CN109189876A

CN109189876A - A kind of data processing method and device

Info

Publication number: CN109189876A
Application number: CN201811010287.0A
Authority: CN
Inventors: 刘均; 张小琼
Original assignee: Shenzhen Launch Technology Co Ltd
Current assignee: Shenzhen Launch Technology Co Ltd
Priority date: 2018-08-31
Filing date: 2018-08-31
Publication date: 2019-01-11
Anticipated expiration: 2038-08-31
Also published as: CN109189876B

Abstract

The embodiment of the invention discloses data processing method and relevant apparatus.Obtain m location information of vehicle diagnostic equipment, the m location information itself geographical location information collected when carrying out vehicle diagnostics for the vehicle diagnostic equipment, wherein m is the integer greater than 2；Classified according to clustering algorithm to the m location information, obtains K location information set；According to preset confidence level condition, the highest first set of confidence level in the K location information set is determined, wherein K is the integer greater than 1 and less than N；According to the location information in the first set, the regional location of the vehicle salvage shop where the vehicle diagnostic equipment is determined.The position result that the technical solution of the embodiment of the present invention is conducive to enhancing positioning is credible, reduces computation complexity, is easy to the processing of mass data.

Description

A kind of data processing method and device

Technical field

The present invention relates to the data processing methods and device of data analysis field more particularly to a kind of vehicle diagnostic equipment.

Background technique

Vehicle diagnostic equipment is the important sources of big data era car networking data as car networking terminal.Although vehicle Diagnostic device is individual large scale equipment, but due to the flexibility that vehicle diagnostic equipment uses, position is variation.It obtains The location information of vehicle diagnostic equipment the characteristics of reflecting vehicle diagnostic equipment shift in position, therefore its position often has not Stationarity.In practical applications, the application scenarios of vehicle diagnostic equipment are mostly vehicle salvage shop, therefore can be examined by vehicle The location information of disconnected equipment determines the geographical location of maintenance factory, but does so and set firstly the need of the relatively-stationary vehicle diagnostics of determination Standby location information.

It is most direct at present on the extraction strategy of the location information of vehicle diagnostic equipment for problem to be solved Method is that one or two location information is extracted from the location information of the multiple equipment of acquisition by random sampling, by analysis Obtain the location information of vehicle diagnostic equipment.But the credible result handled in this way is low.

Summary of the invention

The embodiment of the present invention provides the method for data processing, enhances the credible of the location information of the vehicle diagnostic equipment obtained Property, computation complexity is reduced, the processing of mass data is easy to.

In a first aspect, the embodiment of the invention provides a kind of data processing methods, comprising:

M location information of vehicle diagnostic equipment is obtained, the m location information is that the vehicle diagnostic equipment is carrying out Itself geographical location information collected when vehicle diagnostics, wherein m is the integer greater than 2；According to clustering algorithm to the m A location information is classified, and K location information set is obtained；According to preset confidence level condition, the K position letter is determined The highest first set of confidence level in breath set, wherein K is the integer greater than 1 and less than N；According in the first set Location information determines the regional location of the vehicle salvage shop where the vehicle diagnostic equipment.

By implementing the embodiment of the present invention, according to clustering algorithm to the location informations of a large amount of vehicle diagnostic equipments of acquisition into Row classification, improves positioning accuracy, then goes out the highest first set of confidence level according to preset confidence level conditional filtering, reduces a The influence of other errors present information finally determines the regional location of the vehicle salvage shop where vehicle diagnostic equipment.Energy of the present invention The credibility of the regional location result of enough enhancing positioning, reduces computation complexity, is easy to the processing of mass data.

In one possible implementation, described to be classified according to clustering algorithm to the m location information, it obtains K location information set, comprising:

K cluster centre point is chosen from the sample set that the m location information is constituted；Wherein, the sample set For { x⁽¹⁾, x⁽²⁾..., x^(m), x⁽ⁱ⁾∈Rⁿ, x⁽ⁱ⁾For the location information of i-th of sample in the sample set, i=1, 2 ..., m；The center point set that the K cluster centre point is constituted is { μ⁽¹⁾, μ⁽²⁾..., μ^(K), μ^(j)∈Rⁿ, μ^(j)For institute State the location information of j-th of cluster centre point of center point set, j=1,2 ..., K；RⁿFor n-dimensional vector space, n be greater than or Person is equal to 1 integer；According to formula c^(j)=arg min_j||x⁽ⁱ⁾-μ^(j)||², calculate i-th of sample x⁽ⁱ⁾Affiliated class cluster c^(j)； Wherein, | | x⁽ⁱ⁾-μ^(j)||²For x⁽ⁱ⁾To μ^(j)Euclidean distance square, arg min_j||x⁽ⁱ⁾-μ^(j)||²To work as μ^(j)It is described When cluster centre point, make to obtain μ^(j)Euclidean distance all independent variable x for being minimized of quadratic sum⁽ⁱ⁾Set；According to public affairs FormulaUpdate each described cluster centre point μ^(j), until distortion functionConvergence；Wherein, c^(j)For class cluster j,Believe for the position of the class cluster j Cease x⁽ⁱ⁾,For all location informations in the class cluster j feature and,To count the class cluster j The number of middle location information,For each location information gathering to the class cluster j of the class cluster j The Euclidean distance quadratic sum of class central point；After distortion function convergence, K obtained class cluster corresponds to the K position letter Breath set.

In one possible implementation, the highest first set of confidence level is c in the K location information set^(j) Cluster centre point with the first set is μ^(j), j=k, 0 < k≤K, and k is integer；It is described according in the first set Location information, determine the regional location of the vehicle salvage shop where the vehicle diagnostic equipment, comprising: determine it is described first collection Close c^(j)With the cluster centre point μ of the first set^(j)Afterwards, with the cluster centre point μ^(j)For dot, radius is default value Border circular areas, determine the location information in the first set in the border circular areas；According in the border circular areas Location information determines the regional location of the vehicle salvage shop.

In one possible implementation, described according to preset confidence level condition, determine the K position information set The highest first set of confidence level in conjunction, comprising: calculate location information quantity in each location information set respectively and the ratio of m Value；Determine that the corresponding position information set of maximum ratio is combined into the highest first set of the confidence level.

In one possible implementation, the region of vehicle salvage shop belonging to the determination vehicle diagnostic equipment After position, further includes: obtain the corresponding position of location information of the onboard diagnostic device in the first set and adopted The vehicle-relevant data of collection.

Step 1: K cluster centre point is chosen from the sample set that the m location information is constituted；From the position of acquisition Data select the respective cluster centre point of class cluster described in some class clusters and random initializtion, therefore presetting K is cluster centre The quantity of point, also illustrates that the quantity of class cluster.Wherein, the sample set is { x⁽¹⁾, x⁽²⁾..., x^(m), x⁽ⁱ⁾∈Rⁿ, x⁽ⁱ⁾For The location information of i-th of sample in the sample set, i=1,2 ..., m；The central point that the K cluster centre point is constituted Collection is combined into { μ⁽¹⁾, μ⁽²⁾..., μ^(K), μ^(j)∈Rⁿ, μ^(j)For the location information of j-th of cluster centre point of the center point set, J=1,2 ..., K；RⁿFor n-dimensional vector space, n is the integer more than or equal to 1；

Step 2: according to formula c^(j)=arg min_j||x⁽ⁱ⁾-μ^(j)||², calculate i-th of sample x⁽ⁱ⁾Affiliated class clusterWherein, | | x⁽ⁱ⁾-μ^(j)||²For x⁽ⁱ⁾To μ^(j)Euclidean distance square, arg min_j||x⁽ⁱ⁾- μ (j) 2 is when μ (j) is When the cluster centre point, make all independent variable x that the quadratic sum for the Euclidean distance for obtaining μ (j) is minimized⁽ⁱ⁾Set；

Step 3: according to formulaCalculate each described cluster centre pointWherein, c^(j) For class cluster j,For the location information x of the class cluster j⁽ⁱ⁾,For positions all in class cluster j letter The feature of breath and,For the number for counting location information in the class cluster j；It indicates to calculate for the first time described poly- After class central point, cluster centre point result general name；And so on, it is subsequent to refer to Meaning all similarly.

