CN110119827A - With the prediction technique and device of vehicle type - Google Patents
With the prediction technique and device of vehicle type Download PDFInfo
- Publication number
- CN110119827A CN110119827A CN201810118481.4A CN201810118481A CN110119827A CN 110119827 A CN110119827 A CN 110119827A CN 201810118481 A CN201810118481 A CN 201810118481A CN 110119827 A CN110119827 A CN 110119827A
- Authority
- CN
- China
- Prior art keywords
- vehicle type
- prediction
- vehicle
- probability
- order information
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 31
- 238000000354 decomposition reaction Methods 0.000 claims description 22
- 238000005516 engineering process Methods 0.000 abstract description 7
- 238000010586 diagram Methods 0.000 description 12
- 238000004458 analytical method Methods 0.000 description 4
- 239000011159 matrix material Substances 0.000 description 3
- 238000013178 mathematical model Methods 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 239000012141 concentrate Substances 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/40—Business processes related to the transportation industry
Landscapes
- Business, Economics & Management (AREA)
- Engineering & Computer Science (AREA)
- Human Resources & Organizations (AREA)
- Economics (AREA)
- Strategic Management (AREA)
- Theoretical Computer Science (AREA)
- Tourism & Hospitality (AREA)
- General Physics & Mathematics (AREA)
- Marketing (AREA)
- General Business, Economics & Management (AREA)
- Physics & Mathematics (AREA)
- Operations Research (AREA)
- Quality & Reliability (AREA)
- Game Theory and Decision Science (AREA)
- Entrepreneurship & Innovation (AREA)
- Development Economics (AREA)
- Health & Medical Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Primary Health Care (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
Prediction technique and device provided by the invention with vehicle type, by using the History Order information for obtaining user, prediction model is established with the time of using cars of vehicle type and/or with information such as vehicle places according to each in the History Order information, and predicted with vehicle type according to prediction model under current time and/or current location, so that vehicle type is used in available prediction.Since the user type utilizes prediction model and History Order acquisition of information, its compared with the existing technology for, the accuracy rate of prediction has obtained effective promotion, so that chauffeur platform is matched with lower single interface that vehicle type is switched with the actual demand of user using what the prediction obtained, reduce user's bill time, improves bill efficiency.
Description
Technical field
The present invention relates to Internet technical field more particularly to a kind of prediction techniques and device with vehicle type.
Background technique
With the development of internet technology, carrying out chauffeur on line using intelligent terminal becomes trend.And according to vehicle
The difference of type, chauffeur platform provide lower single port accordingly for each chauffeur type, so that user is at corresponding lower single interface
It places an order.
In the prior art, for the ease of the quick bill of user, chauffeur platform can use vehicle type needed for this to user
It is predicted, and interface is jumped into corresponding lower single interface, to reduce user's single interface the time it takes under switching.
But it is existing to being to be determined according in the upper order of user with vehicle type with the prediction of vehicle type, in reality
In the application of border, the accuracy rate of such prediction mode is relatively low.Once forecasting inaccuracy, user also needs to carry out current lower single interface
Switching, bill time increased instead.
Summary of the invention
For it is above-mentioned refer to it is low to the accuracy rate with vehicle type prediction in the prior art, thus caused by bill low efficiency
The problem of, the present invention provides a kind of prediction techniques and device with vehicle type.
On the one hand, a kind of prediction technique with vehicle type provided by the invention, comprising:
The History Order information for obtaining user, wherein including each with vehicle type of user in the History Order information
Time of using cars and/or with vehicle place；
Prediction model is established according to the History Order information, and to current time and/or is worked as according to the prediction model
It is predicted under preceding place with vehicle type.
It is described that prediction model, and root are established according to the History Order information in a kind of wherein optional embodiment
It is predicted with vehicle type according to the prediction model under current time and/or current location, comprising:
To each use vehicle type in the History Order information, vehicle is used according to the corresponding time of using cars and/or respectively respectively
Place carries out Cluster Decomposition, obtains each corresponding mixed Gauss model of vehicle type；
According to each mixed Gauss model, each probability density with vehicle type under current time and/or current location is determined
Value；
According to each probability density value prediction vehicle type with vehicle type.
In a kind of wherein optional embodiment, according to each probability density value prediction vehicle class with vehicle type
Type, comprising:
According to preset Bayesian model and each probability density value with vehicle type, calculate each with the first of vehicle type
Probability；
Vehicle type is used to use vehicle type as what prediction obtained using probability value in each first probability is highest.
In a kind of wherein optional embodiment, it is described using probability value in each first probability it is highest use vehicle type as
Prediction obtain include: with vehicle type
It determines in each first probability that probability value is highest and uses vehicle type, and judge whether the probability value with vehicle type is greater than
Equal to the first preset threshold；
If so, using vehicle type as prediction acquisition with vehicle type this and exporting.
In a kind of wherein optional embodiment, each use vehicle type in the History Order information, point
Cluster Decomposition is not carried out with vehicle place according to the corresponding time of using cars and/or respectively, obtains each corresponding mixed Gaussian of vehicle type
Before model, further includes:
Each number with vehicle type in statistical history order information, and establish Markov model；
Last time in History Order information is input to the Markov model with vehicle type, obtains each vehicle type
The second probability；
It determines in each second probability that probability value is highest and uses vehicle type, and judge whether the probability value with vehicle type is greater than
Equal to the second preset threshold；
If so, using vehicle type as prediction acquisition with vehicle type this and exporting；Otherwise, execution is described goes through to described
Each use vehicle type in history order information, carries out Cluster Decomposition with vehicle place according to the corresponding time of using cars and/or respectively respectively,
The step of obtaining each mixed Gauss model corresponding with vehicle type.
