example, the list of key values of the longitude key value dictionary and the latitude key value dictionary obtained from the main order route is:

list of longitude key values:

dict_keys([103,104,105,106,107,108,109,110,111,112,102,113])

latitude key value list:

dict_keys([25,26,27,28,29,30,31,32,33,24,34])

it can be seen that the longitude key dictionary is queried to obtain 12 key values, the latitude key dictionary is queried to obtain 11 key values, and the selected longitude key dictionary is more suitable.

Determining a first target intermediate point corresponding to the longitude and latitude of the departure point of each preset sub-order from the main line point set by adopting a selected target query dictionary;

In the embodiment of the invention, the preset sub-order line is as follows: (106, 26) to (111, 30) for example, using the above-determined target query dictionary, namely, the longitude key value dictionary, the result is:

the first result queried according to the departure point of the sub-order is:

lonkey:106

[106.368016，26.498424]

[106.432995，26.526775]

[106.56753，26.585955]

[106.683199，26.694735]

[106.802274，26.699852]

[106.904592，26.832105]

the second result queried according to the sub-order destination point is:

lonkey:111

[111.144582，28.672934]

[111.385084，28.804119]

[111.625586，28.935304]

[111.744527，28.906615]

[111.8709225，29.421682]

[111.93412025，29.6792155]

[111.997318，29.936749]

[111.996426，32.771209]

[111.769822，32.988989]

[111.795469，33.096272]

[111.865764，33.08105]

[111.925304，33.039015]

[111.9253，33.03928]

finally, respectively calculating Euclidean distance between the sub-order starting point and each point in the first result, and selecting the point with the smallest Euclidean distance as a first target intermediate point, namely [106.56753,26.585955] in the example;

the Euclidean distance between the sub-order departure point and each point in the second result is calculated, and the point with the smallest Euclidean distance is selected as the second target intermediate point, namely [111.99731829.936749] in this example.

A substep S4, employing a first connection line between a departure point of the preset sub-order and a destination point of the preset sub-order, and a second connection line between the first target intermediate point and the second target intermediate point corresponding to the preset sub-order as a connection sample pair;

Referring to fig. 5, fig. 5 shows a schematic diagram of a pair of connection samples in an embodiment of the present invention.

In a specific implementation, the first target intermediate point and the second target intermediate point are intermediate points corresponding to the sub-order departure place and destination respectively, and the sub-order departure place and destination are connected to obtain a sub-order departure place and destination connecting line, namely a first connecting line, the first target intermediate point and the second target intermediate point are connected to obtain an intermediate point connecting line closest to the sub-order, namely a second connecting line, and a connecting line sample pair is obtained by adopting a combination of the first connecting line and the second connecting line.

S5, respectively extracting target features in a plurality of connecting line sample pairs through a feature extraction algorithm;

the feature extraction algorithm refers to an algorithm that calculates the correlation between the first connection line and the second connection line in the connection line sample pair.

Referring to fig. 6, fig. 6 shows a schematic diagram of target features in an embodiment of the present invention, where a main order is a main order and a spelling order is a sub order, and the target features may include, but are not limited to, the following features:

feature 1: shortest distance sf_mfs_dist from start point of car pooling order to sampling point on main order line

Feature 2: the shortest distance from the departure place of the carpool order to the sampling point on the main order line is compared with the line of the interval section of the main order

sf_mfs_dist/mfs_mts_dist

Feature 3: the shortest distance from the departure place of the carpool order to the sampling point on the line of the main order is compared with the line of the carpool order

sf_mfs_dist/sub_dist

Feature 4: shortest distance st_mts_dist from car pooling order destination to sampling point on main order line

Feature 5: the shortest distance from the car pooling order destination to the sampling point on the main order line is compared with the line of the main order interval

st_mts_dist/mfs_mts_dist

Feature 6: the shortest distance from the car pooling order destination to the sampling point on the main order line is compared with the car pooling order line

st_mts_dist/sub_dist

Feature 7: near the primary order destination sampling point-near the primary order departure sampling point-the carpooling destination, formed angle theta1

Feature 8: near the main order departure point, - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -) the main order destination, and- -sampling point- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

Feature 9: the linear distance of the carpool order is compared with the linear distance of the main order

sub_dist/main_dist

Feature 10: the ratio of the total distance of the car pooling order detour and the linear distance of the main order interval section