Step 4: judging distortion functionWhether restrain；

Step 5: according to formula c^(j)=arg min_j||x⁽ⁱ⁾-μ^(j)||², calculate i-th of sample x⁽ⁱ⁾Affiliated class cluster

Step 6: according to formulaUpdate each described cluster centre point

Step 7: judging distortion functionWhether restrain；

Step 8: according to formula c^(j)=arg min_j||x⁽ⁱ⁾-μ^(j)||², calculate i-th of sample x⁽ⁱ⁾Affiliated class cluster

Step 9: according to formulaUpdate each described cluster centre point

Step 10: judging distortion functionWhether restrain；

Step 11: according to formula c^(j)=arg min_j||x⁽ⁱ⁾-μ^(j)||², calculate i-th of sample x⁽ⁱ⁾Affiliated class cluster

Step 12: according to formulaUpdate each described cluster centre point

Step 13: judging distortion functionWhether restrain；

……

Step W: judge distortion functionWhether restrain；Work as distortion functionWhen convergence, K cluster centre point μ is obtained^(j)With with K cluster centre point μ^(j) Corresponding class cluster c^(j), i.e., the described K obtained class cluster correspond to the K location information set；Wherein,For the class cluster j each location information to the class cluster j cluster centre point it is European away from From quadratic sum, step W is the last one step when the convergent clustering algorithm of the distortion function executes.

In one possible implementation, in the region position for determining vehicle salvage shop belonging to the vehicle diagnostic equipment After setting, further includes: obtain the corresponding position of location information of the vehicle diagnostic equipment in the first set and acquired Vehicle-relevant data.

Second aspect, the embodiment of the invention provides a kind of data processing equipments, comprising: first acquisition unit, grouping sheet Member, screening unit, determination unit, second acquisition unit；Wherein, the first acquisition unit obtains vehicle for background server M location information of diagnostic device, the m location information are adopted by the vehicle diagnostic equipment when carrying out vehicle diagnostics The geographical location information of itself of collection, wherein m is the integer greater than 2；The taxon is used for according to clustering algorithm to institute It states m location information to classify, obtains K location information set；The screening unit, for according to preset confidence level item Part determines the highest first set of confidence level in the K location information set, wherein K is the integer greater than 1 and less than N； The determination unit, for determining the vehicle where the vehicle diagnostic equipment according to the location information in the first set The regional location of maintenance factory.The second acquisition unit, for being determined belonging to the vehicle diagnostic equipment in the determination unit Vehicle salvage shop regional location after, it is corresponding to obtain location information of the onboard diagnostic device in the first set Position vehicle-relevant data collected.

By implementing the embodiment of the present invention, according to clustering algorithm to the location informations of a large amount of vehicle diagnostic equipments of acquisition into Row classification, improves positioning accuracy, then goes out the highest first set of confidence level according to preset confidence level conditional filtering, reduces a The influence of other errors present information finally determines the regional location of the vehicle salvage shop where vehicle diagnostic equipment.Energy of the present invention The credibility for enough enhancing region position result, reduces computation complexity, is easy to the processing of mass data.

In one possible implementation, the taxon is specifically used for: the sample constituted from the m location information K cluster centre point is chosen in this set；Wherein, the sample set is { x⁽¹⁾, x⁽²⁾..., x^(m), x⁽ⁱ⁾∈Rⁿ, x⁽ⁱ⁾For The location information of i-th of sample in the sample set, i=1,2 ..., m；The central point that the K cluster centre point is constituted Collection is combined into { μ⁽¹⁾, μ⁽²⁾..., μ^(K), μ^(j)∈Rⁿ, μ^(j)For the location information of j-th of cluster centre point of the center point set, J=1,2 ..., K；RⁿFor n-dimensional vector space, n is the integer more than or equal to 1；According to formula c^(j)=arg min_j||x⁽ⁱ⁾-μ^(j)||², calculate i-th of sample x⁽ⁱ⁾Affiliated class cluster c^(j)；Wherein, | | x⁽ⁱ⁾-μ^(j)||²For x⁽ⁱ⁾To μ^(j)Euclidean distance Square, arg min_j||x⁽ⁱ⁾-μ^(j)||²To work as μ^(j)When for the cluster centre point, make to obtain μ^(j)Euclidean distance square With all independent variable x being minimized⁽ⁱ⁾Set；According to formulaIt updates in each described cluster Heart point μ^(j), until distortion functionConvergence；Wherein, c^(j)For class cluster j, For the location information x of the class cluster j⁽ⁱ⁾,For all location informations in the class cluster j feature and,For the number for counting location information in the class cluster j,For the every of the class cluster j A location information to the class cluster j cluster centre point Euclidean distance quadratic sum；After distortion function convergence, obtain K class cluster corresponds to the K location information set.

In one possible implementation, the screening unit, is specifically used for: in K determining location information set The highest first set of confidence level is c^(j)Cluster centre point with the first set is μ^(j), j=k, 0 < k≤K, and k is whole Number；The determination unit is specifically used for: determining the first set c^(j)With the cluster centre point μ of the first set^(j)Afterwards, with The cluster centre point μ^(j)For dot, radius is the border circular areas of default value, is determined in the first set in the circle Location information in region；According to the location information in the border circular areas, the regional location of the vehicle salvage shop is determined.

In one possible implementation, the screening unit, is specifically used for: calculating in each location information set Location information quantity respectively with the ratio of m；Determine that the corresponding position information set of maximum ratio is combined into the confidence level highest One set.

In one possible implementation, described device, further includes: second acquisition unit, for described determining single After member determines the regional location of vehicle salvage shop belonging to the vehicle diagnostic equipment, the onboard diagnostic device is obtained in institute State the corresponding position of the location information vehicle-relevant data collected in first set.

In one possible implementation, the taxon is specifically used for:

Step 1: K cluster centre point is chosen from the sample set that the m location information is constituted；From the position of acquisition Data select the respective cluster centre point of class cluster described in some class clusters and random initializtion.Cluster centre point and each position data Vector length is identical, therefore presets the quantity that K is cluster centre point, also illustrates that the quantity of class cluster.Wherein, the sample set It is combined into { x⁽¹⁾, x⁽²⁾..., x^(m), x⁽ⁱ⁾∈Rⁿ, x⁽ⁱ⁾For the location information of i-th of sample in the sample set, i=1, 2 ..., m；The center point set that the K cluster centre point is constituted is { μ⁽¹⁾, μ⁽²⁾..., μ^(K), μ^(j)∈Rⁿ, μ^(j)For institute State the location information of j-th of cluster centre point of center point set, j=1,2 ..., K；RⁿFor n-dimensional vector space, n be greater than or Person is equal to 1 integer；

Step 3: according to formulaCalculate each described cluster centre pointWherein, c^(j) For class cluster j,For the location information x of the class cluster j⁽ⁱ⁾,For positions all in class cluster j letter The feature of breath and,For the number for counting location information in the class cluster j；It indicates to calculate for the first time described poly- After class central point, cluster centre point result general name；And so on, it is subsequent to refer to The meaning of expression is all same Reason.

Step 4: judging distortion functionWhether restrain；

Step 5: according to formulaCalculate i-th of sample x⁽ⁱ⁾Affiliated class cluster

Step 6: according to formulaUpdate each described cluster centre point

Step 7: judging distortion functionWhether restrain；

Step 9: according to formulaUpdate each described cluster centre point

Step 10: judging distortion functionWhether restrain；

Step 12: according to formulaUpdate each described cluster centre point

Step 13: judging distortion functionWhether restrain；

……

In one possible implementation, the screening unit, is specifically used for: in K determining location information set The highest first set of confidence level is c^(j)Cluster centre point with the first set is μ^(j), j=k, 0 < k≤K, and k is whole Number；The determination unit, is specifically used for: determining the first set c^(j)With the cluster centre point μ of the first set^(j)Afterwards, With the cluster centre point μ^(j)For dot, radius is the border circular areas of default value, is determined in the first set in the circle Location information in shape region；According to the location information in the border circular areas, the regional location of the vehicle salvage shop is determined.

