CN108694486A - A kind of driving behavior intelligent Evaluation method and apparatus based on cloud model - Google Patents
A kind of driving behavior intelligent Evaluation method and apparatus based on cloud model Download PDFInfo
- Publication number
- CN108694486A CN108694486A CN201710225773.3A CN201710225773A CN108694486A CN 108694486 A CN108694486 A CN 108694486A CN 201710225773 A CN201710225773 A CN 201710225773A CN 108694486 A CN108694486 A CN 108694486A
- Authority
- CN
- China
- Prior art keywords
- evaluation
- factor
- driver
- driving
- cloud
- 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
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
- G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
- G06Q10/063—Operations research, analysis or management
- G06Q10/0639—Performance analysis of employees; Performance analysis of enterprise or organisation operations
- G06Q10/06393—Score-carding, benchmarking or key performance indicator [KPI] analysis
Landscapes
- Business, Economics & Management (AREA)
- Human Resources & Organizations (AREA)
- Engineering & Computer Science (AREA)
- Strategic Management (AREA)
- Development Economics (AREA)
- Economics (AREA)
- Entrepreneurship & Innovation (AREA)
- Educational Administration (AREA)
- Operations Research (AREA)
- Marketing (AREA)
- Game Theory and Decision Science (AREA)
- Quality & Reliability (AREA)
- Tourism & Hospitality (AREA)
- Physics & Mathematics (AREA)
- General Business, Economics & Management (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Traffic Control Systems (AREA)
Abstract
The driving behavior intelligent Evaluation method and apparatus based on cloud model that the invention discloses a kind of.The method includes:Acquire travelling data of multiple drivers in multiple driving tasks;It identifies specific driving behavior, and the index needed for factor of evaluation is extracted from collected travelling data;The corresponding evaluation content of factor of evaluation is set;According to the single of the given driver extracted driving task-based appraisal factor, the parameter of the first cloud model is calculated, and generate the first evaluations matrix;According to the first evaluations matrix, the single driving task to giving driver is evaluated;According to the multiple history of collected given driver or given vehicle driving task-based appraisal factor, the parameter of the second cloud model is calculated;According to the second evaluations matrix, using cloud computing, the history driving task to giving driver or given vehicle is evaluated.Method provided by the invention can effectively adapt to the uncertainty and fluctuation of travelling data so that evaluation result is more in line with actual conditions.
Description
Technical field
The present invention relates to technical field of intelligent traffic, more particularly to a kind of driving behavior intelligent Evaluation side based on cloud model
Method and device.
Background technology
China's car ownership gradually increases, and not only automobile market is rapidly developing, many and closely related row of automobile
Industry is also flourishing therewith.But this also brings many new problems to vehicle supervision department down to entire society simultaneously.For safety
Property exigent industry, for example armoured van, the taxi of night trip transport the lorry etc. of valuable cargo or hazardous chemical
Deng it is strong to be faced with administrative vehicle mobility, large number of, it is difficult to the thorny problem of management.First, manager wants to grasp
The real-time running state information of lorry, such as traffic route, current location and speed information, while also can be according to these information pair
Lorry and driver carry out specification.Secondly, for the driver of lorry, it is also desirable to the driving according to running information to oneself
Behavior is improved, to improve driving ability and service quality.In addition, client wishes in the logistics company of selection transport cargo
It hopes and understands the operation level of lorry used and the service scenario of company in advance.And what existing driving behavior evaluation method used
Evaluation parameter is mostly more subjective, and evaluation parameter is fixed, does not embody the uncertainty and fluctuation of data,
When travelling data amount is bigger, or there is (such as drive integral level and rise or fall) when certain Long-term change trend, weakens and comment
Ambiguity, randomness, the statistics of valence.Therefore, a set of adaptation travelling data is established, for the evaluating system for driving operating condition
It is graded and is managed to driving behavior, it is very necessary to realize that safety is gone on a journey, transported safely.
Invention content
In order to solve problems in the prior art, the driving behavior intelligence based on cloud model that an embodiment of the present invention provides a kind of
Evaluation method and device.The technical solution is as follows:
On the one hand, an embodiment of the present invention provides a kind of driving behavior intelligent Evaluation method based on cloud model, the side
Method includes:
Acquire travelling data of multiple drivers in multiple driving tasks;
It identifies specific driving behavior, and the index needed for factor of evaluation, the evaluation is extracted from collected travelling data
Factor is for evaluating driving behavior different in driver's driving procedure;
The corresponding evaluation content of factor of evaluation is set, and the evaluation content is used to reflect the value of driving behavior;
According to the single of the given driver extracted driving task-based appraisal factor, the parameter of the first cloud model is calculated, and
Generate the first evaluations matrix;
According to the first evaluations matrix, the single driving task to giving driver is evaluated;
According to the multiple history of collected given driver or given vehicle driving task-based appraisal factor, the second cloud is calculated
The parameter of model, the parameter of second cloud model is for generating the second evaluations matrix;
According to the second evaluations matrix, using cloud computing, the history driving task to giving driver or given vehicle carries out
Evaluation.
In the above-mentioned driving behavior intelligent Evaluation method based on cloud model of the embodiment of the present invention, the basis is extracted
Given driver single drive a vehicle task-based appraisal factor, calculate the first cloud model parameter, and generate the first evaluations matrix, packet
It includes:
The index needed for single driving task-based appraisal factor by the given driver got, by using average value as base
Standard is standardized;
Determine opinion rating and the corresponding factor of evaluation indication range of each rank;
It is corresponding to be calculated into factor of evaluation using backward cloud generator for index needed for factor of evaluation after standardization
The parameter of first cloud model, and generate the first cloud model;
Calculate the degree of membership of each grade belonging to the factor of evaluation index obtained;
It determines the weight of the corresponding factor of evaluation of each factor of evaluation index, and forms the first evaluations matrix.