(sf_mfs_dist+sub_dist+st_mts_dist)/mfs_mts_dist

Feature 11: the extra distance of the bypass of the carpool order is compared with the linear distance of the carpool order

(sf_mfs_dist+st_mts_dist)/sub_dist

Wherein, the abbreviations of the target features are:

(1)main_dist:main_from---->main_to

(2)sub_dist:sub_from---->sub_to

(3)sf_mfs_dist:sub_from---->main_from_sample

(4)sf_mts_dist:sub_from---->main_to_sample

(5)st_mts_dist:sub_to---->main_to_sample

(6)st_mfs_dist:sub_to---->main_from_sample

(7)mfs_mts_dist:main_from_sample---->main_to_sample

s6, taking a connection line sample pair corresponding to the target feature meeting the preset carpooling condition as the positive sample;

For comparison, referring to fig. 7a and 7b, fig. 7a and 7b show a schematic diagram of a comparison of a wired sample pair and a positive sample according to two alternative embodiments of the present invention, respectively, wherein the positive sample may be labeled 1 at the end of the data display, wherein the unlabeled sample is an unlabeled sample, and the labeled sample is a labeled sample.

In the embodiment of the invention, after the target feature is acquired, the target feature is judged according to the preset carpooling condition set by the user so as to determine the preset sub-order belonging to the positive sample.

For example: the target feature is the ratio of the straight line distance of the carpool order to the straight line distance of the main order, and the preset carpool condition can be set to be that the ratio of the straight line distance of the carpool order to the straight line distance of the main order is smaller than a certain threshold.

The target features may include a plurality of target features, each having a corresponding preset carpooling condition, and when having a plurality of target features, determining that all target features satisfy the preset carpooling condition, determining that the corresponding connection sample pair is a positive sample, or determining that the connection sample pair is a negative sample.

And S7, taking the connection line sample pair of which the target characteristic does not meet the preset carpool condition as the negative sample.

Referring to fig. 8, fig. 8 is a schematic diagram showing a comparison of a wired sample pair and a negative sample in another embodiment of the present invention, wherein the negative sample may be marked as 0 at the end of the data display.

In another example of the present invention, the above step 102 may be replaced with the following steps 202-205:

referring to fig. 9, fig. 9 shows a flowchart of a training process of a preset classification model in an embodiment of the present invention.

Step 202, sequentially inputting the positive sample and the negative sample into the preset classification model to obtain a plurality of classification results;

in the embodiment of the invention, after positive samples and negative samples are sequentially input into the preset classification model, the preset classification model is operated once every positive sample or negative sample is input, so that a classification result is obtained; the classification result comprises meeting the preset carpooling condition and not meeting the preset carpooling condition.

Step 203, judging whether the error rates of the plurality of classification results are smaller than a preset threshold;

and 204, if yes, outputting a target carpooling order identification model.

When the iteration times of the preset classification model reach a threshold value, namely after all positive samples and negative samples are input, comparing the positive samples and the negative samples with correct classification according to the output classification result, and when the error is smaller than the error threshold value, indicating that the preset classification model is trained, and outputting the target carpooling order identification model.

It should be noted that the error threshold may be set by the user according to the actual situation, which is not limited by the embodiment of the present invention.

Step 205, if not, returning to the step of obtaining the training sample set.

In a specific implementation, it may happen that training a preset classification model by using a training sample set still cannot obtain a target carpoolable order recognition model, and at this time, the step of obtaining the training sample set may be returned to, and a new training sample set may be obtained again to continue the training process of the model.

Optionally, the step of inputting the positive sample and the negative sample of the training sample set into the preset classification model may be returned after the structure of the preset classification model is adjusted, which is not limited in the embodiment of the present invention.

Step 206, receiving a goods source main order input by a user, and determining a departure point and a destination point of the goods source main order;

in the embodiment of the present invention, the implementation process of step 206 is similar to that of step 103, and will not be repeated here.