In one possible implementation, described device, further includes: second acquisition unit, for described determining single After member determines the regional location of vehicle salvage shop belonging to the vehicle diagnostic equipment, the vehicle diagnostic equipment is obtained in institute State the corresponding position of the location information vehicle-relevant data collected in first set.

In embodiments of the present invention, divided according to location information of the clustering algorithm to a large amount of vehicle diagnostic equipments of acquisition Class improves positioning accuracy, then goes out the highest first set of confidence level according to preset confidence level conditional filtering, reduces individual wrong The accidentally influence of location information finally efficiently uses the target position information in set according to certain rule in the first aggregate, Determine the regional location of the vehicle salvage shop where vehicle diagnostic equipment.Implement the present invention, region position result can be enhanced Credibility reduces computation complexity, is easy to the processing of mass data.

Detailed description of the invention

In order to more clearly explain the embodiment of the invention or the technical proposal in the existing technology, the present invention will be implemented below Example or attached drawing needed to be used in the description of the prior art are briefly described, it should be apparent that, it is the accompanying drawings in the following description, attached Table is only some embodiments of the present invention, for those of ordinary skill in the art, before not making the creative labor Put, can also according to these attached drawings, subordinate list obtain other attached drawings, subordinate list.

Fig. 1 is the system architecture schematic diagram of vehicle diagnostic equipment data processing provided in an embodiment of the present invention；

Fig. 2 is the interaction schematic diagram of vehicle diagnostic equipment and server provided in an embodiment of the present invention；

Fig. 3 is the table provided in an embodiment of the present invention for arranging and m location information being presented；

Fig. 4-15 is provided in an embodiment of the present invention to be shown according to K mean algorithm m location information data point classification results It is intended to；Wherein, Fig. 4 is the data point distribution figure of m location information provided in an embodiment of the present invention；Fig. 5 is the embodiment of the present invention Data shown in Fig. 4 of offer pass through one time i-th sample x⁽ⁱ⁾Affiliated class clusterCalculating and each described cluster centre PointCalculating classification results figure；Figure 15 is m location information provided in an embodiment of the present invention by K mean cluster algorithm Final classification results figure；

Figure 16 be it is provided in an embodiment of the present invention according to Figure 15 statistics location information set in location information quantity with And its ratio with m；

Figure 17 is vehicle salvage shop's regional location determining in the first set of Figure 16 screening provided in an embodiment of the present invention Schematic diagram；

Figure 18 is the structural schematic diagram of vehicle diagnostic equipment data processing equipment provided in an embodiment of the present invention；

Figure 19 is a kind of structural schematic diagram of equipment provided in an embodiment of the present invention.

Specific embodiment

Following will be combined with the drawings in the embodiments of the present invention, and technical solution in the embodiment of the present invention carries out clear, complete Site preparation description.Obviously, described embodiments are only a part of the embodiments of the present invention, instead of all the embodiments.It is based on Embodiment in the present invention, it is obtained by those of ordinary skill in the art without making creative efforts every other Embodiment shall fall within the protection scope of the present invention.

Description and claims of this specification and term " first ", " second ", " third " and " in the attached drawing Four " etc. are not use to describe a particular order for distinguishing different objects.In addition, term " includes " and " having " and it Any deformation, it is intended that cover and non-exclusive include.Such as it contains the process, method of a series of steps or units, be System, product or equipment are not limited to listed step or unit, but optionally further comprising the step of not listing or list Member, or optionally further comprising other step or units intrinsic for these process, methods, product or equipment.

Referenced herein " embodiment " is it is meant that a particular feature, structure, or characteristic described can wrap in conjunction with the embodiments Containing at least one embodiment of the present invention.Each position in the description occur the phrase might not each mean it is identical Embodiment, nor the independent or alternative embodiment with other embodiments mutual exclusion.Those skilled in the art explicitly and Implicitly understand, embodiment described herein can be combined with other embodiments.The technical side of the embodiment of the present invention Case can be applied to data processing, the fields such as clustering.When the field of method, apparatus application is with scene difference, the present invention Specific equipment, the title in place also can be different in embodiment.

Firstly, the part term in the present invention is explained, in order to those skilled in the art understand that.

(1) vehicle-relevant data is the collected data in the vehicle salvage shop after clustering algorithm determines, subsequent The collected data are handled, realize scale, the management state of the analysis affiliated vehicle salvage shop of vehicle diagnostic equipment Etc. contents purpose, will analysis result feed back to vehicle salvage shop, so that vehicle salvage shop scientifically adjusts the various aspects of itself Situation.

(2) vehicle diagnostic equipment, being may include commonly various in the vehicle salvage shops such as diagnostic device, maintenance of equipment set Standby general name.

(3) Euclidean distance (euclidean metric), can also be referred to as euclidean metric, be one and generally use Distance definition, refer to natural length (i.e. point to origin of the actual distance or vector between two points in an n-dimensional space Distance).In the present invention, location data points are divided into where the nearest cluster centre point of the data point Euclidean distance In class cluster.In order to calculate convenience, with the Euclidean distance of each location information data point to cluster centre point square for divide mark It is quasi-.

(4) K-Medians algorithm is a kind of variant of K-Means, is that gathering for class cluster is calculated with the median of data set Class central point.

(5) average drifting (Meanshift) clustering algorithm is the close quarters that data point is found based on sliding window. Mean shift clustering algorithm is the algorithm based on mass center, passes through be updated to put in sliding window by the candidate point of central point Mean value is completed, positioning the central point of each class cluster.Then it carries out similar window removal to the candidate window to handle, finally Formation center point set and corresponding grouping class cluster.

(6) greatest hope (EM) clustering algorithm for using Gaussian Mixture (GMM) model is to assume initially that data point is in Gauss Distribution, corresponding K-Means assume data point distribution be it is circular, K-Means is a special circumstances of GMMs, the side of being Difference in all dimensions all close to 0 when cluster will show circle.Gaussian Profile (ellipse) gives a possibility that more, The shape of class cluster: mean value and standard deviation can be described by following two parameter.So the class cluster can be any shape The ellipse of shape has standard deviation on direction in x, y.Therefore, each Gaussian Profile is assigned to single class cluster.Do cluster Before, the optimization algorithm of greatest hope (EM) is used first.Find the mean value and standard deviation of data set.

(7) spectral clustering (Spectral Clustering), is the clustering method based on graph theory, by sample number According to the feature vector of Laplacian Matrix clustered, to reach the cluster to sample data.The meaning of spectrum is illustrated, As follows: such as matrix A, the entirety of its all characteristic values are just referred to as the spectrum of A.Algorithm relevant to spectrum is and feature mostly It is worth relevant algorithm.And spectral radius is exactly maximum characteristic value in all characteristic values.

(8) Dbscan clustering algorithm (Density-Based Spatial Clustering of Applications With Noise), it is a more representational density-based algorithms, class cluster is defined as the connected point of density by it Maximum set, can be class cluster having region division highdensity enough, and can be found in the spatial database of noise The cluster of arbitrary shape.

(9) hierarchical clustering algorithm (hierarchical methods), be recursively data object is merged or Division, until certain termination condition meets.According to the isolation of level, particularly may be divided into from top and under division level Cluster and bottom-up Agglomerative Hierarchical Clustering.

