CN109727334A - Recognition methods, device and the vehicle of landform locating for vehicle - Google Patents
Recognition methods, device and the vehicle of landform locating for vehicle Download PDFInfo
- Publication number
- CN109727334A CN109727334A CN201711040120.4A CN201711040120A CN109727334A CN 109727334 A CN109727334 A CN 109727334A CN 201711040120 A CN201711040120 A CN 201711040120A CN 109727334 A CN109727334 A CN 109727334A
- Authority
- CN
- China
- Prior art keywords
- landform
- vehicle parameter
- vehicle
- parameter
- terrain
- 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.)
- Granted
Links
Abstract
The present invention provides a kind of recognition methods of landform locating for vehicle, device and vehicles, this method comprises: obtaining the first driving parameter group that vehicle is sailed in current landform uplink, which includes at least one of steering wheel angle, steering wheel angular speed, longitudinal acceleration, side acceleration and yaw rate vehicle parameter；According to the first driving parameter group, and multiple landform identification section set of different terrain is corresponded respectively to, determines that the current landform belongs to the terrain category probability of each landform in different terrain；The terrain category probability for belonging to each landform in different terrain according to the current landform determines landform belonging to the current landform.Method of the present invention can identify landform locating for vehicle by acting the calculating of relevant vehicle parameter and the precise classification of processing and the landform being likely to be to vehicle to vehicle part collected in vehicle driving and vehicle.
Description
Technical field
This disclosure relates to automatic Pilot technical field, in particular to the recognition methods of landform locating for a kind of vehicle, device and
Vehicle.
Background technique
With the rise of artificial intelligence technology, more and more automobile vendors are provided with Function for Automatic Pilot in the car,
Wherein, the servomechanism based on landform is used for landform or weather environment according to vehicle current shape, selects different
Automatic Pilot strategy.In the prior art, the servomechanism based on landform generally uses contactless inductor, for example,
Visual response device, collects the relevant image information of landform, then by extracting color in these image informations and texture to corresponding
Landform classified and identified.But the image information collected by visual response device is easy by outside environmental elements, example
Such as, the influence of light status and weather in turn results in the situation of classification of landform and identification inaccuracy.
Summary of the invention
In view of this, the present invention is directed to propose a kind of recognition methods of landform locating for vehicle, device and vehicle, outer to reduce
The influence that boundary's environmental factor identifies landform improves the accuracy of landform identification.
In order to achieve the above objectives, the technical scheme of the present invention is realized as follows:
A kind of recognition methods of landform locating for vehicle, which comprises
The first driving parameter group that vehicle is sailed in current landform uplink is obtained, the first driving parameter group includes steering wheel
At least one of corner, steering wheel angular speed, longitudinal acceleration, side acceleration and yaw rate vehicle parameter；
According to the first driving parameter group, and multiple landform identification section set of different terrain is corresponded respectively to,
Determine that the current landform belongs to the terrain category probability of each landform in the different terrain；
The terrain category probability for belonging to each landform in the different terrain according to the current landform determines described current
Landform belonging to landform.
Further, vehicle is being obtained before the first driving parameter group that current landform uplink is sailed, the method is also wrapped
It includes:
Collect multiple groups the second vehicle parameter group that vehicle travels in different terrain respectively, second vehicle parameter of multiple groups
The vehicle parameter that the type for the vehicle parameter that every group of vehicle parameter group in group is included and the first driving parameter group include
Type it is consistent；
According to the second vehicle parameter of multiple groups group, the corresponding terrain information projection matrix of every kind of vehicle parameter is determined；
For any landform in the different terrain, according to the corresponding terrain information projection matrix of every kind of vehicle parameter,
Obtain every kind of vehicle parameter corresponding landform identification section under the landform, wherein whole vehicle parameters pair under the landform
The landform corresponding landform identification section set is formed between the whole landform cog regions answered.