Step 207, generating a line space block diagram based on the departure point and the destination point of the goods source main order, and taking the order in the line space block diagram as an order to be spliced;

In an embodiment of the present invention, the step 207 may be the following sub-steps:

determining the longitude and latitude of the departure point of the goods source main order and the longitude and latitude of the destination point of the goods source main order;

determining a path to be transported based on the longitude and latitude of the departure point of the goods source main order and the longitude and latitude of the destination point of the goods source main order;

extracting a plurality of points to be transported from the path to be transported according to the first-level longitude and latitude; the points to be transported comprise a departure point of the goods source main order, a destination point of the goods source main order and a middle point of the goods source main order;

converting longitude and latitude corresponding to the plurality of points to be transported to generate a line space block diagram; the line space block diagram comprises a plurality of space blocks corresponding to the plurality of points to be transported respectively;

and determining the child orders of which the longitude and latitude of the departure point and the longitude and latitude of the destination point are in the plurality of space blocks as the to-be-carpooled orders.

In specific implementation, determining a path to be transported based on the longitude and latitude of a departure point and the longitude and latitude of a destination point of the goods source main order by combining preset navigation software or a navigation map; based on the space blocks converted from each point to be transported in the path to be transported, determining whether the longitude and latitude of the departure point and the longitude and latitude of the destination point of the sub-order are both in the space blocks, and if yes, determining the sub-order as the order to be assembled.

Step 208, constructing a to-be-carpooled sample pair by adopting the goods source main order and the to-be-carpooled order;

optionally, the step 208 may include the sub-steps of:

acquiring a departure point of the order to be assembled and a destination point of the order to be assembled;

determining a third target intermediate point corresponding to the longitude and latitude of the starting point of the order to be assembled and a fourth target intermediate point corresponding to the longitude and latitude of the destination point of the order to be assembled from the intermediate points of the main order of the goods source by adopting a preset intermediate point searching algorithm;

and adopting a first connecting line between the starting point of the to-be-spliced vehicle order and the destination point of the to-be-spliced vehicle order, and a second connecting line between the third target intermediate point and the fourth target intermediate point as a to-be-spliced vehicle sample pair.

In the embodiment of the application, the training results in sample pairs of inputs required by the target carpool order identification model. After identifying the to-be-spliced vehicle order based on the goods source main order, acquiring a departure point of the to-be-spliced vehicle order and a destination point of the to-be-spliced vehicle order, and determining a third target intermediate point corresponding to the longitude and latitude of the departure point of the to-be-spliced vehicle order and a fourth target intermediate point corresponding to the longitude and latitude of the destination point of the to-be-spliced vehicle order from the intermediate points of the goods source main order based on the preset intermediate point searching algorithm; and connecting a departure point of the to-be-spliced vehicle order and a destination point of the to-be-spliced vehicle order to serve as a third connecting line, connecting a third target intermediate point and a fourth target intermediate point to serve as a fourth connecting line, and adopting the third connecting line and the fourth connecting line as a to-be-spliced vehicle sample pair.

And step 209, inputting the pair of to-be-carpooled samples into the target carpooling order identification model, and outputting a carpooling order.

In another example of the present invention, after the to-be-carpooled order is obtained, the to-be-carpooled sample pair is input into the target carpooling order identification model, the target carpooling order identification model classifies the to-be-carpooling sample pair according to the characteristics of the to-be-carpooling sample pair to determine whether the to-be-carpooled order meets the preset carpooling condition, and the to-be-carpooled order meeting the preset carpooling condition is output as the carpooling order, and in actual operation, the to-be-carpooling order can be displayed on a screen to be checked by a vehicle owner so that the vehicle owner can further confirm whether the carpooling is performed.

Referring to fig. 10, fig. 10 is a block diagram illustrating a device for identifying a carpooling order according to an alternative embodiment of the invention.

A carpooling order identification device, comprising:

a training sample set obtaining module 901, configured to obtain a training sample set, where the training sample set includes a positive sample and a negative sample;

the target carpooling order identification model generation module 902 is configured to train a preset classification model by using the positive sample and the negative sample to obtain a target carpooling order identification model;

the goods source main order receiving module 903 is configured to receive a goods source main order input by a user, and determine a departure point and a destination point of the goods source main order;

the to-be-carpooled order determining module 904 is configured to generate a line space block diagram based on a departure point and a destination point of the goods source main order, and take an order in the line space block diagram as the to-be-carpooled order;

the to-be-carpooled sample pair construction module 905 is configured to construct a to-be-carpooled sample pair by adopting the goods source main order and the to-be-carpooled order;

and the carpooling order output module 906 is used for inputting the pair of samples to be carpooled into the target carpooling order identification model and outputting a carpooling order.