First one of system architecture that the embodiment of the present invention is based on is described below, referring to Figure 1, Fig. 1 is The system architecture schematic diagram of vehicle diagnostic equipment data processing, as shown in Figure 1, at vehicle diagnostic equipment data proposed by the present invention Reason method can be applied to the system architecture.The system architecture contains in vehicle, vehicle diagnostic equipment and background server etc. Hold.The icon of background server, represents background server and can be and be made of several servers.Vehicle 1, vehicle 2 ..., vehicle 12 respectively represent and receive the vehicle of diagnosis in certain position, explanation is numbered only for differentiation.Vehicle The position diagnostic device 1-A, the position vehicle diagnostic equipment 1-B, the position vehicle diagnostic equipment 1-C, the position vehicle diagnostic equipment 1-D, Respectively represent the location information for belonging to the vehicle diagnostic equipment 1 of maintenance factory 1 in this 4 positions A, B, C, D.And so on, it can be with Learn the position vehicle diagnostic equipment 2-A ..., the meanings of these titles of the position vehicle diagnostic equipment 3-D.Maintenance factory 1 and Corresponding border circular areas, maintenance factory 2 and corresponding border circular areas, maintenance factory 3 and corresponding border circular areas, respectively represent vehicle Diagnostic device 1, vehicle diagnostic equipment 2, vehicle salvage shop belonging to vehicle diagnostic equipment 3 regional location.Vehicle diagnostic equipment The location information of itself can upload to background server by transmission modes such as networks.Vehicle, vehicle diagnostic equipment, maintenance factory With the contents such as background server can not exclusive list, so picture specification, the embodiment of the present invention only list centainly for convenience Quantity, but do not represent the use number of practical application.Use position mostly in maintenance factory due to vehicle diagnostic equipment, So most vehicle diagnostic equipment icons concentrate in the border circular areas of maintenance factory；Appliance icon not in border circular areas indicates Vehicle diagnostic equipment is by use, in maintenance factory of the equipment not belonging to it.

It is understood that the system architecture in Fig. 1 is the illustrative embodiment of one of embodiment of the present invention. System architecture in the embodiment of the present invention may include but be not limited only to system above framework.

Below with reference to the embodiment of above system framework and data processing provided in the present invention, to what is proposed in the present invention Technical problem is made a concrete analysis of and is solved.

Fig. 2 is referred to, Fig. 2 is the interaction schematic diagram of vehicle diagnostic equipment and background server, below in conjunction with Fig. 2, from The interaction side of vehicle diagnostic equipment and background server is described, this method embodiment mainly by taking K-Means algorithm as an example into Row explanation, can specifically include step S201- step S204.It optionally, can also include step S205.Wherein, step S202 Provide the possibility implementation of other clustering algorithms.

Step S201: obtaining m location information of vehicle diagnostic equipment, and the m location information is the vehicle diagnostics Equipment itself geographical location information collected when carrying out vehicle diagnostics, wherein m is the integer greater than 2；

Specifically, any one location information in the m location information includes at least longitude and latitude information.Its In, any one location information can store in the form of the n-dimensional vector of space or in the form of orderly real number pair.The acquisition Location information form, may include: the location information of individual equipment itself for obtaining vehicle diagnostic equipment and passing back；Alternatively, Obtain the location information set for the multiple equipment itself that vehicle diagnostic equipment is passed back.Algorithm apply and calculating process in, it is above-mentioned Different storage forms may have the difference in different data processing method details, not influence the core application of algorithm.

Step S202: classified according to clustering algorithm to the m location information, obtain K location information set；

Specifically, before being classified according to clustering algorithm to the m location information, for the ease of showing according to not The classification results that same clustering algorithm is classified, can first pre-process m location information, wherein pretreatment mode It may include: that m location information is subjected to arrangement presentation in table form, refer to Fig. 3, Fig. 3 is that the embodiment of the present invention mentions The table of m location information is presented in the arrangement of confession；Alternatively, m location information is carried out arrangement presentation in the form of images, please join See that Fig. 4, Fig. 4 are the data point distribution figures of m location information provided in an embodiment of the present invention.In embodiments of the present invention, be with For the method for arranging m location information in the form of images.Then, the m location information is carried out according to clustering algorithm Classification, obtains K location information set.Wherein, the classification selection of clustering algorithm depends on type, the purpose of cluster of data.

Main clustering algorithm can be divided into as follows: partition clustering, is based on Density Clustering, fuzzy clustering at hierarchical clustering. All there is the algorithms being used widely in every one kind method, such as: K-means clustering algorithm, level in partition clustering Coagulation type hierarchical clustering algorithm in cluster etc..It should be noted that the research of clustering problem is not limited solely to K-means calculation Method etc. clusters firmly, i.e., each data can only be classified as one kind.Fuzzy clustering is also that relatively broad one is studied in clustering A branch.Fuzzy clustering determines each data membership by membership function in the degree of each cluster, rather than by a data Object is referred to rigidly in certain cluster, such as FCM algorithm.Thus, implementation of the invention is not limited to utilize certain A kind of algorithm realizes the analysis to data, but can according to the specific features of data, using one or more kinds of algorithms or Mode makes every effort to calculate easy and conclusion confidence level.For example, rule of thumb first the pumping based on its distribution can be carried out to sample Then sample carries out hierarchical clustering within the scope of small sample, the K value then obtained with hierarchical clustering, is applied to entire sample and carries out K-means cluster.In embodiments of the present invention, it in conjunction with the feature of location information in the embodiment of the present invention, is calculated respectively with following 7 kinds For method, classify respectively to location information.

In one possible implementation, classified according to K-Means algorithm to the m location information, obtain K A location information set, refers to Fig. 5-15, may include: that K are chosen from the sample set that the m location information is constituted Cluster centre point；Wherein, the sample set is { x⁽¹⁾, x⁽²⁾..., x^(m), x⁽ⁱ⁾∈Rⁿ, x⁽ⁱ⁾It is in the sample set The location information of i sample, i=1,2 ..., m；The center point set that the K cluster centre point is constituted is { μ⁽¹⁾, μ⁽²⁾..., μ^(K), μ^(j)∈Rⁿ, μ^(j)For the location information of j-th of cluster centre point of the center point set, j=1,2 ..., K；RⁿFor n-dimensional vector space, n is the integer more than or equal to 1；According to formula c^(j)=arg min_j||x⁽ⁱ⁾-μ^(j)||², meter Calculate i-th of sample x⁽ⁱ⁾Affiliated class cluster c^(j)；Wherein, | | x⁽ⁱ⁾-μ^(j)||²For x⁽ⁱ⁾To μ^(j)Euclidean distance square, arg min_j||x⁽ⁱ⁾-μ^(j)||²To work as μ^(j)When for the cluster centre point, make to obtain μ^(j)The quadratic sum of Euclidean distance be minimized All independent variable x⁽ⁱ⁾Set；According to formulaUpdate each described cluster centre point μ^(j), directly To distortion functionConvergence；Wherein, c^(j)For class cluster j,For the class cluster The location information x of j⁽ⁱ⁾,For all location informations in the class cluster j feature and,For statistics The number of location information in the class cluster j,For the class cluster j each location information to described The Euclidean distance quadratic sum of the cluster centre point of class cluster j；After distortion function convergence, K obtained class cluster corresponds to the K A location information set.

The advantages of K-Means clustering algorithm provided in the embodiment of the present invention, is that calculating is easy, and algorithm is quick and easy, There is higher efficiency to large data sets, be suitble to excavate large-scale dataset and handle data-intensive class cluster.

In one possible implementation, classified according to K-Medians algorithm to the m location information, obtained May include following 4 steps to K location information set: 1) sample set that location information sample is constituted is { x⁽¹⁾, x⁽²⁾..., x^(m), x⁽ⁱ⁾∈Rⁿ；K cluster centre point is chosen from sample set.2) according to formula c^(j)=arg min_j||x⁽ⁱ⁾-μ^(j)||², calculate the generic c of each sample i^(j), i.e., the described location information sample to the location information sample institute The smallest classification of Euclidean distance of the cluster centre point of the class cluster of category.3) median for calculating each class cluster, determines each Center μ^(j).4) step 2), 3) is constantly repeated, until determining stable K class cluster and the corresponding cluster centre of the K class cluster Point.