On the other hand, a kind of prediction meanss with vehicle type provided by the invention, comprising:
Data obtaining module, for obtaining the History Order information of user, wherein including in the History Order information
Each time of using cars with vehicle type of user and/or with vehicle place；
Prediction module, for establishing prediction model according to the History Order information, and according to the prediction model to working as
It is predicted under preceding time and/or current location with vehicle type.
In a kind of wherein optional embodiment, the prediction module is specifically used for:
To each use vehicle type in the History Order information, vehicle is used according to the corresponding time of using cars and/or respectively respectively
Place carries out Cluster Decomposition, obtains each corresponding mixed Gauss model of vehicle type；
According to each mixed Gauss model, each probability density with vehicle type under current time and/or current location is determined
Value；
According to each probability density value prediction vehicle type with vehicle type.
In a kind of wherein optional embodiment, the prediction module is specifically used for:
According to preset Bayesian model and each probability density value with vehicle type, calculate each with the first of vehicle type
Probability；
Vehicle type is used to use vehicle type as what prediction obtained using probability value in each first probability is highest.
In a kind of wherein optional embodiment, the prediction module is specifically used for:
It determines in each first probability that probability value is highest and uses vehicle type, and judge whether the probability value with vehicle type is greater than
Equal to the first preset threshold；
If so, using vehicle type as prediction acquisition with vehicle type this and exporting.
In a kind of wherein optional embodiment, the prediction module, to each in the History Order information
With vehicle type, Cluster Decomposition is carried out with vehicle place according to the corresponding time of using cars and/or respectively respectively, is obtained each corresponding with vehicle type
Mixed Gauss model before, be also used to:
Each number with vehicle type in statistical history order information, and establish Markov model；
Last time in History Order information is input to the Markov model with vehicle type, obtains each vehicle type
The second probability；
It determines in each second probability that probability value is highest and uses vehicle type, and judge whether the probability value with vehicle type is greater than
Equal to the second preset threshold；
If so, using vehicle type as prediction acquisition with vehicle type this and exporting；Otherwise, execution is described goes through to described
Each use vehicle type in history order information, carries out Cluster Decomposition with vehicle place according to the corresponding time of using cars and/or respectively respectively,
The step of obtaining each mixed Gauss model corresponding with vehicle type.
Prediction technique and device provided by the invention with vehicle type, by using obtain user History Order information,
Prediction model is established with the time of using cars of vehicle type and/or with information such as vehicle places according to each in the History Order information, and
It is predicted with vehicle type according to prediction model under current time and/or current location, so that vehicle is used in available prediction
Type.Since the user type is using prediction model and History Order acquisition of information, compared with the existing technology for, in advance
The accuracy rate of survey has obtained effective promotion, so that chauffeur platform is placed an order using what the prediction obtained with what vehicle type was switched
Interface is matched with the actual demand of user, reduces user's bill time, improves bill efficiency.
Detailed description of the invention
Fig. 1 is a kind of flow diagram for prediction technique with vehicle type that the embodiment of the present invention one provides；
Fig. 2 is a kind of flow diagram of the prediction technique with vehicle type provided by Embodiment 2 of the present invention；
Fig. 3 a is the schematic diagram of the corresponding taxi frequency of History Order information provided by Embodiment 2 of the present invention；
Fig. 3 b is the schematic diagram of the corresponding express frequency of History Order information provided by Embodiment 2 of the present invention；
Fig. 4 is a kind of flow diagram for prediction technique with vehicle type that the embodiment of the present invention three provides；
Fig. 5 is a kind of structural schematic diagram of the prediction meanss with vehicle type provided by the invention.
Specific embodiment
In order to make the object, technical scheme and advantages of the embodiment of the invention clearer, below in conjunction with the embodiment of the present invention
In attached drawing, technical scheme in the embodiment of the invention is clearly and completely described.
With the development of internet technology, carrying out chauffeur on line using intelligent terminal becomes trend.Along with chauffeur
Business sustainable development, user have on chauffeur platform it is a variety of it is alternative use vehicle types, for example: taxi, express,
Special train, windward driving are hired a car etc..
Conventionally, as every kind with the billing mechanism of vehicle type and the difference of chauffeur mechanism, different uses vehicle class
Type has different lower single interfaces.And quickly place an order for the ease of user or quick bill, when user opens chauffeur platform or touching
When sending out the chauffeur interface of chauffeur platform, chauffeur platform can to user this it is possible predicted with vehicle type, and switch to pre-
Survey with the corresponding lower single interface of vehicle type so that user quickly places an order.
But existing vehicle type is used according to user's upper order when this possible prediction with vehicle type to user
Determining.By taking the upper order of user is " express " as an example, when user opens chauffeur platform, chauffeur platform interface will directly switch
To lower single interface of " express ", such mode is not inconsistent with the actual demand of user largely, once that is, user this
It is not " express " with vehicle type, also needs to release lower single interface and switch to other lower single interfaces to carry out lower single operation, send out
Single-action rate is lower instead.Therefore, how the accuracy rate of summary prediction so that this prediction obtain it is practical with vehicle type and user
It is consistent as far as possible with vehicle type, to improve bill efficiency as technical problem urgently to be resolved.
For it is above-mentioned refer to it is low to the accuracy rate with vehicle type prediction in the prior art, thus caused by bill low efficiency
The problem of, the present invention provides a kind of prediction technique with vehicle type, Fig. 1 provides a kind of with vehicle class for the embodiment of the present invention one
The flow diagram of the prediction technique of type.