Further, described according to the second vehicle parameter of multiple groups group, determine the corresponding landform letter of every kind of vehicle parameter
Cease projection matrix, comprising:
For any vehicle parameter obtained in any landform in the different terrain:
The peak value and valley that the vehicle parameter includes are obtained, to form the peak value array and valley array of the vehicle parameter；
According to the peak value array and the valley array, the first peak average value and the first paddy of the vehicle parameter are calculated
It is worth average value；
It is average according to the peak value array, the valley array, first peak average value and first valley
Value constructs the first driving parameter matrix of the vehicle parameter；
According to first peak average value, the first valley average value and the first driving parameter matrix, obtain
Take the discrete matrix of the first driving parameter matrix；
After obtaining the corresponding discrete matrix of the vehicle parameter under every kind of landform, according to whole discrete squares of acquisition
Battle array, determines the corresponding terrain information projection matrix of the vehicle parameter.
Further, the first landform is any landform in the different terrain, the first vehicle parameter is described second
Any vehicle parameter in vehicle parameter group is obtained according to the corresponding terrain information projection matrix of the first described vehicle parameter
Take the first described vehicle parameter corresponding landform identification section under first landform, comprising:
According to the corresponding terrain information projection matrix of the first described vehicle parameter, determine that the first described vehicle parameter exists
The first projection information under first landform, wherein first projection information is the landform of the first vehicle parameter
The product of the transposed matrix of information projection matrix the first driving parameter matrix corresponding with the first described vehicle parameter；
Know using the maximum value of first projection information, minimum value, average value, variance and adjustment average as landform
The input of other formula identifies that data, the landform identify formula to obtain the first landform of the first vehicle parameter are as follows:
Wherein, Sig1iIndicate that the first landform of i-th kind of vehicle parameter in the second vehicle parameter group obtained in landform 1 is known
Other data, Y1imaxIndicate the maximum value of the first projection information of i-th kind of vehicle parameter, Y1iminIndicate i-th kind of driving
The minimum value of first projection information of parameter, SDy1iIndicate the variance of the first projection information of i-th kind of vehicle parameter,
μy1i_adjIndicate the adjustment average of the first projection information of i-th kind of vehicle parameter；Work as Y1imaxTake the first described driving
The maximum value of first projection information of parameter, Y1iminThe minimum value of first projection information of the first vehicle parameter described in taking,
SDy1iTake variance, the μ of the first projection information of the first vehicle parametery1i_adjTake the first of the first vehicle parameter
When the adjustment average of projection information, Sig1iData are identified for the first landform of the first vehicle parameter；
Data are identified according to the first landform of the first vehicle parameter, determine the first described vehicle parameter described
Corresponding landform identifies section under first landform.
Further, described according to the first driving parameter group, and correspond respectively to multiple landform of different terrain
It identifies section set, determines that the current landform belongs to the terrain category probability of each landform in the different terrain, comprising:
Obtain the corresponding terrain information projection matrix of every kind of vehicle parameter in the first driving parameter group；
Institute is determined according to every kind of vehicle parameter and the corresponding terrain information projection matrix of every kind of vehicle parameter
State every kind of vehicle parameter corresponding second landform identification data under the current landform；
By every kind of vehicle parameter, corresponding second landform identifies data and described more under the current landform
A landform identification section set, determines the terrain category probability.