It is described to be commented according to first in the above-mentioned driving behavior intelligent Evaluation method based on cloud model of the embodiment of the present invention
Valence matrix, the single driving task to giving driver are evaluated, including:
Commenting for the single driving task of given driver is obtained by the first evaluations matrix according to maximum membership grade principle
Valence result.
In the above-mentioned driving behavior intelligent Evaluation method based on cloud model of the embodiment of the present invention, the basis collects
Given driver or given vehicle multiple history drive a vehicle task-based appraisal factor, calculate the second cloud model parameter, including:
By the index needed for the multiple history of the given driver got or given vehicle driving task-based appraisal factor, lead to
It crosses and is standardized on the basis of average value;
By the index needed for each factor of evaluation after standardization, using backward cloud generator, calculate each evaluation because
The parameter of corresponding second cloud model of element.
It is described to be commented according to second in the above-mentioned driving behavior intelligent Evaluation method based on cloud model of the embodiment of the present invention
Valence matrix, using cloud computing, the history driving task to giving driver or given vehicle is evaluated, including:
By the parameter of calculated multiple second cloud models, synthetic evaluation matrix is obtained;
It determines the weight of each factor of evaluation, and the second evaluations matrix is obtained using cloud computing;
By the second evaluations matrix, the evaluation result of the history driving task of given driver or given vehicle is obtained.
On the other hand, an embodiment of the present invention provides a kind of driving behavior intelligent Evaluation device based on cloud model, it is described
Device includes:
Acquisition module, for acquiring travelling data of multiple drivers in multiple driving tasks;
It identifies extraction module, for identification specific driving behavior, and factor of evaluation institute is extracted from collected travelling data
The index needed, the factor of evaluation is for evaluating driving behavior different in driver's driving procedure;
Setup module, for the corresponding evaluation content of factor of evaluation to be arranged, the evaluation content is for reflecting driving behavior
Value;
First computing module calculates for driving a vehicle task-based appraisal factor according to the single of given driver extracted
The parameter of one cloud model, and generate the first evaluations matrix;
First processing module, for according to the first evaluations matrix, the single driving task to giving driver to be evaluated;
Second computing module, for being commented according to the multiple history of collected given driver or given vehicle driving task
Valence factor calculates the parameter of the second cloud model, and the parameter of second cloud model is for generating the second evaluations matrix;
Second processing module, for according to the second evaluations matrix, using cloud computing, to giving driver or given vehicle
History driving task is evaluated.
In the above-mentioned driving behavior intelligent Evaluation device based on cloud model of the embodiment of the present invention, described first calculates mould
Block includes:
First Standardisation Cell, for the finger needed for the single driving task-based appraisal factor by the given driver got
Mark, by being standardized on the basis of average value;
Determination unit, for determining opinion rating and the corresponding factor of evaluation indication range of each rank;
First computing unit is calculated for the index needed for the factor of evaluation after standardizing using backward cloud generator
Go out the parameter of corresponding first cloud model of factor of evaluation, and generates the first cloud model;
First computing unit is additionally operable to calculate the degree of membership of each grade belonging to the factor of evaluation index obtained;
First generation unit, the weight for determining the corresponding factor of evaluation of each factor of evaluation index, and form first
Evaluations matrix.
In the above-mentioned driving behavior intelligent Evaluation device based on cloud model of the embodiment of the present invention, the first processing mould
Block, including:
First processing units, for obtaining given driver's by the first evaluations matrix according to maximum membership grade principle
The evaluation result of single driving task.
In the above-mentioned driving behavior intelligent Evaluation device based on cloud model of the embodiment of the present invention, described second calculates mould
Block, including:
Second Standardisation Cell, for commenting the multiple history driving task of the given driver got or given vehicle
Index needed for valence factor, by being standardized on the basis of average value;
Second computing unit, for the index needed for each factor of evaluation after standardizing, using backward cloud generator,
Calculate the parameter of corresponding second cloud model of each factor of evaluation.
In the above-mentioned driving behavior intelligent Evaluation device based on cloud model of the embodiment of the present invention, the second processing mould
Block, including:
Second generation unit obtains synthetic evaluation matrix for the parameter by calculated multiple second cloud models;
Second generation unit is additionally operable to determine the weight of each factor of evaluation, and obtains second using cloud computing and comment
Valence matrix;
Second processing unit, for by the second evaluations matrix, obtaining the history driving of given driver or given vehicle
The evaluation result of task.
The advantageous effect that technical solution provided in an embodiment of the present invention is brought is:
By acquiring travelling data of multiple drivers in multiple driving tasks;Then specific driving behavior is identified, and
From the index needed for collected travelling data extraction factor of evaluation;The corresponding evaluation content of factor of evaluation is set;According to extraction
The single driving task-based appraisal factor of the given driver arrived, calculates the parameter of the first cloud model, and generate the first evaluations matrix;
According to the first evaluations matrix, the single driving task to giving driver is evaluated;According to collected given driver or
The multiple history driving task-based appraisal factor of given vehicle, calculates the parameter of the second cloud model;According to the second evaluations matrix, utilize
Cloud computing, the history driving task to giving driver or given vehicle are evaluated.It in this way should the driving row based on cloud model
The task that can not only drive a vehicle to the single for giving driver for intelligent Evaluation method carries out effective evaluation, can also be driven to given
The history of member or given vehicle driving task carries out overall merit, and during carrying out intelligent Evaluation, phase is carried out using cloud model
Pass is handled, and can effectively embody the uncertainty and fluctuation of data so that evaluation result is more in line with actual conditions, effectively carries
The high reference value of evaluation result.