The advantages of K-Medians clustering algorithm provided in the embodiment of the present invention, is to count using the median of data Central point is calculated, calculated result is avoided to be influenced by abnormal data.

In one possible implementation, according to mean shift clustering algorithm, the m location information is divided Class obtains K location information set, may include following 4 steps: 1) determining sliding window radius r, poly- with what is randomly selected Class central point C radius is that the round sliding window of r starts to slide.It is mobile to the higher region of density in each iteration, directly To convergence.2) new region is slided into each time, the point for calculating the mean value in sliding window as central point, in sliding window Quantity be window in density.In moving each time, window can be mobile towards the higher region of density.3) moving window, The density in central point and window in calculation window, until there is no direction that can accommodate more points in window, i.e., one It is directly moved in circle until density is not further added by.4) 3) step 1) is to that can generate many sliding windows, when multiple sliding windows When overlapping, retains the window comprising most multiple spot, then clustered according to the sliding window where data point, eventually form stabilization Window, i.e. center point set and corresponding grouping class cluster.

The advantages of mean shift clustering algorithm provided in the embodiment of the present invention, is, requires no knowledge about the number of class cluster Amount, the quantity of class cluster can be separated by the calculating of algorithm automatically；In calculating process, cluster centre can be poly- towards Dmax density Collection, is influenced smaller by data mean value.

In one possible implementation, according to greatest hope (EM) clustering algorithm with Gaussian Mixture (GMM) model, Classify to the m location information, obtain K location information set, may include following 4 steps: 1) selecting class cluster Quantity and each class cluster of random initializtion Gaussian Distribution Parameters (mean value and variance).Can also one first be provided according to data Relatively accurate mean value and variance.2) Gaussian Profile for giving each class cluster calculates each data point and belongs to the general of each class cluster Rate.One point may more belong to such cluster closer to the center of Gaussian Profile.3) joined based on these probability calculation Gaussian Profiles The maximization so that data point is counted, the weighting of data point probability can be used to calculate these new parameters, weight is exactly Data point belongs to the probability of such cluster.4) iteration 2) and 3) until the variation in iteration is little.

The advantages of greatest hope (EM) clustering algorithm of use Gaussian Mixture (GMM) model provided in the embodiment of the present invention It is, since GMMs uses mean value and standard deviation, identifiable class cluster shape can be ellipse, and be not limited to circle；By Probability is used in GMMs, a data point may belong to multiple clusters, improve the accuracy of calculating.

In one possible implementation, according to spectral clustering, classify to the m location information, obtain K A location information set may include following 4 steps: 1) a-th of sample and b-th of sample measurement are similar, i.e., Gauss is similar DegreeWherein σ is hyper parameter, and 1≤a≤m, 1≤b≤m, a ≠ b, a, b are integer. 2) similarity matrix W=S is formed_abM*m, symmetrical matrix, wherein S_aa1 should be equal to, but calculate all write as 0 for convenience, So similarity matrix reformed on leading diagonal be all 0 symmetrical matrix.3) a-th of sample is calculated to other all samples Similarity and d_a=S_a1+S_a2+…+S_a(m-1)(about S_aAddition, some will such as be polymerized to K class just and only will use preceding K S_a It is added or is set a threshold value, casts out the S less than threshold value_a)；The d in graph theory_aDegree of being called, it can be understood as connect the power on side Value.By the degree d of all the points_aComposition degree matrix D (diagonal matrix).4) Laplacian Matrix L=D-W is formed, L is symmetrical positive semidefinite square Battle array, minimal eigenvalue is 0, and corresponding feature vector is complete 1 vector.The characteristic value of L is arranged from small to large, λ₁...λ_m, corresponding Feature vector u₁...u_mIf requirement is polymerized to K class, the corresponding feature vector of K characteristic value before just taking forms matrix U_m* K, this Sample thinks that the feature of corresponding first sample is exactly u₁₁, u₁₂..., u_1K, the feature of second sample is exactly u₂₁, u₂₂..., u_2K, the feature of m-th of sample is exactly u_m1, u_m2..., u_mK, K mean value is done to this m sample, finally to the cluster of this m sample It as a result is exactly the cluster result of original position-information.

The advantages of spectral clustering provided in the embodiment of the present invention, is the La Pula based on graph theory to sample data The feature vector of this matrix is clustered, to reach the cluster to sample data；When data sample distribution is presented aspherical, It can identify and handle.

In one possible implementation, according to Dbscan clustering algorithm, classify to the m location information, K location information set is obtained, may include following 2 steps: 1) determining radius r and minPoints first；Do not have from one The arbitrary number strong point being accessed starts, by this point centered on, r be radius circle in include point quantity whether be greater than or Equal to minPoints, if it is greater than or be equal to minPoints, then change the time and be marked as central point, it is on the contrary then can be by Labeled as noise point.2) the step of repeating 1) is if a noise point is present in some central point In the circle of radius, then this point is marked as marginal point, otherwise is still noise point.Repeat step 1), it is known that all points All it is accessed.

The advantages of Dbscan clustering algorithm provided in the embodiment of the present invention, is Name-based Routing, does not need to know The quantity of road class cluster, can be with the class number of clusters amount of the data sample of automatically derived required processing by calculating.

In one possible implementation, according to the Agglomerative Hierarchical Clustering in hierarchical clustering algorithm, to the m position Information is classified, and K location information set is obtained, and may include following 3 steps: 1) each data point be considered as one it is single Cluster, then select the module of distance between two clusters of a measurement, calculate all individuals and the distance between individuals, look for One kind is polymerized to two nearest samples.2) regard microcommunity above as a new individual, then with remaining individual, Distance between all individuals and individual is calculated, two nearest individuals is looked for be polymerized to one kind, and so on.3) finally, until Until obtaining required class number of clusters amount.

The advantages of hierarchical clustering algorithm provided in the embodiment of the present invention is can be by being arranged different related ginsengs Numerical value obtains the level clustering structure on different grain size；In cluster vpg connection, hierarchical clustering is suitable for the poly- of arbitrary shape Class, and it is insensitive to the input sequence of sample.Selection for distance metric is simultaneously insensitive, and requires no knowledge about class The quantity of cluster subjective can divide class number of clusters amount.

Step S203: according to preset confidence level condition, confidence level highest in the K location information set is determined One set, wherein K is the integer greater than 1 and less than N；

Specifically, the preset confidence level condition may include being made with the location information quantity in location information set To judge the whether believable condition of location information set.The location information quantity using in location information set is as judging position The whether believable condition of information aggregate is set, can specifically include following three kinds of situations: 1. calculate in each location information set The ratio of location information quantity and m；Determine that the corresponding position information set of maximum ratio is combined into highest first collection of the confidence level It closes.2. calculating the ratio of the location information quantity and m in each location information set；Determine the ratio pair more than default value The position information set answered is combined into the highest first set of the confidence level.3. the location information in more each location information set Then population size sorts according to the sequence of quantity from big to small, the maximum corresponding position information set of access amount is combined into described The highest first set of confidence level.Wherein, the setting of the default value may include: by method user of service according to itself Practical experience setting；Alternatively, being set corresponding with data characteristics silent by server according to the data characteristics of the location information of acquisition Recognize numerical value.The cluster result of 7 kind clustering algorithms of the confidence level condition for illustrating in step S202 in this method step all may be used It is applicable in.

In one possible implementation, the location information quantity in each location information set is calculated respectively with m's Ratio；Determine that the corresponding position information set of maximum ratio is combined into the highest first set of the confidence level.Referring to Figure 16, Figure 16 It is the ratio of the location information quantity and itself and m in the location information set provided in an embodiment of the present invention according to Figure 15 statistics Value, as shown in figure 16, accounting is exactly the ratio of location information quantity and m in each location information set, compares the size of ratio, Determine that accounting isPosition information set be combined into the highest first set of the confidence level；Wherein, the numerical value of m is each position letter Location information quantity summation in breath set.

Step S204: according to the location information in the first set, the vehicle where the vehicle diagnostic equipment is determined The regional location of maintenance factory.