Further, described to be thrown according to every kind of vehicle parameter and the corresponding terrain information of every kind of vehicle parameter
Shadow matrix determines every kind of vehicle parameter corresponding second landform identification data under the current landform, comprising:
Any vehicle parameter being directed in the first driving parameter group:
The peak value and valley that the vehicle parameter includes are obtained, to form the peak value array and valley array of the vehicle parameter；
According to the peak value array and the valley array, the second peak average value and the second paddy of the vehicle parameter are calculated
It is worth average value；
It is average according to the peak value array, the valley array, second peak average value and second valley
Value, constructs the second vehicle parameter matrix of the vehicle parameter；
According to the second vehicle parameter matrix terrain information projection matrix corresponding with the vehicle parameter, the driving is obtained
Second projection information of the parameter under the current landform；
Know using the maximum value of second projection information, minimum value, average value, variance and adjustment average as landform
The input of other formula identifies data, the landform identification to obtain second landform of the vehicle parameter under the current landform
Formula are as follows:
Wherein, SigxiIndicate that the second landform of i-th kind of vehicle parameter in the obtain on landform x first driving parameter group is known
Other data, YximaxIndicate the maximum value of the second projection information of i-th kind of vehicle parameter, YximinIndicate i-th kind of driving
The minimum value of second projection information of parameter, SDyxiIndicate the variance of the second projection information of i-th kind of vehicle parameter,
μyxi_adjIndicate the adjustment average of the second projection information of i-th kind of vehicle parameter；Work as YximaxTake the of the vehicle parameter
The maximum value of two projection informations, YximinTake minimum value, the SD of the second projection information of the vehicle parameteryxiTake the vehicle parameter
The variance of second projection information, μyxi_adjWhen taking the adjustment average of the second projection information of the vehicle parameter, Sig1iFor the row
Second landform of vehicle parameter identifies data.
Further, landform cog region has been respectively corresponded under each landform of the every kind of vehicle parameter in the different terrain
Between, the landform identification Interval Set of the landform is formed between the corresponding whole landform cog regions of whole vehicle parameters under same landform
It closes,
By every kind of vehicle parameter, corresponding second landform identifies data and described more under the current landform
A landform identification section set, determines the terrain category probability, comprising:
Any vehicle parameter being directed in the first driving parameter group, determines the vehicle parameter in the current position
Whether the second landform identification data under shape, which belong to the vehicle parameter corresponding any landform under the different terrain, is known
Other section；
When the vehicle parameter under the current landform the second landform identification data belong to the vehicle parameter it is described not
The similar general of the corresponding landform in section is identified with the landform belonging to when corresponding any landform identification section, determining under landform
Rate is 1, and determines that the likelihood probability of the corresponding landform in other landform identification section in addition to affiliated landform identifies section is 0；
Alternatively,
When second landform identification data of the vehicle parameter under the current landform are not belonging to the vehicle parameter described
When under different terrain within corresponding any landform identification section, data and the landform are identified according to second landform
The distance for identifying each of section set landform identification section, calculates the second landform identification data and the vehicle parameter
The likelihood probability of the corresponding landform in corresponding each landform identification section under the different terrain；
After obtaining the corresponding whole likelihood probabilities of whole vehicle parameters that the first driving parameter group includes, it will correspond to
The average value of the corresponding whole likelihood probabilities of whole vehicle parameters of same landform is as described current in the different terrain
Landform belongs to the terrain category probability of the landform.
Further, the terrain category for belonging to each landform in the different terrain according to the current landform is general
Rate determines landform belonging to the current landform, comprising:
The maximum landform of terrain category probability value is determined as landform belonging to the current landform.
Compared with the existing technology, the recognition methods of landform locating for vehicle of the present invention has the advantage that
(1) method of the present invention can obtain the first driving parameter group that vehicle is sailed in current landform uplink, this
One driving parameter group includes in steering wheel angle, steering wheel angular speed, longitudinal acceleration, side acceleration and yaw rate
At least one vehicle parameter；According to the first driving parameter group, and correspond respectively to multiple landform identification of different terrain
Section set, determines that the current landform belongs to the terrain category probability of each landform in different terrain；According to the current landform category
The terrain category probability of each landform in different terrain, determines landform belonging to the current landform.It can be by vehicle row
Collected vehicle part and vehicle act the calculating of relevant vehicle parameter and processing and are likely to be to vehicle in sailing
The classification of landform identifies that the present invention provides a kind of methods of new identification landform, without adopting to landform locating for vehicle
Collect Image Acquisition, avoids environmental factor on influence caused by landform identification, recognition result is more acurrate, is conducive to further preferably
Vehicle is controlled, while also improving the intelligence degree of vehicle.