Description of the drawings
To describe the technical solutions in the embodiments of the present invention more clearly, make required in being described below to embodiment
Attached drawing is briefly described, it should be apparent that, drawings in the following description are only some embodiments of the invention, for
For those of ordinary skill in the art, without creative efforts, other are can also be obtained according to these attached drawings
Attached drawing.
Fig. 1 is a kind of driving behavior intelligent Evaluation method flow diagram based on cloud model that the embodiment of the present invention one provides;
Fig. 2 is a kind of driving behavior intelligent Evaluation apparatus structure signal based on cloud model provided by Embodiment 2 of the present invention
Figure;
Fig. 3 is a kind of driving behavior intelligent Evaluation apparatus structure signal based on cloud model provided by Embodiment 2 of the present invention
Figure.
Specific implementation mode
To make the object, technical solutions and advantages of the present invention clearer, below in conjunction with attached drawing to embodiment party of the present invention
Formula is described in further detail.
Embodiment one
The driving behavior intelligent Evaluation method based on cloud model that an embodiment of the present invention provides a kind of, referring to Fig. 1, this method
May include:
Step S11 acquires travelling data of multiple drivers in multiple driving tasks.
In the present embodiment, can by onboard diagnostic system (On-Board Diagnostic, referred to as " OBD ") interface,
Or other onboard sensors obtain required initial data, these initial data may include:Wheelpath, engine turn
Speed, running speed, travel acceleration, transverse acceleration, oil consumption, brake information, throttle information, clutch information, running brake
Information, direction information and user's scoring etc..
With the development of the advanced information technology such as car networking, the data that people generate in driving behavior are more and more, more
To be more difficult to handle.The huge of motor-driven vehicle going that system acquisition arrives etc. can not only be made full use of using big data technology
Data more can provide strong technical support for driving behavior evaluation method by some specialized processing.
Step S12, identifies specific driving behavior, and the index needed for the collected travelling data extraction factor of evaluation,
Factor of evaluation is for evaluating driving behavior different in driver's driving procedure.
In the present embodiment, identify that specific driving behavior may include:It is related to sequential and sequence of movement, it is dynamic to driving behavior
It is matched and is judged, and then extracted specific driving behavior from driving data and such as overtake other vehicles, bypass the immediate leadership and hang shelves.Acquisition is commented
Index may include needed for valence factor:Preliminary treatment is carried out to data and calculates required index such as maximum speed to extract, it is average
Oil consumption etc..
Step S13, the corresponding evaluation content of setting factor of evaluation, evaluation content are used to reflect the value of driving behavior.
In the present embodiment, the factor of evaluation obtained according to step S12, to different evaluation index (i.e. factors of evaluation pair
The index answered) the corresponding evaluation content of setting.In conjunction with the feasibility of data acquisition and preliminary treatment system, in practical applications,
The evaluation index in terms of safety, economy and quality three, the practical and related working people of incorporation engineering can be respectively set
The experience of member chooses following factor of evaluation:
According to upper table it is found that the set of corresponding factor of evaluation can be expressed as CI, CIIAnd CIII.Remember that N indicates evaluation
The number of element in sets of factors, then can obtain NI=3, NII(safety)=10, NII(economy)=8, NII(quality)=3.Profit
With the method for layered queuing networks, the III level factor of evaluation in table can be used to the safety of driver's driving behavior, economical
Make evaluation respectively with quality, can also herein basis on provided using II grades of indexs as factor of evaluation under I grades of indexs
Driving behavior evaluation result.
Step S14 calculates the first cloud model according to the single of the given driver extracted driving task-based appraisal factor
Parameter, and the first evaluation square is generated, which may include:It is expected that, entropy, super entropy.
Specifically, above-mentioned steps S14 can be realized in the following way:
S141, the index needed for single driving task-based appraisal factor by the given driver got, by be averaged
It is standardized on the basis of value.
In the present embodiment, it to make that there is evaluable property between each index, needs first to transform to different evaluation points
Under same scale, by being standardized on the basis of average value, evaluation parameter is obtained.The specific method is as follows:
For factor of evaluation Cq, by big data obtain a large amount of drivers driving data corresponding with the factor of evaluation it
Afterwards, the universal average value of the factor is sought
Driver is x about one group of data (index i.e. needed for factor of evaluation) of this factorqi, i=(1,2 ...);Then withOn the basis of be standardized after be:
S142 determines opinion rating and the corresponding factor of evaluation indication range of each rank.
In the present embodiment, if being divided into P rank to the opinion rating of driving task, comprehensive big data uses clustering
Method determines that data (i.e. collected index) are grouped, and driving data is divided into P cluster using clustering method, each
Data in cluster have similar feature, and the specific method is as follows:
(1) it is equipped with n object, each object contains q feature, analyzes the data mode used matrix may be used and carry out table
Show, is shown below:
(2) it determines per a kind of central point bj=(bj1bj2…bjq), j=1,2 ... p, wherein p are center of a sample's point number.
(3) object function is determinedQi Zhong ||xj-bj||2For with bjCentered on point
Middle element xjDistance center point distance, is defined as with Euclidean distance
(4) x is adjustedjAffiliated group so that object function E (b1,…bk) minimum, obtain final Clustering.
Index needed for factor of evaluation after standardization is calculated factor of evaluation pair by S143 using backward cloud generator
The parameter for the first cloud model answered, and generate the first cloud model.