As shown in Figure 1, the prediction technique includes:
Step 101, the History Order information for obtaining user wherein include that each of user uses vehicle class in History Order information
Time of using cars of type and/or with vehicle place.
Step 102 is established prediction model according to History Order information, and to current time and/or is worked as according to prediction model
It is predicted under preceding place with vehicle type.
It should be noted that provided by the invention concretely use vehicle type with the executing subject of the prediction technique of vehicle type
Prediction meanss, which can be realized by way of hardware and/or software.Chauffeur platform can be generally integrated in be based on
Cloud server in, there is the data server of types of databases to be used cooperatively with the storage that chauffeur platform is based on, in addition, in advance
Surveying the server that device is based on can be same server with data server, or to be under the jurisdiction of same server cluster not
Same server, the present invention are not limited this.
In the present embodiment, History Order information refers to that user is serviced by the way that chauffeur platform is completed with vehicle.This uses vehicle
Service, which may include, is directed to taxi, express, special train, suitable by what real-time chauffeur or the reservation chauffeurs mode such as chauffeur carried out
Windmill is hired a car, the service of the vehicle type such as share-car.
It specifically may include user each order in History Order information with vehicle type, time of using cars and/or with vehicle
Point, in addition, being also possible that O/No., transaction amount etc. in order information.Further, the time of using cars can be specially
Order initiates time, order end time, order waiting time etc..It then may include having starting point, terminating point etc. with vehicle place
Deng.
Due to including much information in History Order information, by the analysis to these information would know that user right
With the Behavior law in the selection of vehicle type, i.e., probability is used with vehicle type under some special scenes.
For example, in the History Order information of a certain user, the order overwhelming majority that Mon-Fri issues is
This hire a car with vehicle type, and the order overwhelming majority that Saturday and Sunday issue be express this use vehicle type, and by with
The Behavior law of the user is analyzed on vehicle time dimension it is found that the user on weekdays be taxi with vehicle type
Probability is bigger, and the user is bigger with the probability that vehicle type is express festivals or holidays.For another example in the history of a certain user
In order information, use vehicle place be the order overwhelming majority from a certain office building as initial place be taxi this with vehicle class
Type, and use vehicle place to be from the order overwhelming majority in order as the initial place of a certain cell be express this with vehicle type,
And by analyzing the Behavior law of the user place of working it is found that the possible user of the office building in the dimension of in-use automotive place
Point, it is required bigger with the probability that vehicle type is taxi when user sets out with going to other purposes to job site,
The cell may be the location of the family of user, when user from family in going to other purposes when, it is required to use vehicle type
It is bigger for the probability of express.Certainly, also the behavior of user can be advised for time of using cars dimension and with vehicle place dimension simultaneously
Rule carries out comprehensive analysis, uses probability with vehicle type to obtain user under special scenes.
Therefore, the user behavior rule can be indicated by mathematical model and constructs corresponding prediction model to realize
To user's prediction of vehicle type, which may include but be not limited to Markov model, Gauss model, mixed Gaussian mould
The one of which model such as type, Bayesian model or a variety of models.Wherein, which can be used for according to the current time of using cars
And/or the current determining historic scenery similar with current scene in vehicle place, and find and made under historic scenery with vehicle type
Vehicle type is used for what prediction obtained.
The prediction technique with vehicle type that the embodiment of the present invention one provides is believed by using the History Order for obtaining user
Breath establishes prediction model with the time of using cars of vehicle type and/or with information such as vehicle places according to each in the History Order information,
And predicted with vehicle type according to prediction model under current time and/or current location, to can get the use of prediction
Vehicle type.Since the user type is using prediction model and History Order acquisition of information, compared with the existing technology for,
The accuracy rate of prediction has obtained effective promotion so that chauffeur platform using the prediction obtain switched with vehicle type under
Single interface is matched with the actual demand of user, reduces user's bill time, improves bill efficiency.
On the basis of example 1, Fig. 2 is a kind of prediction technique with vehicle type provided by Embodiment 2 of the present invention
Flow diagram.As shown in Fig. 2, the prediction technique includes:
Step 201, the History Order information for obtaining user wherein include that each of user uses vehicle class in History Order information
Time of using cars of type and/or with vehicle place.
Step 202 uses vehicle type to each in History Order information, respectively according to corresponding time of using cars and/or each
Cluster Decomposition is carried out with vehicle place, obtains each corresponding mixed Gauss model of vehicle type.
Step 203, according to each mixed Gauss model, determine each under current time and/or current location with vehicle type
Probability density value.
Step 204, according to each probability density value prediction vehicle type with vehicle type.
In the present embodiment two, similarly with embodiment one, History Order information refers to user by chauffeur platform
That completes is serviced with vehicle.This may include with vehicle service and is directed to by what the chauffeurs mode such as real-time chauffeur or reservation chauffeur carried out
In taxi, express, special train, windward driving, hire a car, the service of the vehicle type such as share-car.Specifically may include in History Order information
Each order of user with vehicle type, time of using cars and/or with vehicle place, ordered in addition, being also possible that in order information
Single number, transaction amount etc..Further, the time of using cars can be specially that order initiates time, order end time, order etc.
To time etc..It then may include having starting point, terminating point etc. with vehicle place.
And what is different from the first embodiment is that the present embodiment two provides the specific implementation that a kind of pair of user type is predicted
Mode.In the present embodiment two, use vehicle type for each in History Order information, can respectively according to its it is corresponding with vehicle when
Between and/or respectively carry out Cluster Decomposition with vehicle place, obtain each corresponding mixed Gauss model of vehicle type, below will be to vehicle
Time be illustrated for Cluster Decomposition.