Specifically, the first set is the highest first set c of confidence level in K location information set^(j), the first collection The cluster centre point of conjunction is μ^(j), j=k, 0 < k≤K, and k is integer.

In one possible implementation, the first set c is determined^(j)With the cluster centre point of the first set μ^(j)Afterwards, with the cluster centre point μ^(j)For dot, radius is the border circular areas of default value, determine in the first set Location information in the border circular areas；According to the location information in the border circular areas, the area of the vehicle salvage shop is determined Domain position.7, Figure 17 is vehicle maintenance determining in the first set of Figure 16 screening provided in an embodiment of the present invention referring to Figure 1 The schematic diagram of factory's regional location, as shown in figure 17, the solid line border circular areas in the first set shown in dotted ellipse region is root It is determined according to the operation in above-mentioned possible implementation；Then, the vehicle is determined according to the location information in solid line border circular areas The regional location of maintenance factory.Wherein, the default value of radius be can be according to the experience of user or other reasonable manners Setting.

Step S205: it obtains the corresponding position of location information of the onboard diagnostic device in the first set and is adopted The vehicle-relevant data of collection.

Specifically, the vehicle-relevant data may include following 3 kinds of situations: vehicle diagnostic equipment is examined in 1. a period of times Disconnected vehicle fleet size；2. the frequency that vehicle diagnostic equipment uses in a period of time；3. the vehicle trouble of vehicle diagnostic equipment detection Problem statistics.Step S205 is to handle the vehicle-relevant data in the vehicle salvage shop after step S204 execution.Step S205 is a kind of optional step in the embodiment of the present invention.

It is referred in above-mentioned step S202, Fig. 3 is the table provided in an embodiment of the present invention for arranging and m location information being presented Lattice, as shown in figure 3, location information x⁽ⁱ⁾It is expressed as about longitudeLatitudeOrderly real number pairThis Inventive embodiments provide a kind of method for arranging and showing m location information in the table；Wherein, 0 < i≤m, i is whole Number.

Fig. 4 is the data point distribution figure of m location information provided in an embodiment of the present invention, as shown in figure 4, horizontal axis, the longitudinal axis The scale of mark only illustrates as the embodiment of the present invention and uses that the scale marked when specifically used is according to the specific data of acquisition Feature and choose.Location information x in figure⁽ⁱ⁾It is expressed as the data point of bivector, K=3, dot indicates location information data Point, crunode indicate the cluster centre point of each class cluster.Due to location information x⁽ⁱ⁾Storage form and Collator Mode all there is multiplicity Property, it is not limited to said combination and data information is handled.

Fig. 5 is data shown in Fig. 4 provided in an embodiment of the present invention by one time i-th sample x⁽ⁱ⁾Affiliated class cluster's It calculates and each described cluster centre pointCalculating classification results figure.And so on, Fig. 6-14 is by K mean algorithm Each circulation step after, the classification results figure of location information data point；Final classification results, referring to Figure 15, Tu15Shi M location information provided in an embodiment of the present invention passes through the final classification results figure of K mean cluster algorithm, as shown in figure 15, The position of cluster centre is different from its initial position, is the definitive result obtained by reasonable computation.

The embodiment of the present invention is classified according to location information of the clustering algorithm to a large amount of vehicle diagnostic equipments of acquisition, is mentioned Then high position precision goes out the highest first set of confidence level according to preset confidence level conditional filtering, reduce single error position The influence of confidence breath finally efficiently uses the target position information in set according to certain rule in the first aggregate, determines The regional location of vehicle salvage shop where vehicle diagnostic equipment.Implement the present invention, the credible of region position result can be enhanced Property, computation complexity is reduced, the processing of mass data is easy to.

It is above-mentioned to illustrate the method for the embodiment of the present invention, the relevant apparatus of the embodiment of the present invention is provided below.This Installation practice is also mainly illustrated by taking K-Means algorithm as an example, wherein the application of taxon is referred to other clusters and calculates The possibility implementation of method.

8, Figure 18 is the structural representation of vehicle diagnostic equipment data processing equipment provided in an embodiment of the present invention referring to Figure 1 Figure, the vehicle diagnostic equipment data processing equipment 18 may include: first acquisition unit 1801, taxon 1802, screening Unit 1803, determination unit 1804 and second acquisition unit 1805.Wherein,

First acquisition unit 1801, for obtaining m location information of vehicle diagnostic equipment；

Taxon 1802 obtains K position letter for classifying according to clustering algorithm to the m location information Breath set；

In one possible implementation, taxon 1802 are used for according to K-Means algorithm to the m position Information is classified, and is obtained K location information set, is referred to Fig. 5-15, may include: to constitute from the m location information Sample set in choose K cluster centre point；Wherein, the sample set is { x⁽¹⁾, x⁽²⁾..., x^(m), x⁽ⁱ⁾∈Rⁿ, x⁽ⁱ⁾For the location information of i-th of sample in the sample set, i=1,2 ..., m；During the K cluster centre point is constituted Heart point set is combined into { μ⁽¹⁾, μ⁽²⁾..., μ^(K), μ^(j)∈Rⁿ, μ^(j)For the position of j-th of cluster centre point of the center point set Information, j=1,2 ..., K；RⁿFor n-dimensional vector space, n is the integer more than or equal to 1；According to formula c^(j)=arg min_j ||x⁽ⁱ⁾-μ^(j)||², calculate i-th of sample x⁽ⁱ⁾Affiliated class cluster c^(j)；Wherein, | | x⁽ⁱ⁾-μ^(j)||²For x⁽ⁱ⁾To μ^(j)It is European away from From square, arg min_j||x⁽ⁱ⁾-μ^(j)||²To work as μ^(j)When for the cluster centre point, make to obtain μ^(j)Euclidean distance it is flat Side and all independent variable x being minimized⁽ⁱ⁾Set；According to formulaUpdate each described cluster Central point μ^(j), until distortion functionConvergence；Wherein, c^(j)For class cluster j,For the location information x of the class cluster j⁽ⁱ⁾,For the feature of all location informations in the class cluster j With,For the number for counting location information in the class cluster j,For the class cluster j Each location information to the class cluster j cluster centre point Euclidean distance quadratic sum；After distortion function convergence, obtain The K class cluster arrived corresponds to the K location information set.

In one possible implementation, taxon 1802 are used for according to K-Medians algorithm to the m position Confidence breath is classified, and K location information set is obtained, and the sample set that may include: 1) location information sample composition is { x⁽¹⁾, x⁽²⁾..., x^(m), x⁽ⁱ⁾∈Rⁿ；K cluster centre point is chosen from sample set.2) according to formula c^(j)=arg min_j| |x⁽ⁱ⁾-μ^(j)||², calculate the generic c of each sample i^(j), i.e., the described location information sample to the location information sample The smallest classification of Euclidean distance of the cluster centre point of affiliated class cluster.3) median for calculating each class cluster, determines each Center μ^(j).4) step 2), 3) is constantly repeated, until determining in stable K class cluster and the corresponding cluster of the K class cluster Heart point.

In one possible implementation, taxon 1802 are used for according to mean shift clustering algorithm, to the m A location information is classified, and K location information set is obtained, and may include: 1) to determine sliding window radius r, to select at random The cluster centre point C radius taken is that the round sliding window of r starts to slide；It is moved in each iteration to the higher region of density It is dynamic, until convergence.2) new region is slided into each time, calculates the mean value in sliding window as central point, sliding window The quantity of interior point is the density in window；In moving each time, window can be mobile towards the higher region of density.3) mobile Window, the density in central point and window in calculation window, until there is no direction that can accommodate more points in window, It is moved to always in circle until density is not further added by.4) 3) step 1) is to that can generate many sliding windows, when multiple slidings When windows overlay, retains the window comprising most multiple spot, then clustered, eventually formed according to the sliding window where data point Stable window, i.e. center point set and corresponding grouping class cluster.