In the present embodiment, the concept of cloud model is, in opinion rating domain, arbitrary element x's in fuzzy set
The distribution of degree of membership referred to as (is subordinate to) cloud.By the definition of cloud it is found that with traditional fuzzy degree of membership unlike, it is a certain on domain
The degree of membership of point is not fixed, but with certain probability in slight change, cloud model just establishes first in fuzzy set
Mapping relations between element and degree of membership.
The numerical characteristic (i.e. the parameter of cloud model) of cloud can use tri- numerical value of desired Ex, entropy En and super entropy He to characterize,
It combines ambiguity and randomness constitutes qualitative and quantitative mutual mapping.Having determined that factor of evaluation and its classification
On the basis of standard, it can determine that 3 numerical characteristics of cloud model, specific algorithm are as follows according to following methods:
If factor of evaluation CqCorresponding statistical sample is xl, l=1,2 ...;
It is expected that Ex:According to xlSample is asked to be desired for
Entropy En:Reflect the uncertainty of qualitativing concept.In general, entropy is bigger, ambiguity and randomness are also bigger.The algorithm of entropy
It is as follows:
Super entropy He:It is the uncertain measurement of entropy, reflects the coherency in sample point uncertainty.The size and water dust of He
Dispersion, i.e. the randomness of degree of membership is associated.The computational methods of super entropy are:
Wherein, sample variance
According to fixed cloud model parameter Ex, En and He, with positive normal state cloud generator respectively to factor of evaluation
Index generates corresponding comprehensive cloud model.
According to the mathematical model of Membership Cloud Generators, σ is enabledx 2=He realizes that algorithm is as follows:
It enables and generates desired value x0=Ex, bandwidthNormal random number x;
X=N (x0,b2)
Calculate desired value x0, the degree of membership of the expectation curves of the normal state membership clouds of bandwidth b at x
Calculate pointThe variances sigma at placex;
Calculate separately left bandwidth bl, the wide br of right belt by drop half normal state rule variation.
Wherein
Calculate pointThe variances sigma at placex
Generating desired value isVariance is σxNormal random number μ;
μ=N (x0,b2)
Thus obtain water dust ε (x, μ).
S144 calculates the degree of membership of each grade belonging to the factor of evaluation index of acquisition.
In the present embodiment, for factor of evaluation Xi, for factor of evaluation Xi, according to determining water dust ε (x, μ), by specific
Data to be evaluated be input in water dust, degree of membership of the data to each rank can be generated.Recycling this operation can obtain entirely
Degree of membership of portion's data to P rank.It is organized into the synthetic evaluation matrix of following form:
S145 determines the weight of the corresponding factor of evaluation of each factor of evaluation index, and forms the first evaluations matrix.
In the present embodiment, the weight of each factor of evaluation is obtained using Principal Component Analysis, the specific method is as follows:
It seeks the related coefficient of each two factor, calculates between variable relevant related coefficient two-by-two
Wherein x, y indicate arbitrary 2 in n evaluation points respectively,Being averaged for two variable datas is indicated respectively
Value.By the X of gainedxyForm correlation matrix:
X={ Xxy}n×n
Seek characteristic root, feature vector and the variance contribution ratio of matrix X:
|λjEj-R|=0, j=1,2 ..., n
Wherein λjIndicate characteristic value, EjIndicate j rank unit matrixs, wjTabular form variance contribution ratio, that is, weight.
Finally obtain the weight matrix of following form:
A=[w1 w2 … wn]
In turn, the formula of the first evaluations matrix B is as follows:
Step S15, according to the first evaluations matrix, the single driving task to giving driver is evaluated.
In the present embodiment, above-mentioned steps S15 can be realized in the following way:According to maximum membership grade principle, pass through
First evaluations matrix obtains the evaluation result of the single driving task of given driver.
In the present embodiment, according to maximum membership grade principle, that in the first evaluations matrix B corresponding to maximum value is selected
Opinion rating is as final evaluation result.
It should be noted that by evaluating the single driving task for giving driver, can be driven in real time to each
The driving situation for the person of sailing effectively is supervised.
Step S16, according to the multiple history of collected given driver or given vehicle driving task-based appraisal factor, meter
The parameter of the second cloud model is calculated, the parameter of the second cloud model is for generating the second evaluations matrix.
In the present embodiment, above-mentioned steps S16 can be realized in the following way:
S161, by the finger needed for the multiple history of the given driver got or given vehicle driving task-based appraisal factor
Mark, by being standardized on the basis of average value.
It should be noted that standardisation process here is consistent with the standardisation process in step S14, do not repeating here.
Index needed for each factor of evaluation after standardization is calculated each comment by S162 using backward cloud generator
The parameter of corresponding second cloud model of valence factor.
In the present embodiment, by the index of obtained each factor of evaluation, second is obtained using backward cloud generator model
The parameter (Ex, En, He) of cloud model.First ask the evaluations matrix of one of factor, other factors method similar.Reverse cloud occurs
The specific algorithm of device is as follows:
If the statistical sample after standardization is xl, l=1,2 ... according to xlSample is asked it is expected:
Seek the single order absolute center centre-to-centre spacing of sample:
Seek sample variance:
Ask expectation:
Seek entropy:
Seek super entropy:
Step S17, according to the second evaluations matrix, using cloud computing, the history to giving driver or given vehicle is driven a vehicle
Task is evaluated.
Above-mentioned steps S17 can be realized in the following way:
S171 obtains synthetic evaluation matrix by the parameter of calculated multiple second cloud models.
Specifically, the synthetic evaluation matrix R of all factors is:
S172 determines the weight of each factor of evaluation, and obtains the second evaluations matrix using cloud computing.
Specifically, weight coefficient matrix A is:
Second evaluations matrix B=A*R, in conjunction with following cloud computing:
It obtains:
It finally obtains, B=[Ex(B),En(B),He(B)]T
Comparative evaluation is carried out if it is to k driver then:
The fuzzy overall evaluation achievement of each evaluation unit is obtained by above formula, and then can analyze selected driver more
Behavior evaluation result in secondary driving task.