Table 1 is the History Order information of a certain user, and Fig. 3 a is History Order information pair provided by Embodiment 2 of the present invention
The schematic diagram for the taxi frequency answered, Fig. 3 b are the corresponding express frequency of History Order information provided by Embodiment 2 of the present invention
Schematic diagram can first be handled the History Order information in table 1 by taking table 1 and Fig. 3 a and Fig. 3 b as an example, to extract the moment,
It will then carry out with vehicle type with the moment corresponding to obtain histogram shown in Fig. 3 a and Fig. 3 b, can clearly be obtained by histogram
Know, the use of the taxi of the user concentrates near 12 noon, and the frequency of use of express occurs respectively there are two peak value
In 7 points and at 18 points in afternoon.Therefore, for this dimension of the time of using cars of the user, the result of the Cluster Decomposition of taxi
Cluster centre for the time of using cars is 12 points, and it is 7 points and 18 that the Cluster Decomposition result of express, which is the cluster centre of time of using cars,
Point.
Then, determine Gaussian Profile centered on each center using the cluster centre, that is, be directed to taxi this
For vehicle type, the center that corresponds to is 12 points of Gauss model, and be directed to express this with for vehicle type, correspond to
The mixed Gauss model that is formed by stacking by two Gaussian Profiles that 7 points and center are 18 points of center.
Table 1
Wherein, above-mentioned mixed Gauss model to be established specifically based on the time of using cars various ways realization can be used, wherein
In a kind of mode, each use can be obtained by way of it will carry out constantly vectorization and calculate corresponding mean value of each moment and variance
The mixed Gauss model based on the time of using cars of vehicle type.
Aforesaid way only establishes prediction model according to History Order information in the present invention by taking time of using cars dimension as an example
It is illustrated, in other embodiments, can also establish mixed Gauss model according to vehicle place, or simultaneously according to the time of using cars
And mixed Gauss model is established with vehicle place, this will not be repeated here.Certainly, at the time of aforesaid way is according only in the time of using cars
This parameter, establishes mixed Gauss model, in other modes, can also be belonged to according to the date parameter in the time of using cars, festivals or holidays
Property other time of using cars information such as parameter carry out the foundation of mixed Gauss model, be also similar with vehicle place.
Then, current time can be substituted into the corresponding mixed Gauss model of each use vehicle type, when obtaining in this prior
Between under each user type probability density value, and according to each probability density value to the prediction with vehicle type.
Mixed Gauss model is established by using above-mentioned to realize the use according to the History Order information of user to user
Vehicle type is predicted, so that the accuracy of prediction is effectively raised, so that chauffeur platform uses vehicle class using the prediction
Corresponding chauffeur interface is pushed to user by type, improves user's bill efficiency.
On the basis of the above embodiment, it is preferred that in step 204 according to each pre- with the probability density value of vehicle type
It surveys and uses vehicle type, specifically include: according to preset Bayesian model and each probability density value with vehicle type, calculating each vehicle class
First probability of type；Vehicle type is used to use vehicle type as what prediction obtained using probability value in each first probability is highest.
Specifically, Bayesian model can be used for indicating the probability under each dynamic condition with vehicle type, be appreciated that
For to the mathematical expression form with vehicle type probability of happening.Therefore, in the present embodiment, by each probability with vehicle type
Density value, which carries out processing, can calculate each the first probability with vehicle type of acquisition, which illustrates its corresponding vehicle type
The current time of using cars and/or it is current use vehicle place this case that under generation a possibility that.Therefore, it is chosen in each second probability
Probability value is highest to use vehicle type to use vehicle type as prediction, will further improve the accuracy of prediction.
More preferably, in order to make more matching with vehicle type with user demand for prediction output, by probability in each first probability
It is worth and highest vehicle type is used to comprise determining that in each first probability that probability value is highest as what prediction obtained with vehicle type and use vehicle class
Type, and judge whether the probability value with vehicle type is more than or equal to the first preset threshold；If so, using vehicle type as in advance this
Survey obtain with vehicle type and export.
Specifically, due to the difference of the History Order information of each user, above-mentioned prediction model may not be suitable for
The prediction with vehicle type to the user of a certain type, such as the user that quantity on order is few.Therefore, by the way that the first default threshold is arranged
Value to above-mentioned acquisition probability value is maximum with vehicle type to be verified, and only the probability value at this with vehicle type meets first in advance
If when threshold value, just by this with vehicle type export, otherwise, then can be used other prediction models or it is existing in mode export prediction
With vehicle type.Certainly, for the user of this part, once its History Order information is supported enough to its prediction model
It establishes, mode above-mentioned also can be used and predicted.
Prediction technique provided by Embodiment 2 of the present invention with vehicle type, on the basis of example 1, by history
The Cluster Decomposition of order information, and combine mixed Gauss model, to be effectively predicted to vehicle type, predictablity rate
It is high.
On the basis of example 2, Fig. 4 is a kind of prediction technique with vehicle type that the embodiment of the present invention three provides
Flow diagram.As shown in figure 4, the prediction technique includes:
Step 301, the History Order information for obtaining user wherein include that each of user uses vehicle class in History Order information
Time of using cars of type and/or with vehicle place.
Each number with vehicle type in step 302, statistical history order information, and establish Markov model.
Last time in History Order information is input to Markov model with vehicle type by step 303, and acquisition is each to use vehicle
Second probability of type.