In one possible implementation, taxon 1802, for according to the maximum with Gaussian Mixture (GMM) model It is expected that (EM) clustering algorithm, classifies to the m location information, K location information set is obtained, may include: 1) to select The quantity of class cluster and the Gaussian Distribution Parameters (mean value and variance) of each class cluster of random initializtion are selected, can also first be given according to data A relatively accurate mean value and variance out.2) Gaussian Profile for giving each class cluster calculates each data point and belongs to each class The probability of cluster；One point may more belong to such cluster closer to the center of Gaussian Profile.3) it is based on these probability calculations Gauss Distribution parameter makes the maximization of data point, and the weighting of data point probability can be used to calculate these new parameters, weigh It is again exactly the probability that data point belongs to such cluster.4) iteration 2) and 3) until the variation in iteration is little.

In one possible implementation, taxon 1802 are used for according to spectral clustering, to the m position Information is classified, and K location information set is obtained, and may include following 4 steps: 1) a-th of sample and b-th of sample degree Measure similar, i.e. Gauss similarityWherein σ is hyper parameter, 1≤a≤m, 1≤b≤m, A ≠ b, a, b are integer.2) similarity matrix W=S is formed_abM*m, symmetrical matrix, wherein S_aa1 should be equal to, but be Facilitate calculating all to be write as 0, thus similarity matrix reformed into all be on leading diagonal 0 symmetrical matrix.3) a-th of sample is calculated This to other all samples similarity and d_a=S_a1+S_a2+…+S_a(m-1)(about S_aAddition, some will such as be polymerized to K class It just only will use preceding K S_aIt is added or is set a threshold value, casts out the S less than threshold value_a)；The d in graph theory_aDegree of being called, can be with It is interpreted as the weight on connection side.By the degree d of all the points_aComposition degree matrix D (diagonal matrix).4) Laplacian Matrix L=D- is formed W, L are symmetric positive semidefinite matrix, and minimal eigenvalue is 0, and corresponding feature vector is complete 1 vector.From small to large the characteristic value of L Arrangement, λ₁...λ_m, character pair vector u₁...u_mIf require be polymerized to K class, just take before the corresponding feature of K characteristic value to Amount forms matrix U_m* K thinks that the feature of corresponding first sample is exactly u in this way₁₁, u₁₂..., u_1K, the spy of second sample Sign is exactly u₂₁, u₂₂..., u_2K, the feature of m-th of sample is exactly u_ml, u_m2..., u_mK, K mean value is done to this m sample, finally Cluster result to this m sample is exactly the cluster result of original position-information.

The advantages of spectral clustering provided in the embodiment of the present invention, is, passes through the drawing to sample data based on graph theory The feature vector of this matrix of pula is clustered, to reach the cluster to sample data；Aspheric is presented when data sample is distributed When shape, it can identify and handle.

In one possible implementation, taxon 1802 are used for according to Dbscan clustering algorithm, to the m Location information is classified, and K location information set is obtained, and may include: 1) to determine radius r and minPoints first；From one A not visited arbitrary number strong point starts, by this point centered on, r be radius circle in include point quantity whether Then change the time if it is greater than or equal to minPoints more than or equal to minPoints and be marked as central point, it is on the contrary Noise point can be then marked as.2) the step of repeating 1), if a noise point is present in some central Point is in the circle of radius, then this point is marked as marginal point, otherwise is still noise point.Repeat step 1), it is known that All points are all accessed.

In one possible implementation, taxon 1802, for according to the cohesion level in hierarchical clustering algorithm Cluster, classify to the m location information, obtain K location information set, may include: 1) each data point be considered as Then one single cluster selects the module of distance between two clusters of a measurement, calculates between all individuals and individual Distance, find two nearest samples and be polymerized to one kind.2) regard microcommunity above as a new individual, then with it is surplus Under individual, calculate it is all individual and individuals between distance, look for two nearest individuals to be polymerized to one kind, and so on.3) Until obtaining required class number of clusters amount.

Screening unit 1803, for determining confidence level in the K location information set according to preset confidence level condition Highest first set；

In one possible implementation, screening unit 1803, for calculating the position in each location information set Information content respectively with the ratio of m；Determine that the corresponding position information set of maximum ratio is combined into highest first collection of the confidence level It closes.

Determination unit 1804, for determining the vehicle diagnostic equipment institute according to the location information in the first set Vehicle salvage shop regional location.

In one possible implementation, determination unit 1804, for the screening unit execute corresponding contents it Afterwards, the screening module is c specifically for the highest first set of confidence level in K determining location information set^(j)And institute The cluster centre point for stating first set is μ^(j), j=k, 0 < k≤K, and k is integer；Determine the first set c^(j)With it is described The cluster centre point μ of first set^(j)Afterwards, with the cluster centre point μ^(j)For dot, radius is the border circular areas of default value, Determine the location information in the first set in the border circular areas；According to the location information in the border circular areas, really The regional location of the fixed vehicle salvage shop.

Second acquisition unit 1805, for obtaining location information pair of the onboard diagnostic device in the first set The position answered vehicle-relevant data collected.Second acquisition unit 1805 is a kind of optional unit in the embodiment of the present invention.

As shown in figure 19, Figure 19 is a kind of structural schematic diagram of equipment provided in an embodiment of the present invention.Vehicle diagnostic equipment Processing unit 18 can realize with the structure in Figure 19, which may include at least one storage unit 1901, at least One communication component 1902, at least one processing component 1903.In addition, the equipment can also include the general portions such as antenna, power supply Part, this will not be detailed here.

Storage unit 1901, can be read-only memory (read-only memory, ROM) or can store static information and The other kinds of static storage device of instruction, random access memory (random access memory, RAM) or can deposit The other kinds of dynamic memory for storing up information and instruction, is also possible to Electrically Erasable Programmable Read-Only Memory (Electrically Erasable Programmable Read-Only Memory, EEPROM), CD-ROM (Compact Disc Read-Only Memory, CD-ROM) or other optical disc storages, optical disc storage (may include compression optical disc, laser disc, Optical disc, Digital Versatile Disc, Blu-ray Disc etc.), magnetic disk storage medium or other magnetic storage apparatus or can be used in carrying Storage have instruction or data structure form desired program code and can by any other medium of computer access, But not limited to this.Memory, which can be, to be individually present, and is connected by bus with processor.Memory can also be with processor collection At together.

Communication component 1902, can be for other equipment or communication, such as upgrade server, cipher key service Device, equipment of vehicle-mounted inside etc..

Processing component 1903 can be general central processor (CPU), microprocessor, application-specific integrated circuit (application-specific integrated circuit, ASIC), or it is one or more for controlling above scheme journey The integrated circuit that sequence executes.

When equipment shown in Figure 19 is vehicle diagnostic equipment data processing equipment 18, processing component obtains vehicle diagnostic equipment M location information；Classified according to clustering algorithm to the m location information, obtains K location information set；According to Preset confidence level condition determines the highest first set of confidence level in the K location information set；According to first collection Location information in conjunction determines the regional location of the vehicle salvage shop where the vehicle diagnostic equipment.

It should be noted that vehicle diagnostic equipment data processing equipment 18 described in apparatus of the present invention embodiment is each The function of functional unit, reference can be made in embodiment of the method described in above-mentioned Fig. 2-17 vehicle diagnostic equipment data processing method phase Description is closed, details are not described herein again.

The embodiment of the present invention also provides a kind of computer storage medium, wherein the computer storage medium can be stored with journey Sequence, the program may include some or all of any one recorded in above method embodiment step when executing.

The embodiment of the present invention also provides a kind of computer program, which may include instruction, when the computer When program is computer-executed, computer is allowed to execute the part of any one including recording in above method embodiment Or Overall Steps.

In the above-described embodiments, it all emphasizes particularly on different fields to the description of each embodiment, there is no the portion being described in detail in some embodiment Point, reference can be made to the related descriptions of other embodiments.