S173 obtains the evaluation knot of the history driving task of given driver or given vehicle by the second evaluations matrix
Fruit.
The embodiment of the present invention is by acquiring travelling data of multiple drivers in multiple driving tasks;Then it identifies specific
Driving behavior, and extract the index needed for factor of evaluation from collected travelling data;It is arranged in the corresponding evaluation of factor of evaluation
Hold;It is driven a vehicle task-based appraisal factor according to the single of the given driver extracted, calculates the parameter of the first cloud model, and generate the
One evaluations matrix;According to the first evaluations matrix, the single driving task to giving driver is evaluated;It is given according to collected
The multiple history driving task-based appraisal factor for determining driver or given vehicle, calculates the parameter of the second cloud model;It is commented according to second
Valence matrix, using cloud computing, the history driving task to giving driver or given vehicle is evaluated.It should be based on cloud mould in this way
The driving behavior intelligent Evaluation method of type not only can carry out effective evaluation to the single driving task for giving driver, can be with
History driving task to giving driver or given vehicle carries out overall merit and utilizes cloud during carrying out intelligent Evaluation
Model carries out relevant treatment, can effectively embody the uncertainty and fluctuation of data so that evaluation result is more in line with reality
Situation effectively increases the reference value of evaluation result.
Embodiment two
An embodiment of the present invention provides a kind of driving behavior intelligent Evaluation device based on cloud model, realizes embodiment one
The driving behavior intelligent Evaluation method based on cloud model, referring to Fig. 2, which may include:Acquisition module 100 is known
Other extraction module 200, setup module 300, the first computing module 400, first processing module 500, the second computing module 600,
Two processing modules 700.
Acquisition module 100, for acquiring travelling data of multiple drivers in multiple driving tasks.
In the present embodiment, required original number can be obtained by vehicle-mounted OBD interfaces or other onboard sensors
According to these initial data may include:Wheelpath, engine speed, running speed, travel acceleration, transverse acceleration, oil
Consumption, brake information, throttle information, clutch information, running brake information, direction information and user's scoring etc..
With the development of the advanced information technology such as car networking, the data that people generate in driving behavior are more and more, more
To be more difficult to handle.The huge of motor-driven vehicle going that system acquisition arrives etc. can not only be made full use of using big data technology
Data more can provide strong technical support for driving behavior evaluation method by some specialized processing.
It identifies extraction module 200, for identification specific driving behavior, and factor of evaluation is extracted from collected travelling data
Required index, factor of evaluation is for evaluating driving behavior different in driver's driving procedure.
In the present embodiment, identify that specific driving behavior may include:It is related to sequential and sequence of movement, it is dynamic to driving behavior
It is matched and is judged, and then extracted specific driving behavior from driving data and such as overtake other vehicles, bypass the immediate leadership and hang shelves.Acquisition is commented
Index may include needed for valence factor:Preliminary treatment is carried out to data and calculates required index such as maximum speed to extract, it is average
Oil consumption etc..
Setup module 300, for the corresponding evaluation content of factor of evaluation to be arranged, evaluation content is for reflecting driving behavior
Value.
In the present embodiment, the factor of evaluation that can be obtained according to setup module 300, (comments different evaluation indexes
The corresponding index of valence factor) the corresponding evaluation content of setting.
First computing module 400 is calculated for the single driving task-based appraisal factor according to the given driver extracted
The parameter of first cloud model, and the first evaluation square is generated, which may include:It is expected that, entropy, super entropy.
First processing module 500, for according to the first evaluations matrix, the single driving task to giving driver to be commented
Valence.
In the present embodiment, it by evaluating the single driving task for giving driver, can be driven in real time to each
The driving situation for the person of sailing effectively is supervised.
Second computing module 600, for being appointed according to the driving of the multiple history of collected given driver or given vehicle
Business factor of evaluation calculates the parameter of the second cloud model, and the parameter of the second cloud model is for generating the second evaluations matrix.
Second processing module 700 is used for according to the second evaluations matrix, using cloud computing, to given driver or given vehicle
History driving task evaluated.
In the present embodiment, the multiple history driving task for giving driver or given vehicle is evaluated, it can be right
Entire driving situation has good integral monitoring.
Specifically, referring to Fig. 3, the first computing module 400 may include:
First Standardisation Cell 401, needed for the single driving task-based appraisal factor by the given driver got
Index, by being standardized on the basis of average value.
In the present embodiment, it to make that there is evaluable property between each index, needs first to transform to different evaluation points
Under same scale, by being standardized on the basis of average value, evaluation parameter is obtained.Specific method embodiment one has been retouched
It states.
Determination unit 402, for determining opinion rating and the corresponding factor of evaluation indication range of each rank.
In the present embodiment, if being divided into P rank to the opinion rating of driving task, comprehensive big data uses clustering
Method determines that data (i.e. collected index) are grouped, and driving data is divided into P cluster using clustering method, each
There is data in cluster similar feature, specific method embodiment one to have been described.
First computing unit 403, for the index needed for the factor of evaluation after standardizing, using backward cloud generator,
The parameter of corresponding first cloud model of factor of evaluation is calculated, and generates the first cloud model.
In the present embodiment, the concept of cloud model is, in opinion rating domain, arbitrary element x's in fuzzy set
The distribution of degree of membership referred to as (is subordinate to) cloud.By the definition of cloud it is found that with traditional fuzzy degree of membership unlike, it is a certain on domain
The degree of membership of point is not fixed, but with certain probability in slight change, cloud model just establishes first in fuzzy set
Mapping relations between element and degree of membership.