Optionally, the training sample set obtaining module 901 includes:

The planning line determining submodule is used for determining a planning line from a preset map according to the longitude and latitude of a departure point and the longitude and latitude of a destination point of a preset main order;

the main line point set determining sub-module is used for executing line point sampling operation on the planning line to obtain a main line point set;

the target intermediate point searching sub-module is used for determining a first target intermediate point corresponding to the longitude and latitude of the departure point of each preset sub-order and a second target intermediate point corresponding to the longitude and latitude of the destination point of each preset sub-order from the main line point set by adopting a preset intermediate point searching algorithm;

a connection sample pair generating sub-module, configured to use a first connection line between a departure point of the preset sub-order and a destination point of the preset sub-order, and a second connection line between the first target intermediate point and the second target intermediate point corresponding to the preset sub-order as a connection sample pair;

the target feature extraction submodule is used for respectively extracting target features in a plurality of connecting line sample pairs through a feature extraction algorithm;

a positive sample determining sub-module, configured to use a pair of connection samples corresponding to the target feature satisfying a preset carpool condition as the positive sample;

And the negative sample sub-module is used for taking the connecting line sample pair which does not meet the target characteristics and corresponds to the preset carpooling condition as the negative sample.

Optionally, the main line point set determining submodule includes:

the line point extraction unit is used for extracting a plurality of line points from the planned line according to the first-level longitude and latitude; the line points comprise a departure point, a destination point and a plurality of intermediate points of the preset main order;

the space block determining unit is used for converting longitude and latitude corresponding to the plurality of line points to obtain a plurality of space blocks corresponding to the plurality of line points respectively; the line space block diagram comprises space blocks, and the types of the space blocks comprise geohash blocks or google S2 blocks;

and the first main line point set construction unit is used for constructing a main line point set by adopting the plurality of line points if the plurality of space blocks are connected in pairs.

Optionally, the main line point set determining sub-module further includes:

the to-be-interpolated line point determining unit is used for acquiring line points corresponding to the space blocks which are not connected between every two if the space blocks which are not connected between every two exist as to-be-interpolated line points;

the interpolation point determining unit is used for interpolating the area between the line points to be interpolated by adopting an interpolation algorithm to obtain at least one interpolation point;

And the second main line point set constructing unit is used for constructing the main line point set by adopting the plurality of line points and the at least one interpolation point.

Optionally, the target intermediate point searching sub-module includes:

the dictionary creation unit is used for creating a longitude key value query dictionary and a latitude key value query dictionary by adopting the main line point set according to longitude and latitude respectively;

a target query dictionary determining unit configured to determine, as a target query dictionary, a dictionary having the largest number of key values from the longitude key value dictionary and the latitude key value dictionary;

the first target intermediate point determining unit is used for determining a first target intermediate point corresponding to the longitude and latitude of the departure point of each preset sub-order from the main line point set by adopting the target query dictionary;

and the second target intermediate point determining unit is used for determining a second target intermediate point corresponding to the longitude and latitude of the destination point of each preset sub-order from the main line point set by adopting the target query dictionary.

Optionally, the target collage order identification model generating module 902 includes:

the classification result determining submodule is used for sequentially inputting the positive sample and the negative sample into the preset classification model to obtain a plurality of classification results; the classification result comprises meeting the preset carpooling condition and not meeting the preset carpooling condition;

The error rate judging sub-module is used for judging whether the error rates of the plurality of classification results are smaller than a preset threshold value or not;

and the target carpooling order identification model generation sub-module is used for outputting the target carpooling order identification model if yes.

Optionally, the target collage order identification model generating module 902 further includes:

and the return sub-module is used for returning to the step of acquiring the training sample set if not.