(2) method of the present invention can collect the driving ginseng of multiple groups second that vehicle travels in different terrain respectively
Array, the type for the vehicle parameter that every group of vehicle parameter group in multiple groups the second vehicle parameter group is included and first driving
The type for the vehicle parameter that parameter group includes is consistent；According to multiple groups the second vehicle parameter group, determine that every kind of vehicle parameter is corresponding
Terrain information projection matrix；For any landform in different terrain, thrown according to the corresponding terrain information of every kind of vehicle parameter
Shadow matrix obtains every kind of vehicle parameter corresponding landform under the landform and identifies section, wherein whole drivings under the landform
The landform corresponding landform identification section set is formed between the corresponding whole landform cog regions of parameter.It can be big by what is be collected into
The vehicle part and vehicle measured from different terrain act relevant vehicle parameter, obtain the various rows shown in different terrain
The characteristic of vehicle parameter, for the classification and identification to vehicle landform.
Another object of the present invention is to propose a kind of identification device of landform locating for vehicle, to reduce outside environmental elements
Influence to the Function for Automatic Pilot based on landform.
In order to achieve the above objectives, the technical scheme of the present invention is realized as follows:
A kind of identification device of landform locating for vehicle, described device include:
Parameter acquisition module, the first driving parameter group sailed for obtaining vehicle in current landform uplink, the first row
Vehicle parameter group include in steering wheel angle, steering wheel angular speed, longitudinal acceleration, side acceleration and yaw rate extremely
A kind of few vehicle parameter；
Probability determination module is used for according to the first driving parameter group, and corresponds respectively to the multiple of different terrain
Landform identifies section set, determines that the current landform belongs to the terrain category probability of each landform in the different terrain；
Landform determining module, for belonging to the terrain category of each landform in the different terrain according to the current landform
Probability determines landform belonging to the current landform.
Compared with the existing technology, the identification device of landform locating for vehicle of the present invention has the advantage that
Device of the present invention can obtain the first driving parameter group that vehicle is sailed in current landform uplink, this first
Vehicle parameter group includes in steering wheel angle, steering wheel angular speed, longitudinal acceleration, side acceleration and yaw rate
At least one vehicle parameter；According to the first driving parameter group, and correspond respectively to multiple landform cog regions of different terrain
Between gather, determine that the current landform belongs to the terrain category probability of each landform in different terrain；Belonged to according to the current landform
The terrain category probability of each landform in different terrain, determines landform belonging to the current landform.It can be by vehicle driving
In collected vehicle part and vehicle act calculating and processing and the ground being likely to be to vehicle of relevant vehicle parameter
The classification of shape identifies that the present invention avoids environmental factor from knowing landform without acquiring Image Acquisition to landform locating for vehicle
It is influenced caused by not, recognition result is more acurrate, is conducive to further preferably control vehicle, while also improving vehicle
Intelligence degree.
Another object of the present invention is to propose that a kind of vehicle, the vehicle include the identification of landform locating for above-mentioned vehicle
Device, to reduce influence of the outside environmental elements to the Function for Automatic Pilot based on landform.