Step 304 determines the highest vehicle type of probability value in each second probability, and judges that this uses the probability value of vehicle type
Whether the second preset threshold is more than or equal to.
If so, thening follow the steps 305；If it is not, thening follow the steps 306.
This is used vehicle type as prediction acquisition with vehicle type and exported by step 305.
Step 306 uses vehicle type to each in History Order information, respectively according to corresponding time of using cars and/or each
Cluster Decomposition is carried out with vehicle place, obtains each corresponding mixed Gauss model of vehicle type.
Step 307, according to each mixed Gauss model, determine each under current time and/or current location with vehicle type
Probability density value.
Step 308, according to each probability density value prediction vehicle type with vehicle type.
In order to improve forecasting efficiency, the present embodiment three has carried out Markov model and aforementioned mixed Gauss model into one
Step mixing, so that the forecasting efficiency of entire prediction meanss gets a promotion, but also the predictable user scope of prediction meanss obtains
Further expansion.
Specifically, first as in embodiment one or embodiment two, History Order information refers to that user is flat by chauffeur
Platform is completed to be serviced with vehicle.This may include through real-time chauffeur with vehicle service or reserves what the chauffeurs modes such as chauffeur carried out
It is directed to taxi, express, special train, windward driving, hires a car, the service of the vehicle type such as share-car.Specifically may be used in History Order information
Including user each order with vehicle type, time of using cars and/or with vehicle place, in addition, being also possible that in order information
There are O/No., transaction amount etc..Further, the time of using cars can be specially that order is initiated the time, the order end time, ordered
Single waiting time etc..It then may include having starting point, terminating point etc. with vehicle place.
It then,, can be first with Ma Erke after the History Order information for obtaining user unlike previous embodiment
Husband's model is predicted with vehicle type the user's:
Table 2 is the History Order information of a certain user, counts each number with vehicle type in the History Order information,
To obtain statistical matrix as shown in table 3.
Table 2
Table 3
| Next time uses taxi | Next time uses express | |
| Last time uses taxi | 18 | 2 |
| Last time uses express | 2 | 0 |
According to History Order information determine user it is the last use vehicle type, and according to it is the last with vehicle type and
Above-mentioned statistical matrix jump the calculating of matrix, each the second probability with vehicle type predicted, that is, the express predicted
Probability be 0, the probability of the taxi of prediction is 1.It is then determined that the highest vehicle type of probability value in each second probability, and
When the probability value with vehicle type is more than or equal to the second preset threshold, vehicle type is used to use vehicle type as what prediction obtained this
It is exported, otherwise, then carries out the prediction with vehicle type using the mixed Gauss model in such as embodiment two, do not do herein superfluous
It states.
In the present embodiment, it as shown in above-mentioned example, is directed to some largely using certain single user with vehicle type
For, more effectively this kind of user can be predicted by Markov model, and for mixed Gauss model its
Calculation amount needed for predicting is also smaller, is conducive to the treatment effeciency for improving entire prediction meanss.Certainly, it is accordingly set in present embodiment
The second preset threshold has been set, so that when finding that Markov model is not suitable for the History Order information of user, it can be according to
Mixed Gauss model above-mentioned is effectively predicted to such user's with vehicle type.
Fig. 5 is a kind of structural schematic diagram of the prediction meanss with vehicle type provided by the invention, as shown in figure 5, the prediction
Structure, comprising:
Data obtaining module 10 includes wherein useful in History Order information for obtaining the History Order information of user
Each time of using cars with vehicle type at family and/or with vehicle place.
Prediction module 20, for establishing prediction model according to History Order information, and according to prediction model to current time
And/or it is predicted under current location with vehicle type.
Wherein the prediction meanss can be realized by way of hardware and/or software.Chauffeur platform institute base can be generally integrated in
In cloud server in, there is the data server of types of databases to be used cooperatively with the storage that chauffeur platform is based on, in addition,
The server that prediction meanss are based on can be same server with data server, or to be under the jurisdiction of same server cluster
Different server, the present invention are not limited this.
In the present embodiment, History Order information refers to that user is serviced by the way that chauffeur platform is completed with vehicle.This uses vehicle
Service, which may include, is directed to taxi, express, special train, suitable by what real-time chauffeur or the reservation chauffeurs mode such as chauffeur carried out
Windmill is hired a car, the service of the vehicle type such as share-car.
And specifically may include in History Order information user each order with vehicle type, time of using cars and/or use vehicle
Place, in addition, being also possible that O/No., transaction amount etc. in order information.Further, the time of using cars can be specific
Time, order end time, order waiting time etc. are initiated for order.It then may include having starting point, terminating point with vehicle place
Etc..
Due to including much information in History Order information, by the analysis to these information would know that user right
With the Behavior law in the selection of vehicle type, i.e., probability is used with vehicle type under some special scenes.
For example, in the History Order information of a certain user, the order overwhelming majority that Mon-Fri issues is
This hire a car with vehicle type, and the order overwhelming majority that Saturday and Sunday issue be express this use vehicle type, and by with
The Behavior law of the user is analyzed on vehicle time dimension it is found that the user on weekdays be taxi with vehicle type
Probability is bigger, and the user is bigger with the probability that vehicle type is express festivals or holidays.For another example in the history of a certain user
In order information, use vehicle place be the order overwhelming majority from a certain office building as initial place be taxi this with vehicle class
Type, and use vehicle place to be from the order overwhelming majority in order as the initial place of a certain cell be express this with vehicle type,
And by analyzing the Behavior law of the user place of working it is found that the possible user of the office building in the dimension of in-use automotive place
Point, it is required bigger with the probability that vehicle type is taxi when user sets out with going to other purposes to job site,
The cell may be the location of the family of user, when user from family in going to other purposes when, it is required to use vehicle type
It is bigger for the probability of express.Certainly, also the behavior of user can be advised for time of using cars dimension and with vehicle place dimension simultaneously
Rule carries out comprehensive analysis, uses probability with vehicle type to obtain user under special scenes.