The numerical characteristic (i.e. the parameter of cloud model) of cloud can use tri- numerical value of desired Ex, entropy En and super entropy He to characterize,
It combines ambiguity and randomness constitutes qualitative and quantitative mutual mapping.Having determined that factor of evaluation and its classification
On the basis of standard, it can determine that 3 numerical characteristics of cloud model, specific method embodiment one have been retouched according to following methods
It states.
First computing unit 403 is additionally operable to calculate the degree of membership of each grade belonging to the factor of evaluation index obtained.
In the present embodiment, for factor of evaluation Xi, for factor of evaluation Xi, according to determining water dust ε (x, μ), by specific
Data to be evaluated be input in water dust, degree of membership of the data to each rank can be generated.Recycling this operation can obtain entirely
Degree of membership of portion's data to P rank.
First generation unit 404, the weight for determining the corresponding factor of evaluation of each factor of evaluation index, and form the
One evaluations matrix.
In the present embodiment, the weight of each factor of evaluation is obtained using Principal Component Analysis, in specific method embodiment one
It has described.
Further, referring to Fig. 3, first processing module 500 may include:
First processing units 501, for obtaining given driving by the first evaluations matrix according to maximum membership grade principle
The evaluation result of the single driving task of member.
In the present embodiment, according to maximum membership grade principle, that in the first evaluations matrix B corresponding to maximum value is selected
Opinion rating is as final evaluation result.
Specifically, referring to Fig. 3, the second computing module 600 may include:
Second Standardisation Cell 601, for appointing the multiple history driving of the given driver got or given vehicle
Index needed for factor of evaluation of being engaged in, by being standardized on the basis of average value.It has been described in specific method embodiment one.
Second computing unit 602 is occurred for the index needed for each factor of evaluation after standardizing using reverse cloud
Device calculates the parameter of corresponding second cloud model of each factor of evaluation.
In the present embodiment, by the index of obtained each factor of evaluation, second is obtained using backward cloud generator model
The parameter (Ex, En, He) of cloud model.First ask the evaluations matrix of one of factor, other factors method similar.Specific method is real
It applies and has been described in example one.
Further, referring to Fig. 3, Second processing module 700 may include:
Second generation unit 701 obtains overall merit square for the parameter by calculated multiple second cloud models
Battle array.
Second generation unit 701, the weight for determining each factor of evaluation, and obtain the second evaluation square using cloud computing
Battle array.
Second processing unit 702, for by the second evaluations matrix, obtaining the history row of given driver or given vehicle
The evaluation result of vehicle task.
The processing procedure of above-mentioned each unit has been described in embodiment one.
The embodiment of the present invention is by acquiring travelling data of multiple drivers in multiple driving tasks;Then it identifies specific
Driving behavior, and extract the index needed for factor of evaluation from collected travelling data;It is arranged in the corresponding evaluation of factor of evaluation
Hold;It is driven a vehicle task-based appraisal factor according to the single of the given driver extracted, calculates the parameter of the first cloud model, and generate the
One evaluations matrix;According to the first evaluations matrix, the single driving task to giving driver is evaluated;It is given according to collected
The multiple history driving task-based appraisal factor for determining driver or given vehicle, calculates the parameter of the second cloud model;It is commented according to second
Valence matrix, using cloud computing, the history driving task to giving driver or given vehicle is evaluated.It should be based on cloud mould in this way
The driving behavior intelligent Evaluation method of type not only can carry out effective evaluation to the single driving task for giving driver, can be with
History driving task to giving driver or given vehicle carries out overall merit and utilizes cloud during carrying out intelligent Evaluation
Model carries out relevant treatment, can effectively embody the uncertainty and fluctuation of data so that evaluation result is more in line with reality
Situation effectively increases the reference value of evaluation result.
The embodiments of the present invention are for illustration only, can not represent the quality of embodiment.
It should be noted that:The driving behavior intelligent Evaluation device based on cloud model that above-described embodiment provides is realizing base
When the driving behavior intelligent Evaluation method of cloud model, only the example of the division of the above functional modules, actually answer
In, it can be completed, i.e., divided the internal structure of equipment by different function modules as needed and by above-mentioned function distribution
At different function modules, to complete all or part of the functions described above.In addition, above-described embodiment provide based on cloud
The driving behavior intelligent Evaluation device of model belongs to same structure with the driving behavior intelligent Evaluation embodiment of the method based on cloud model
Think, specific implementation process refers to embodiment of the method, and which is not described herein again.
One of ordinary skill in the art will appreciate that realizing that all or part of step of above-described embodiment can pass through hardware
It completes, relevant hardware can also be instructed to complete by program, the program can be stored in a kind of computer-readable
In storage medium, storage medium mentioned above can be read-only memory, disk or CD etc..
The foregoing is merely presently preferred embodiments of the present invention, is not intended to limit the invention, it is all the present invention spirit and
Within principle, any modification, equivalent replacement, improvement and so on should all be included in the protection scope of the present invention.
Claims (10)
1. a kind of driving behavior intelligent Evaluation method based on cloud model, which is characterized in that the method includes:
Acquire travelling data of multiple drivers in multiple driving tasks;
It identifies specific driving behavior, and the index needed for factor of evaluation, the factor of evaluation is extracted from collected travelling data
For evaluating driving behavior different in driver's driving procedure;
The corresponding evaluation content of factor of evaluation is set, and the evaluation content is used to reflect the value of driving behavior;
According to the single of the given driver extracted driving task-based appraisal factor, the parameter of the first cloud model is calculated, and generate
First evaluations matrix;
According to the first evaluations matrix, the single driving task to giving driver is evaluated;
According to the multiple history of collected given driver or given vehicle driving task-based appraisal factor, the second cloud model is calculated
Parameter, the parameter of second cloud model is for generating the second evaluations matrix;
According to the second evaluations matrix, using cloud computing, the history driving task to giving driver or given vehicle is evaluated.