Detailed description of the invention
The attached drawing for constituting a part of the invention is used to provide further understanding of the present invention, schematic reality of the invention
It applies example and its explanation is used to explain the present invention, do not constitute improper limitations of the present invention.In the accompanying drawings:
Fig. 1 is the flow chart of the recognition methods of landform locating for a kind of vehicle described in the embodiment of the present invention；
Fig. 2 is the flow chart of the recognition methods of the landform according to locating for another vehicle shown in embodiment illustrated in fig. 1；
Fig. 3 is the flow chart that method is determined according to a kind of topographic projection's matrix shown in embodiment illustrated in fig. 2；
Fig. 4 is to identify that the flow chart of the method for determination is gathered in section according to a kind of landform shown in embodiment illustrated in fig. 2；
Fig. 5 is the flow chart according to a kind of terrain category probability determination method shown in embodiment illustrated in fig. 1；
Fig. 6 is the flow chart that data capture method is identified according to a kind of landform shown in embodiment illustrated in fig. 5；
Fig. 7 is the flow chart according to a kind of terrain category method for calculating probability shown in embodiment illustrated in fig. 5；
Fig. 8 is the block diagram of the identification device of landform locating for a kind of vehicle described in the embodiment of the present invention；
Fig. 9 is the block diagram of the identification device of the landform according to locating for another vehicle shown in embodiment illustrated in fig. 8；
Figure 10 is the block diagram according to a kind of probability determination module shown in embodiment illustrated in fig. 8.
Specific embodiment
It should be noted that in the absence of conflict, the feature in embodiment and embodiment in the present invention can phase
Mutually combination.
The present invention will be described in detail below with reference to the accompanying drawings and embodiments.
Fig. 1 should referring to Fig. 1 for the flow chart of the recognition methods of landform locating for a kind of vehicle described in the embodiment of the present invention
The recognition methods of landform locating for vehicle may comprise steps of:
In a step 101, the first driving parameter group that vehicle is sailed in current landform uplink is obtained.
Wherein, which includes steering wheel angle, steering wheel angular speed, longitudinal acceleration, lateral acceleration
At least one of degree and yaw rate vehicle parameter.
In a step 102, according to the first driving parameter group, and multiple landform identification of different terrain is corresponded respectively to
Section set, determines that the current landform belongs to the terrain category probability of each landform in aforementioned different terrain.
Illustratively, the different terrain in the embodiment of the present disclosure includes the extreme landform that can be explicitly described, for example, sandstone
Landform, muddy landform, accumulated snow landform etc., or can be the more complicated landform that a variety of extreme landform combine.
Wherein, correspond respectively to different terrain multiple landform identification section set can according to different terrain characteristic
The landform identification section set being artificially arranged, or, it is preferable that it can be for after a large amount of landform related datas be trained
The set of class interval in classification of landform model.The training process can be illustrated in subsequent steps, and vehicle needs
Trained classification of landform model is received and saved after the training.It should be noted that each landform cog region
Between identify that section, each landform identification section correspond to a kind of driving ginseng acquired in a kind of landform comprising multiple landform in set
Number.When calculating the terrain category, can first calculate certain vehicle parameter under current landform with it is same under each landform
The likelihood probability of vehicle parameter, then integrate the likelihood probability of multiple vehicle parameters, obtain the current landform belong to it is aforementioned differently
The terrain category probability of each landform in shape.
In step 103, the terrain category probability of each landform in aforementioned different terrain is belonged to according to the current landform, really
Landform belonging to the fixed current landform.
Illustratively, it is with the maximum landform class which, which may include: landform belonging to the determining current landform,
The landform of other probability.
In conclusion the recognition methods of landform locating for the vehicle that the disclosure provides, can obtain vehicle in current landform
Traveling first driving parameter group, this first driving parameter group include steering wheel angle, steering wheel angular speed, longitudinal acceleration,
At least one of side acceleration and yaw rate vehicle parameter；It is and right respectively according to the first driving parameter group
It should identify that section is gathered in multiple landform of different terrain, determine that the current landform belongs to the landform of each landform in different terrain
Class probability；The terrain category probability for belonging to each landform in different terrain according to the current landform determines the current landform institute
The landform of category.Can by vehicle part collected in vehicle driving and vehicle are acted relevant vehicle parameter calculating and
The precise classification of processing and the landform being likely to be to vehicle identifies, the present invention is without adopting to landform locating for vehicle
Collect Image Acquisition, avoids environmental factor on influence caused by landform identification, recognition result is more acurrate, is conducive to further preferably
Vehicle is controlled, while also improving the intelligence degree of vehicle.