Therefore, the user behavior rule can be indicated by mathematical model and constructs corresponding prediction model to realize
To user's prediction of vehicle type, which may include but be not limited to Markov model, Gauss model, mixed Gaussian mould
The one of which model such as type, Bayesian model or a variety of models.Wherein, which can be used for according to the current time of using cars
And/or the current determining historic scenery similar with current scene in vehicle place, and find and made under historic scenery with vehicle type
Vehicle type is used for what prediction obtained.
Preferably, prediction module 20 is specifically used for each use vehicle type in History Order information, respectively according to correspondence
Time of using cars and/or respectively carry out Cluster Decomposition with vehicle place, obtain each corresponding mixed Gauss model of vehicle type；According to each
Mixed Gauss model determines each probability density value with vehicle type under current time and/or current location；Vehicle class is used according to each
The probability density value prediction vehicle type of type.More preferably, prediction module 20 is specifically used for: according to preset Bayesian model and
The probability density value for respectively using vehicle type, calculates each the first probability with vehicle type；By the highest use of probability value in each first probability
Vehicle type uses vehicle type as what prediction obtained.Further, prediction module 20 is specifically used for determining probability in each first probability
Be worth it is highest use vehicle type, and judge whether the probability value with vehicle type is more than or equal to the first preset threshold；If so, should
It uses vehicle type as prediction acquisition with vehicle type and exports.
In addition, prediction module 20, vehicle type is being used to each in History Order information, when respectively according to corresponding with vehicle
Between and/or respectively carry out Cluster Decomposition with vehicle place, before obtaining each corresponding mixed Gauss model with vehicle type, be also used to: system
Each number with vehicle type in History Order information is counted, and establishes Markov model；It will be upper in History Order information
It once is input to Markov model with vehicle type, obtains each the second probability with vehicle type；Determine probability in each second probability
Be worth it is highest use vehicle type, and judge whether the probability value with vehicle type is more than or equal to the second preset threshold；If so, should
It uses vehicle type as prediction acquisition with vehicle type and exports；Otherwise, it executes to each use vehicle type in History Order information,
Cluster Decomposition is carried out with vehicle place according to the corresponding time of using cars and/or respectively respectively, is obtained each high with the corresponding mixing of vehicle type
The step of this model.
It is apparent to those skilled in the art that for convenience and simplicity of description, the system of foregoing description
Specific work process and corresponding beneficial effect, can refer to corresponding processes in the foregoing method embodiment, herein no longer
It repeats.
The prediction meanss with vehicle type that the embodiment of the present invention four provides are believed by using the History Order for obtaining user
Breath establishes prediction model with the time of using cars of vehicle type and/or with information such as vehicle places according to each in the History Order information,
And predicted with vehicle type according to prediction model under current time and/or current location, to can get the use of prediction
Vehicle type.Since the user type is using prediction model and History Order acquisition of information, compared with the existing technology for,
The accuracy rate of prediction has obtained effective promotion so that chauffeur platform using the prediction obtain switched with vehicle type under
Single interface is matched with the actual demand of user, reduces user's bill time, improves bill efficiency.
Those of ordinary skill in the art will appreciate that: realize that all or part of the steps of above-mentioned each method embodiment can lead to
The relevant hardware of program instruction is crossed to complete.Program above-mentioned can be stored in a computer readable storage medium.The journey
When being executed, execution includes the steps that above-mentioned each method embodiment to sequence；And storage medium above-mentioned include: ROM, RAM, magnetic disk or
The various media that can store program code such as person's CD.
Finally, it should be noted that the above embodiments are only used to illustrate the technical solution of the present invention., rather than its limitations；To the greatest extent
Pipe present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that: its according to
So be possible to modify the technical solutions described in the foregoing embodiments, or to some or all of the technical features into
Row equivalent replacement；And these are modified or replaceed, various embodiments of the present invention technology that it does not separate the essence of the corresponding technical solution
The range of scheme.
Claims (10)
1. a kind of prediction technique with vehicle type characterized by comprising
The History Order information for obtaining user, wherein including that each of user uses vehicle with vehicle type in the History Order information
Time and/or with vehicle place；
Prediction model is established according to the History Order information, and according to the prediction model to current time and/or current position
It is predicted under point with vehicle type.
2. prediction technique according to claim 1, which is characterized in that described established according to the History Order information is predicted
Model, and predicted with vehicle type according to the prediction model under current time and/or current location, comprising:
To each use vehicle type in the History Order information, vehicle place is used according to the corresponding time of using cars and/or respectively respectively
Cluster Decomposition is carried out, each corresponding mixed Gauss model of vehicle type is obtained；
According to each mixed Gauss model, each probability density value with vehicle type under current time and/or current location is determined；
According to each probability density value prediction vehicle type with vehicle type.
3. prediction technique according to claim 2, which is characterized in that according to described each pre- with the probability density value of vehicle type
It surveys and uses vehicle type, comprising:
According to preset Bayesian model and each probability density value with vehicle type, calculate each with the first of vehicle type general
Rate；
Vehicle type is used to use vehicle type as what prediction obtained using probability value in each first probability is highest.