Optionally, the waiting-for-car order determining module 904 includes:

the goods source main order longitude and latitude determining submodule is used for determining the longitude and latitude of a departure point of the goods source main order and the longitude and latitude of a destination point of the goods source main order;

the to-be-transported path determining submodule is used for determining a to-be-transported path based on the longitude and latitude of the departure point of the goods source main order and the longitude and latitude of the destination point of the goods source main order;

the point to be transported extraction submodule is used for extracting a plurality of points to be transported from the path to be transported according to the first-level longitude and latitude; the points to be transported comprise a departure point of the goods source main order, a destination point of the goods source main order and a middle point of the goods source main order;

the line space block diagram generation sub-module is used for converting longitude and latitude corresponding to the plurality of points to be transported to generate a line space block diagram; the line space block diagram comprises a plurality of space blocks corresponding to the plurality of points to be transported respectively;

And the waiting car order determining sub-module is used for determining sub-orders of which the longitude and latitude of the departure point and the longitude and latitude of the destination point are in the plurality of space blocks as waiting car orders.

Optionally, the to-be-carpooled sample pair construction module 905 includes:

the waiting car order data acquisition sub-module is used for acquiring a departure point of the waiting car order and a destination point of the waiting car order;

a target intermediate point determining submodule, configured to determine a third target intermediate point corresponding to the longitude and latitude of the departure point of the order to be assembled and a fourth target intermediate point corresponding to the longitude and latitude of the destination point of the order to be assembled from the intermediate points of the main order of the goods source by adopting a preset intermediate point searching algorithm;

and the to-be-carpooled sample pair generating sub-module is used for adopting a third connecting line between the departure point of the to-be-carpooled order and the destination point of the to-be-carpooled order and a fourth connecting line between the third target intermediate point and the fourth target intermediate point as the to-be-carpooled sample pair.

It will be clear to those skilled in the art that, for convenience and brevity of description, specific working procedures of the apparatus and units described above may refer to corresponding procedures in the foregoing method embodiments, which are not described herein again.

In the several embodiments provided by the present invention, it should be understood that the disclosed apparatus and method may be implemented in other manners. For example, the apparatus embodiments described above are merely illustrative, e.g., the division of the units is merely a logical function division, and there may be additional divisions when actually implemented, e.g., multiple units or components may be combined or integrated into another system, or some features may be omitted or not performed. Alternatively, the coupling or direct coupling or communication connection shown or discussed with each other may be an indirect coupling or communication connection via some interfaces, devices or units, which may be in electrical, mechanical or other form.

The units described as separate units may or may not be physically separate, and units shown as units may or may not be physical units, may be located in one place, or may be distributed on a plurality of network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

In addition, each functional unit in the embodiments of the present invention may be integrated in one processing unit, or each unit may exist alone physically, or two or more units may be integrated in one unit. The integrated units may be implemented in hardware or in software functional units.

The integrated units, if implemented in the form of software functional units and sold or used as stand-alone products, may be stored in a computer readable storage medium. Based on such understanding, the technical solution of the present invention may be embodied essentially or in part or all of the technical solution or in part in the form of a software product stored in a storage medium, including instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method according to the embodiments of the present invention. And the aforementioned storage medium includes: a U-disk, a removable hard disk, a Read-Only Memory (ROM), a random access Memory (RAM, random Access Memory), a magnetic disk, or an optical disk, or other various media capable of storing program codes.

The above embodiments are only for illustrating the technical solution of the present invention, and not for limiting the same; although the invention has been described in detail with reference to the foregoing embodiments, it will be understood by those of ordinary skill in the art that: the technical scheme described in the foregoing embodiments can be modified or some technical features thereof can be replaced by equivalents; such modifications and substitutions do not depart from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for identifying a collage order, comprising:

obtaining a training sample set, the training sample set comprising a positive sample and a negative sample, the step of obtaining the training sample set comprising: determining a planning line from a preset map according to the latitude and longitude of a departure point and the latitude and longitude of a destination point of a preset main order; performing line point sampling operation on the planning line to obtain a main line point set; a preset intermediate point searching algorithm is adopted, and a first target intermediate point corresponding to the longitude and latitude of the departure point of each preset sub-order and a second target intermediate point corresponding to the longitude and latitude of the destination point of each preset sub-order are determined from the main line point set; adopting a first connecting line between a starting point of the preset sub-order and a destination point of the preset sub-order, and a second connecting line between the first target intermediate point and the second target intermediate point corresponding to the preset sub-order as a connecting line sample pair; respectively extracting target features in a plurality of connecting line sample pairs through a feature extraction algorithm; taking a connecting line sample pair, which is corresponding to the target characteristics meeting the preset carpooling condition, as the positive sample; taking a connection sample pair of which the target characteristics do not meet the preset carpool conditions as the negative sample;