2. according to the method described in claim 1, it is characterized in that, the single driving for the given driver that the basis is extracted
Task-based appraisal factor calculates the parameter of the first cloud model, and generates the first evaluations matrix, including:
By the index needed for the single of the given driver got driving task-based appraisal factor, by the basis of average value into
Row standardization;
Determine opinion rating and the corresponding factor of evaluation indication range of each rank;
Index needed for factor of evaluation after standardization is calculated into factor of evaluation corresponding first using backward cloud generator
The parameter of cloud model, and generate the first cloud model;
Calculate the degree of membership of each grade belonging to the factor of evaluation index obtained;
It determines the weight of the corresponding factor of evaluation of each factor of evaluation index, and forms the first evaluations matrix.
3. according to the method described in claim 2, it is characterized in that, described according to the first evaluations matrix, to giving driver's
Single driving task is evaluated, including:
The evaluation knot of the single driving task of given driver is obtained by the first evaluations matrix according to maximum membership grade principle
Fruit.
4. according to the method described in claim 1, it is characterized in that, described according to collected given driver or given vehicle
Multiple history drive a vehicle task-based appraisal factor, calculate the second cloud model parameter, including:
By the multiple history of the given driver got or given vehicle drive a vehicle task-based appraisal factor needed for index, by with
It is standardized on the basis of average value;
Index needed for each factor of evaluation after standardization is calculated into each factor of evaluation pair using backward cloud generator
The parameter for the second cloud model answered.
5. according to the method described in claim 4, it is characterized in that, described according to the second evaluations matrix, using cloud computing, to giving
The history driving task for determining driver or given vehicle is evaluated, including:
By the parameter of calculated multiple second cloud models, synthetic evaluation matrix is obtained;
It determines the weight of each factor of evaluation, and the second evaluations matrix is obtained using cloud computing;
By the second evaluations matrix, the evaluation result of the history driving task of given driver or given vehicle is obtained.
6. a kind of driving behavior intelligent Evaluation device based on cloud model, which is characterized in that described device includes:
Acquisition module, for acquiring travelling data of multiple drivers in multiple driving tasks;
It identifies extraction module, for identification specific driving behavior, and is extracted needed for factor of evaluation from collected travelling data
Index, the factor of evaluation is for evaluating driving behavior different in driver's driving procedure;
Setup module, for the corresponding evaluation content of factor of evaluation to be arranged, the evaluation content is used to reflect the valence of driving behavior
Value;
First computing module calculates the first cloud for the single driving task-based appraisal factor according to the given driver extracted
The parameter of model, and generate the first evaluations matrix;
First processing module, for according to the first evaluations matrix, the single driving task to giving driver to be evaluated;
Second computing module, for according to the multiple history of collected given driver or given vehicle drive a vehicle task-based appraisal because
Element calculates the parameter of the second cloud model, and the parameter of second cloud model is for generating the second evaluations matrix;
Second processing module is used for according to the second evaluations matrix, using cloud computing, to giving the history of driver or given vehicle
Driving task is evaluated.
7. device according to claim 6, which is characterized in that first computing module includes:
First Standardisation Cell, the index for the single of the given driver got to be driven a vehicle needed for task-based appraisal factor,
By being standardized on the basis of average value;
Determination unit, for determining opinion rating and the corresponding factor of evaluation indication range of each rank;
First computing unit is calculated and is commented using backward cloud generator for the index needed for the factor of evaluation after standardizing
The parameter of corresponding first cloud model of valence factor, and generate the first cloud model;
First computing unit is additionally operable to calculate the degree of membership of each grade belonging to the factor of evaluation index obtained;
First generation unit, the weight for determining the corresponding factor of evaluation of each factor of evaluation index, and form the first evaluation
Matrix.
8. device according to claim 7, which is characterized in that the first processing module, including:
First processing units, for obtaining the single of given driver by the first evaluations matrix according to maximum membership grade principle
The evaluation result of driving task.
9. device according to claim 6, which is characterized in that second computing module, including:
Second Standardisation Cell, for by the multiple history of the given driver got or given vehicle driving task-based appraisal because
The required index of element, by being standardized on the basis of average value;
Second computing unit is calculated for the index needed for each factor of evaluation after standardizing using backward cloud generator
Go out the parameter of corresponding second cloud model of each factor of evaluation.