Fig. 2 is the flow chart of the recognition methods of the landform according to locating for another vehicle shown in embodiment illustrated in fig. 1, is such as schemed
Shown in 2, vehicle is being obtained before the first driving parameter group that current landform uplink is sailed, the recognition methods of landform locating for the vehicle
It can also include the following steps.
At step 104, multiple groups the second vehicle parameter group that vehicle travels in different terrain respectively is collected.
Wherein, the type for the vehicle parameter that every group of vehicle parameter group in aforementioned the second vehicle parameter of multiple groups group is included with
The type for the vehicle parameter that the first driving parameter group includes is consistent.
Illustratively, vehicle is being obtained before the first driving parameter group that current landform uplink is sailed, need to collect vehicle point
Multiple groups the second vehicle parameter group not travelled in different terrain, to establish in the second vehicle parameter group each vehicle parameter not
The corresponding relationship of same value and different terrain.
In step 105, according to aforementioned the second vehicle parameter of multiple groups group, the corresponding terrain information of every kind of vehicle parameter is determined
Projection matrix.
In step 106, for any landform in aforementioned different terrain, believed according to the corresponding landform of every kind of vehicle parameter
Projection matrix is ceased, every kind of vehicle parameter corresponding landform under the landform is obtained and identifies section.
Wherein, the landform is formed correspondingly between the corresponding whole landform cog regions of whole vehicle parameters under the landform
Shape identifies section set.
Illustratively, the step 104 to step 106 may be collectively referred to as through collected a large amount of terrain data (the first drivings
Parameter group) process that landform disaggregated model is trained, which can be in off-board equipment or on mobile unit
It carries out.It when being carried out in off-board equipment, needs before vehicle driving, by trained classification of landform model (i.e. landform
Information projection matrix and the landform identify section) push to vehicle storage device.
Fig. 3 is the flow chart that method is determined according to a kind of topographic projection's matrix shown in embodiment illustrated in fig. 2, such as Fig. 3 institute
Show, which may comprise steps of.
For any vehicle parameter obtained in any landform in aforementioned different terrain:
In step 1051, the peak value and valley that the vehicle parameter includes are obtained, to form the peak value of the vehicle parameter
Group and valley array.
Illustratively, any landform in aforementioned different terrain is indicated with landform 1, is indicated with i-th kind of vehicle parameter in landform 1
Any vehicle parameter of upper acquisition can calculate the peak value and valley that vehicle parameter includes by following method:
Obtain Δ i=di (k)-di (k-1) and β i=di (k-1)-di (k), wherein di (k-1) indicates i-th kind of driving
Kth -1 value of parameter, di (k) indicate k-th of value of i-th kind of vehicle parameter, and Δ i and β i are respectively i-th kind of vehicle parameter
The positive difference and reverse difference of k-th of value and kth -1 value.When Δ i<0, and when β i>0, di_p (k)=di is determined
(k), i.e. di (k) is one of peak value of i-th kind of vehicle parameter；When Δ i>0, and when β i<0, di_v (k)=di is determined
(k), i.e. di (k) is one of valley of i-th kind of vehicle parameter.
All peak values and valley that every kind of vehicle parameter includes can be obtained by the above method.It should be noted that one
The quantity of peak value and valley that kind vehicle parameter includes is not necessarily identical, subsequent calculating for convenience, in construction vehicle parameter
Peak value array and valley array when, it is necessary first to unification is carried out to the quantity of the peak value and valley that get, for example, when i-th
When kind vehicle parameter includes 6 peak values and 5 valleies, the unified value Sizei_p_v of peak value number and valley number is obtained first
=min (6,5)=5, and then construct array di_p_resized=[P (1) P (2) P (3) P (4) P (5)] and di_v_
Resized=[V (1) V (2) V (3) V (4) V (5)], wherein di_p_resized is the peak value array of i-th kind of vehicle parameter,
Di_v_resized is the valley array of i-th kind of vehicle parameter.