Training a preset classification model by adopting the positive sample and the negative sample to obtain a target carpoolable order identification model;

receiving a goods source main order input by a user, and determining a departure point and a destination point of the goods source main order;

generating a line space block diagram based on the starting point and the destination point of the goods source main order, and taking the order in the line space block diagram as an order to be spliced; the line space block diagram comprises space blocks, and the types of the space blocks comprise geohash blocks or google S2 blocks;

constructing a to-be-carpooled sample pair by adopting the goods source main order and the to-be-carpooled order;

and inputting the to-be-carpooled sample pair into the target carpooling order identification model, and outputting the carpooling order.

2. The method of claim 1, wherein the step of performing a line point sampling operation on the planned line to obtain a main line point set comprises:

3. The method as recited in claim 2, further comprising:

4. The method of claim 1, wherein the step of determining, from the main line point set, a first target intermediate point corresponding to a latitude and longitude of a departure point of each preset sub-order and a second target intermediate point corresponding to a latitude and longitude of a destination point of each preset sub-order using a preset intermediate point search algorithm comprises:

creating a longitude key value dictionary and a latitude key value dictionary by adopting the main line point set according to longitude and latitude respectively;

5. The method of claim 1, wherein training a preset classification model using the positive and negative samples to obtain a target collapsable order identification model comprises:

sequentially inputting the positive sample and the negative sample into the preset classification model to obtain a plurality of classification results; the classification result comprises meeting the preset carpooling condition and not meeting the preset carpooling condition;

judging whether the error rates of the plurality of classification results are smaller than a preset threshold value or not;

if yes, outputting a target carpooling order identification model.

6. The method as recited in claim 5, further comprising:

and if not, returning to the step of acquiring the training sample set.

7. The method of claim 1, wherein the step of generating a line space block diagram based on the origin and destination points of the source main order, and taking the order in the line space block diagram as the order for the car to be taken includes:

8. The method of claim 7, wherein said step of constructing a pair of to-be-carpooled samples using said source master order and said to-be-carpooled order comprises:

And adopting a third connecting line between the starting point of the to-be-spliced vehicle order and the destination point of the to-be-spliced vehicle order and a fourth connecting line between the third target intermediate point and the fourth target intermediate point as a to-be-spliced vehicle sample pair.

9. A device for identifying a poolable vehicle order, comprising:

the training sample set acquisition module is used for acquiring a training sample set, wherein the training sample set comprises a positive sample and a negative sample, and a planning line is determined from a preset map according to the longitude and latitude of a departure point and the longitude and latitude of a destination point of a preset main order; performing line point sampling operation on the planning line to obtain a main line point set; a preset intermediate point searching algorithm is adopted, and a first target intermediate point corresponding to the longitude and latitude of the departure point of each preset sub-order and a second target intermediate point corresponding to the longitude and latitude of the destination point of each preset sub-order are determined from the main line point set; adopting a first connecting line between a starting point of the preset sub-order and a destination point of the preset sub-order, and a second connecting line between the first target intermediate point and the second target intermediate point corresponding to the preset sub-order as a connecting line sample pair; respectively extracting target features in a plurality of connecting line sample pairs through a feature extraction algorithm; taking a connecting line sample pair, which is corresponding to the target characteristics meeting the preset carpooling condition, as the positive sample; taking a connection sample pair of which the target characteristics do not meet the preset carpool conditions as the negative sample;

The target carpooling order recognition model generation module is used for training a preset classification model by adopting the positive sample and the negative sample to obtain a target carpooling order recognition model;

the goods source main order receiving module is used for receiving the goods source main order input by a user and determining the departure point and the destination point of the goods source main order;

the to-be-carpooled order determining module is used for generating a line space block diagram based on the starting point and the destination point of the goods source main order, and taking the order in the line space block diagram as the to-be-carpooled order; the line space block diagram comprises space blocks, and the types of the space blocks comprise geohash blocks or google S2 blocks;

the to-be-carpooled sample pair construction module is used for constructing a to-be-carpooled sample pair by adopting the goods source main order and the to-be-carpooled order;

and the carpooling order output module is used for inputting the pair of the samples to be carpooled into the target carpooling order identification model and outputting a carpooling order.