4. prediction technique according to claim 3, which is characterized in that described by the highest use of probability value in each first probability
Vehicle type as prediction obtain include: with vehicle type
It determines in each first probability that probability value is highest and uses vehicle type, and judge whether the probability value with vehicle type is more than or equal to
First preset threshold；
If so, using vehicle type as prediction acquisition with vehicle type this and exporting.
5. according to the described in any item prediction techniques of claim 2-4, which is characterized in that described in the History Order information
It is each use vehicle type, according to the corresponding time of using cars and/or respectively carries out Cluster Decomposition, each use vehicle class of acquisition with vehicle place respectively
Before the corresponding mixed Gauss model of type, further includes:
Each number with vehicle type in statistical history order information, and establish Markov model；
Last in History Order information is input to the Markov model with vehicle type, is obtained each with the of vehicle type
Two probability；
It determines in each second probability that probability value is highest and uses vehicle type, and judge whether the probability value with vehicle type is more than or equal to
Second preset threshold；
If so, using vehicle type as prediction acquisition with vehicle type this and exporting；Otherwise, execution is described orders the history
Each use vehicle type in single information, carries out Cluster Decomposition with vehicle place according to the corresponding time of using cars and/or respectively respectively, obtains
Respectively the step of mixed Gauss model corresponding with vehicle type.
6. a kind of prediction meanss with vehicle type characterized by comprising
Data obtaining module, for obtaining the History Order information of user, wherein including user in the History Order information
Each time of using cars with vehicle type and/or with vehicle place；
Prediction module, for establishing prediction model according to the History Order information, and according to the prediction model to it is current when
Between and/or current location under predicted with vehicle type.
7. prediction meanss according to claim 6, which is characterized in that the prediction module is specifically used for:
To each use vehicle type in the History Order information, vehicle place is used according to the corresponding time of using cars and/or respectively respectively
Cluster Decomposition is carried out, each corresponding mixed Gauss model of vehicle type is obtained；
According to each mixed Gauss model, each probability density value with vehicle type under current time and/or current location is determined；
According to each probability density value prediction vehicle type with vehicle type.
8. prediction meanss according to claim 7, which is characterized in that the prediction module is specifically used for:
According to preset Bayesian model and each probability density value with vehicle type, calculate each with the first of vehicle type general
Rate；
Vehicle type is used to use vehicle type as what prediction obtained using probability value in each first probability is highest.
9. prediction meanss according to claim 8, which is characterized in that the prediction module is specifically used for:
It determines in each first probability that probability value is highest and uses vehicle type, and judge whether the probability value with vehicle type is more than or equal to
First preset threshold；
If so, using vehicle type as prediction acquisition with vehicle type this and exporting.
10. according to the described in any item prediction meanss of claim 7-9, which is characterized in that the prediction module is gone through to described
Each use vehicle type in history order information, carries out Cluster Decomposition with vehicle place according to the corresponding time of using cars and/or respectively respectively,
Before obtaining each corresponding mixed Gauss model with vehicle type, it is also used to:
Each number with vehicle type in statistical history order information, and establish Markov model；
Last in History Order information is input to the Markov model with vehicle type, is obtained each with the of vehicle type
Two probability；
It determines in each second probability that probability value is highest and uses vehicle type, and judge whether the probability value with vehicle type is more than or equal to
Second preset threshold；
If so, using vehicle type as prediction acquisition with vehicle type this and exporting；Otherwise, execution is described orders the history
Each use vehicle type in single information, carries out Cluster Decomposition with vehicle place according to the corresponding time of using cars and/or respectively respectively, obtains
Respectively the step of mixed Gauss model corresponding with vehicle type.
Priority Applications (5)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810118481.4A CN110119827A (en) | 2018-02-06 | 2018-02-06 | With the prediction technique and device of vehicle type |
| CN201980003279.XA CN110869953B (en) | 2018-02-06 | 2019-02-03 | System and method for recommending traffic travel service |
| PCT/CN2019/074723 WO2019154398A1 (en) | 2018-02-06 | 2019-02-03 | Systems and methods for recommending transportation services |
| US16/729,281 US20200134747A1 (en) | 2018-02-06 | 2019-12-27 | Systems and methods for recommending transportation services |
| US17/936,352 US20230044760A1 (en) | 2018-02-06 | 2022-09-28 | Systems and methods for recommending transportation services |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810118481.4A CN110119827A (en) | 2018-02-06 | 2018-02-06 | With the prediction technique and device of vehicle type |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN110119827A true CN110119827A (en) | 2019-08-13 |
Family
ID=67519388
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810118481.4A Pending CN110119827A (en) | 2018-02-06 | 2018-02-06 | With the prediction technique and device of vehicle type |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110119827A (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111383055A (en) * | 2020-03-12 | 2020-07-07 | 北京嘀嘀无限科技发展有限公司 | Method, device and equipment for selecting products in center cabin of taxi and computer readable storage medium |