It should be noted that for the various method embodiments described above, for simple description, therefore, it is stated as a series of Combination of actions, but those skilled in the art should understand that, the present invention is not limited by the sequence of acts described because According to the present invention, certain steps may can be performed in other orders or simultaneously.Secondly, those skilled in the art also should Know, the embodiments described in the specification are all preferred embodiments, related actions and modules not necessarily this hair Necessary to bright.

In several embodiments provided by the present invention, it should be understood that disclosed device, it can be by another way It realizes.For example, the apparatus embodiments described above are merely exemplary, such as the division of said units, it is only a kind of Logical function partition, there may be another division manner in actual implementation, such as multiple units or components can combine or can To be integrated into another system, or some features can be ignored or not executed.The unit of above-mentioned apparatus embodiment can be or It may not be and be physically separated, some or all of unit therein can be selected to realize this hair according to the actual needs The purpose of bright example scheme.

In addition, each functional unit in various embodiments of the present invention can integrate in one processing unit, it is also possible to Each unit physically exists alone, and can also be integrated in one unit with two or more units.Above-mentioned integrated unit Both it can take the form of hardware realization, can also realize in the form of software functional units.If above-mentioned integrated unit It is realized in the form of SFU software functional unit and when sold or used as an independent product, can store computer-readable at one It takes in storage medium.

Based on this understanding, technical solution of the present invention substantially in other words the part that contributes to existing technology or The all or part of person's technical solution can be embodied in the form of software products, which is stored in one In a storage medium, may include some instructions use so that computer equipment (can for personal computer, server or Person's network equipment etc. specifically can be the processor in computer equipment) execute the complete of each embodiment above method of the present invention Portion or part steps.Wherein, storage medium above-mentioned may include: USB flash disk, mobile hard disk, magnetic disk, CD, read-only memory (Read-Only Memory, abbreviation: ROM) or random access memory (Random Access Memory, abbreviation: RAM) Etc. the various media that can store program code.The above, the above embodiments are merely illustrative of the technical solutions of the present invention, and It is non-that it is limited；Although the present invention is described in detail referring to the foregoing embodiments, those skilled in the art answer Work as understanding: it is still possible to modify the technical solutions described in the foregoing embodiments, or special to part of technology Sign is equivalently replaced；And these are modified or replaceed, various embodiments of the present invention that it does not separate the essence of the corresponding technical solution The spirit and scope of technical solution.

Claims

1. a kind of data processing method characterized by comprising

M location information of vehicle diagnostic equipment is obtained, the m location information is that the vehicle diagnostic equipment is carrying out vehicle Itself geographical location information collected when diagnosis, wherein m is the integer greater than 2；

Classified according to clustering algorithm to the m location information, obtains K location information set；

According to preset confidence level condition, the highest first set of confidence level in the K location information set is determined, wherein K For the integer greater than 1 and less than N；

According to the location information in the first set, the region position of the vehicle salvage shop where the vehicle diagnostic equipment is determined It sets.

2. the method according to claim 1, wherein it is described according to clustering algorithm to the m location information into Row classification, obtains K location information set, comprising:

K cluster centre point is chosen from the sample set that the m location information is constituted；Wherein, the sample set is { x⁽¹⁾, x⁽²⁾..., x^(m), x⁽ⁱ⁾∈Rⁿ, x⁽ⁱ⁾For the location information of i-th of sample in the sample set, i=1,2 ..., m； The center point set that the K cluster centre point is constituted is { μ⁽¹⁾, μ⁽²⁾..., μ^(K), μ^(j)∈Rⁿ, μ^(j)For the central point Gather the location information of j-th of cluster centre point, j=1,2 ..., K；RⁿFor n-dimensional vector space, n is more than or equal to 1 Integer；

According to formula c^(j)=arg min_j||x⁽ⁱ⁾-μ^(j)||², calculate i-th of sample x⁽ⁱ⁾Affiliated class cluster c^(j)；Wherein, | | x⁽ⁱ⁾-μ^(j)||²For x⁽ⁱ⁾To μ^(j)Euclidean distance square, arg min_j||x⁽ⁱ⁾-μ^(j)||²To work as μ^(j)For the cluster centre When point, make to obtain μ^(j)Euclidean distance all independent variable x for being minimized of quadratic sum⁽ⁱ⁾Set；

According to formulaUpdate each described cluster centre point μ^(j), until distortion functionConvergence；Wherein, c^(j)For class cluster j,For the location information of the class cluster j x⁽ⁱ⁾,For all location informations in the class cluster j feature and,To count in the class cluster j The number of location information,For the class cluster j each location information to the class cluster j cluster The Euclidean distance quadratic sum of central point；

After distortion function convergence, K obtained class cluster corresponds to the K location information set.

3. according to the method described in claim 2, it is characterized in that, confidence level highest in the K location information set One collection is combined into c^(j)Cluster centre point with the first set is μ^(j), j=k, 0 < k≤K, and k is integer；It is described according to institute The location information in first set is stated, determines the regional location of the vehicle salvage shop where the vehicle diagnostic equipment, comprising:

Determine the first set c^(j)With the cluster centre point μ of the first set^(j)Afterwards, with the cluster centre point μ^(j)For Dot, radius are the border circular areas of default value, determine the location information in the first set in the border circular areas；

According to the location information in the border circular areas, the regional location of the vehicle salvage shop is determined.

4. method according to claim 1 to 3, which is characterized in that it is described according to preset confidence level condition, Determine the highest first set of confidence level in the K location information set, comprising:

Calculate location information quantity in each location information set respectively with the ratio of m；

Determine that the corresponding position information set of maximum ratio is combined into the highest first set of the confidence level.

5. method according to claim 1 to 3, which is characterized in that the determination vehicle diagnostic equipment institute After the regional location of the vehicle salvage shop of category, further includes:

It is related to obtain the corresponding position of location information of the onboard diagnostic device in first set vehicle collected Data.

6. a kind of data processing equipment characterized by comprising

First acquisition unit, for obtaining m location information of vehicle diagnostic equipment, the m location information is the vehicle Diagnostic device itself geographical location information collected when carrying out vehicle diagnostics, wherein m is the integer greater than 2；

Taxon obtains K location information set for classifying according to clustering algorithm to the m location information；

Screening unit determines in the K location information set confidence level highest for according to preset confidence level condition One set, wherein K is the integer greater than 1 and less than N；

Determination unit, for determining the vehicle where the vehicle diagnostic equipment according to the location information in the first set The regional location of maintenance factory.

7. device according to claim 6, which is characterized in that the taxon is specifically used for:

According to formulaUpdate each described cluster centre point μ^(j), until distortion functionConvergence；Wherein, c^(j)For class cluster j,Believe for the position of the class cluster j Cease x⁽ⁱ⁾,For all location informations in the class cluster j feature and,To count the class cluster j The number of middle location information,For each location information gathering to the class cluster j of the class cluster j The Euclidean distance quadratic sum of class central point；

8. device according to claim 7, which is characterized in that the screening unit is specifically used for:

The highest first set of confidence level is c in K determining location information set^(j)With the cluster centre of the first set Point is μ^(j), j=k, 0 < k≤K, and k is integer；

The determination unit is specifically used for:

Determine the first set c^(j)With the cluster centre point μ of the first set^(j)Afterwards, with the cluster centre point μ^(j)For Dot, radius are the border circular areas of default value, determine the location information in the first set in the border circular areas；Root According to the location information in the border circular areas, the regional location of the vehicle salvage shop is determined.

9. according to device described in claim 6-8 any one, which is characterized in that the screening unit is specifically used for:

Calculate location information quantity in each location information set respectively with the ratio of m；Determine the corresponding position of maximum ratio Information aggregate is the highest first set of the confidence level.

10. according to device described in claim 6-8 any one, which is characterized in that described device, further includes:

Second acquisition unit, for determining the region of vehicle salvage shop belonging to the vehicle diagnostic equipment in the determination unit After position, the corresponding position vehicle collected of location information of the onboard diagnostic device in the first set is obtained Related data.