10. device according to claim 9, which is characterized in that the Second processing module, including:
Second generation unit obtains synthetic evaluation matrix for the parameter by calculated multiple second cloud models;
Second generation unit is additionally operable to determine the weight of each factor of evaluation, and obtains the second evaluation square using cloud computing
Battle array;
Second processing unit, for by the second evaluations matrix, obtaining the history driving task of given driver or given vehicle
Evaluation result.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710225773.3A CN108694486A (en) | 2017-04-07 | 2017-04-07 | A kind of driving behavior intelligent Evaluation method and apparatus based on cloud model |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710225773.3A CN108694486A (en) | 2017-04-07 | 2017-04-07 | A kind of driving behavior intelligent Evaluation method and apparatus based on cloud model |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN108694486A true CN108694486A (en) | 2018-10-23 |
Family
ID=63842167
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710225773.3A Pending CN108694486A (en) | 2017-04-07 | 2017-04-07 | A kind of driving behavior intelligent Evaluation method and apparatus based on cloud model |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108694486A (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110646007A (en) * | 2019-09-20 | 2020-01-03 | 中国科学技术大学 | Vehicle driving method based on formal representation |
| CN110705854A (en) * | 2019-09-20 | 2020-01-17 | 中国科学技术大学 | Driving level evaluation method and system |
| CN113128810A (en) * | 2019-12-31 | 2021-07-16 | 中移动信息技术有限公司 | Training method, device and equipment of scoring model and storage medium |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20140358430A1 (en) * | 2012-02-07 | 2014-12-04 | Xiamen King Long Motor Vehicle Inspection Co., Ltd | Driving evaluation system and method |
| CN105730450A (en) * | 2016-01-29 | 2016-07-06 | 北京荣之联科技股份有限公司 | Driving behavior analyzing method and evaluation system based on vehicle-mounted data |
| CN106022561A (en) * | 2016-05-05 | 2016-10-12 | 广州星唯信息科技有限公司 | Driving comprehensive evaluation method |
| CN106033001A (en) * | 2015-03-12 | 2016-10-19 | 比亚迪股份有限公司 | A vehicle fuel consumption evaluation method and device |
| CN106202872A (en) * | 2016-06-27 | 2016-12-07 | 江苏迪纳数字科技股份有限公司 | Vehicle driving behavior scoring method |
-
2017
- 2017-04-07 CN CN201710225773.3A patent/CN108694486A/en active Pending
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20140358430A1 (en) * | 2012-02-07 | 2014-12-04 | Xiamen King Long Motor Vehicle Inspection Co., Ltd | Driving evaluation system and method |
| CN106033001A (en) * | 2015-03-12 | 2016-10-19 | 比亚迪股份有限公司 | A vehicle fuel consumption evaluation method and device |
| CN105730450A (en) * | 2016-01-29 | 2016-07-06 | 北京荣之联科技股份有限公司 | Driving behavior analyzing method and evaluation system based on vehicle-mounted data |
| CN106022561A (en) * | 2016-05-05 | 2016-10-12 | 广州星唯信息科技有限公司 | Driving comprehensive evaluation method |
| CN106202872A (en) * | 2016-06-27 | 2016-12-07 | 江苏迪纳数字科技股份有限公司 | Vehicle driving behavior scoring method |
Non-Patent Citations (1)
| Title |
|---|
| 高洪波: "基于云模型的智能驾驶车辆变粒度测评研究", 《电子学报》 * |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110646007A (en) * | 2019-09-20 | 2020-01-03 | 中国科学技术大学 | Vehicle driving method based on formal representation |
| CN110705854A (en) * | 2019-09-20 | 2020-01-17 | 中国科学技术大学 | Driving level evaluation method and system |
| CN110646007B (en) * | 2019-09-20 | 2021-10-01 | 中国科学技术大学 | Vehicle driving method based on formal representation |
| CN113128810A (en) * | 2019-12-31 | 2021-07-16 | 中移动信息技术有限公司 | Training method, device and equipment of scoring model and storage medium |
| CN113128810B (en) * | 2019-12-31 | 2024-05-28 | 中移动信息技术有限公司 | Training method, device, equipment and storage medium of scoring model |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Yuhui et al. | Development of a representative driving cycle for urban buses based on the K-means cluster method | |
| CN110634080B (en) | Abnormal electricity utilization detection method, device, equipment and computer readable storage medium | |
| CN109840612A (en) | User's driving behavior analysis method and system | |
| CN109670970B (en) | Driving behavior scoring method and device and computer readable storage medium | |
| CN108513676A (en) | A kind of road conditions recognition methods, device and equipment | |
| CN104794184B (en) | A kind of illegal vehicle recognition methods of the Bayesian Classification Arithmetic based on large-scale data | |
| Abas et al. | Development of Malaysian urban drive cycle using vehicle and engine parameters | |
| CN102829980A (en) | Intelligent degree evaluation method of intelligent vehicle | |
| CN111222584A (en) | Lithium battery real-time evaluation method based on big data and deep neural network | |
| CN108694486A (en) | A kind of driving behavior intelligent Evaluation method and apparatus based on cloud model | |
| DE112018006555T5 (en) | PROGRAM CONTROLLED IDENTIFICATION OF A PERSONALITY OF AN AUTONOMOUS VEHICLE | |
| Vong et al. | Simultaneous-fault detection based on qualitative symptom descriptions for automotive engine diagnosis | |
| CN113155173A (en) | Perception performance evaluation method and device, electronic device and storage medium | |
| CN107437147A (en) | Reduce the vehicle travel risk dynamic assessment method and its system of freight logistics scene | |
| DE112021004024T5 (en) | APPROPRIATE DETECTION AND LOCATION OF ANOMALIES | |
| Chen et al. | Driving behaviors analysis based on feature selection and statistical approach: a preliminary study | |
| Fridelin et al. | Implementation of microservice architectures on SEMAR extension for air quality monitoring | |
| Wu et al. | Clustering of several typical behavioral characteristics of commercial vehicle drivers based on GPS data mining: Case study of highways in China | |
| CN117036060A (en) | Vehicle insurance fraud recognition method, device and storage medium | |
| Vdovic et al. | Eco-efficient driving pattern evaluation for sustainable road transport based on contextually enriched automotive data | |
| Wang et al. | Changing lane probability estimating model based on neural network | |
| CN114372680A (en) | Spatial crowdsourcing task allocation method based on worker loss prediction | |
| CN111079940B (en) | Decision tree model establishing method and using method for real-time fake-licensed car analysis | |
| Carrillo et al. | Design of efficient deep learning models for determining road surface condition from roadside camera images and weather data | |
| Silva et al. | Towards a customized vehicular maintenance based on 2-layers data-stream application |
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 | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20181023 |