| CN111898954A (en) * | 2020-07-31 | 2020-11-06 | 沙师弟(重庆)网络科技有限公司 | Vehicle matching method based on improved Gaussian mixture model clustering |
| CN112766748A (en) * | 2021-01-22 | 2021-05-07 | 北京嘀嘀无限科技发展有限公司 | Method, device, equipment, medium and product for adjusting service resources |
| CN114492878A (en) * | 2020-11-13 | 2022-05-13 | 上海擎感智能科技有限公司 | Reservation control method, terminal and storage medium for travel vehicle service |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104599217A (en) * | 2015-01-27 | 2015-05-06 | 北京嘀嘀无限科技发展有限公司 | Method and device for determining current destination of passenger |
| CN105183909A (en) * | 2015-10-09 | 2015-12-23 | 福州大学 | Social network user interest predicting method based on Gaussian mixture model |
| CN105303817A (en) * | 2015-09-16 | 2016-02-03 | 北京嘀嘀无限科技发展有限公司 | Travel manner planning method and device |
| US9473803B2 (en) * | 2014-08-08 | 2016-10-18 | TCL Research America Inc. | Personalized channel recommendation method and system |
| CN106897919A (en) * | 2017-02-28 | 2017-06-27 | 百度在线网络技术（北京）有限公司 | With the foundation of car type prediction model, information providing method and device |
| CN107018493A (en) * | 2017-04-20 | 2017-08-04 | 北京工业大学 | A kind of geographical position Forecasting Methodology based on continuous sequential Markov model |
-
2018
- 2018-02-06 CN CN201810118481.4A patent/CN110119827A/en active Pending
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US9473803B2 (en) * | 2014-08-08 | 2016-10-18 | TCL Research America Inc. | Personalized channel recommendation method and system |
| CN104599217A (en) * | 2015-01-27 | 2015-05-06 | 北京嘀嘀无限科技发展有限公司 | Method and device for determining current destination of passenger |
| CN105303817A (en) * | 2015-09-16 | 2016-02-03 | 北京嘀嘀无限科技发展有限公司 | Travel manner planning method and device |
| CN105183909A (en) * | 2015-10-09 | 2015-12-23 | 福州大学 | Social network user interest predicting method based on Gaussian mixture model |
| CN106897919A (en) * | 2017-02-28 | 2017-06-27 | 百度在线网络技术（北京）有限公司 | With the foundation of car type prediction model, information providing method and device |
| CN107018493A (en) * | 2017-04-20 | 2017-08-04 | 北京工业大学 | A kind of geographical position Forecasting Methodology based on continuous sequential Markov model |
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111383055A (en) * | 2020-03-12 | 2020-07-07 | 北京嘀嘀无限科技发展有限公司 | Method, device and equipment for selecting products in center cabin of taxi and computer readable storage medium |
| CN111383055B (en) * | 2020-03-12 | 2023-11-17 | 北京嘀嘀无限科技发展有限公司 | Method, device and equipment for selecting car renting center bin and computer readable storage medium |
| CN111898954A (en) * | 2020-07-31 | 2020-11-06 | 沙师弟(重庆)网络科技有限公司 | Vehicle matching method based on improved Gaussian mixture model clustering |
| CN111898954B (en) * | 2020-07-31 | 2024-01-12 | 沙师弟(重庆)网络科技有限公司 | Vehicle matching method based on improved Gaussian mixture model clustering |
| CN114492878A (en) * | 2020-11-13 | 2022-05-13 | 上海擎感智能科技有限公司 | Reservation control method, terminal and storage medium for travel vehicle service |
| CN112766748A (en) * | 2021-01-22 | 2021-05-07 | 北京嘀嘀无限科技发展有限公司 | Method, device, equipment, medium and product for adjusting service resources |
| CN112766748B (en) * | 2021-01-22 | 2024-06-07 | 北京嘀嘀无限科技发展有限公司 | Method, device, equipment, medium and product for adjusting service resources |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111159533B (en) | Intelligent charging service recommendation method and system based on user image | |
| CN110119827A (en) | With the prediction technique and device of vehicle type | |
| CN112258251B (en) | Grey correlation-based integrated learning prediction method and system for electric vehicle battery replacement demand | |
| CN102393839B (en) | Parallel data processing system and method | |
| CN111832869B (en) | Vehicle scheduling method and device, electronic equipment and storage medium | |
| CN109308538B (en) | Method and device for predicting transaction conversion rate | |
| CN105023196A (en) | Analysis method and device for charging transaction data of charging stations | |
| CN104991924A (en) | Method and apparatus for determining address of new supply point | |
| CN105389975A (en) | Chauffeured car scheduling method and apparatus | |
| CN108345670A (en) | A kind of service hot spot discovery method being used for 95598 electric power work orders | |
| CN115700717A (en) | Power distribution analysis method based on electric automobile power consumption demand | |
| US20240227611A1 (en) | Method and apparatus for operating electric vehicle charging infrastructure | |
| CN109920269A (en) | A kind of virtual parking space management system in parking lot and device | |
| CN116681450A (en) | Customer credit evaluation method and system supporting intelligent fee-forcing | |
| CN115759629A (en) | Site selection method, device, equipment and medium for power change station | |
| CN113990068B (en) | Traffic data processing method, device, equipment and storage medium | |
| CN119443569A (en) | A charging pile construction planning method and system based on big data analysis | |
| CN113094120A (en) | Event processing method and device | |
| CN111861192A (en) | Site selection method and device for electric vehicle charging station | |
| CN111507753A (en) | Information pushing method and device and electronic equipment | |
| CN112782584A (en) | Method, system, medium, and device for predicting remaining usage limit of battery power | |
| WO2023241388A1 (en) | Model training method and apparatus, energy replenishment intention recognition method and apparatus, device, and medium | |
| CN111353797A (en) | Resource allocation method and device and electronic equipment | |
| CN116629528A (en) | Method, device, equipment and storage medium for scheduling vehicles | |
| CN113890948A (en) | Resource allocation method based on voice outbound robot dialogue data and related equipment |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| AD01 | Patent right deemed abandoned | ||
| AD01 | Patent right deemed abandoned |
Effective date of abandoning: 20220826 |