In step 1052, according to the peak value array and the valley array, the first peak averaging of the vehicle parameter is calculated
Value and the first valley average value.
Illustratively, the first peak average value and the first valley of vehicle parameter can be calculated by following equation (1) and (2)
Average value:
Wherein, Sizei_p_v is the number of element in the peak value array and valley array of i-th kind of vehicle parameter, μi_pIt is
First peak average value of i kind vehicle parameter, μi_vFirst valley average value of i-th kind of vehicle parameter.
In step 1053, according to the peak value array, the valley array, first peak average value and first valley
Average value constructs the first driving parameter matrix of the vehicle parameter.
Illustratively, the constructive formula of the first driving parameter matrix of the vehicle parameter can be with are as follows:
Wherein, X1iIndicate first driving parameter matrix of i-th kind of vehicle parameter under landform 1, di_p_resized i-th
The peak value array di_v_resized of kind vehicle parameter is the valley array of i-th kind of vehicle parameter, μi_pFor i-th kind of vehicle parameter
The first peak average value, μi_vFirst valley average value of i-th kind of vehicle parameter, Sizei_p_v are i-th kind of vehicle parameter
The number of element in peak value array and valley array, landform 1 are any landform in aforementioned different terrain.
In step 1054, according to first peak average value, the first valley average value and first vehicle parameter
Matrix obtains the discrete matrix of the first driving parameter matrix.
It is exemplary, the average value of first peak average value and the first valley average value can be calculated, as the row
The average value of all values in vehicle parameter, then, the constructive formula of the first driving parameter matrix of the vehicle parameter can be with are as follows:
S1i=(X1i-μ1i)(X1i-μ1i)T (4)
Wherein, S1iIndicate the discrete matrix of first driving parameter matrix of the i-th middle rolling car parameter under landform 1, X1iIt is
First driving parameter matrix of the i kind vehicle parameter under landform 1, μ1iAll values in the i-th kind of vehicle parameter obtained in landform 1
Average value.
In step 1055, after obtaining the corresponding discrete matrix of the vehicle parameter under every kind of landform, according to acquisition
Whole discrete matrix, determines the corresponding terrain information projection matrix of the vehicle parameter.
Illustratively, after obtaining the corresponding discrete matrix of the vehicle parameter under every kind of landform, it is necessary first to calculate this and obtain
The sum of the whole discrete matrix taken.It should be noted that whole discrete matrix of the acquisition are in all possible landform
The corresponding whole discrete matrix of the same vehicle parameter of upper acquisition.After this, which is determined by formula (5)
The terrain information projection matrix answered:
Wherein, WiIndicate the corresponding terrain information projection matrix of i-th kind of vehicle parameter,For in all possible landform
The transposed matrix of the corresponding discrete matrix of i-th kind of vehicle parameter got, μ1iThe i-th kind of vehicle parameter obtained in landform 1
The average value of middle all values, μ2iThe average value of all values in the i-th kind of vehicle parameter obtained in landform 2.
Fig. 4 is to identify that the flow chart of the method for determination is gathered in section according to a kind of landform shown in embodiment illustrated in fig. 2, is such as schemed
Shown in 4, the first landform is any landform in aforementioned different terrain, the first vehicle parameter is in the second vehicle parameter group
Any vehicle parameter, the step 106 may comprise steps of.
In step 1061, according to the corresponding terrain information projection matrix of the first vehicle parameter, determine this first
First projection information of the vehicle parameter under first landform.
Wherein, which is the transposed matrix of the terrain information projection matrix of the first vehicle parameter and is somebody's turn to do
The product of the corresponding first driving parameter matrix of the first vehicle parameter.
It is in step 1062, the maximum value of first projection information, minimum value, average value, variance and adjustment is average
Input of the number as landform identification formula (6) identifies that data, the landform are known to obtain the first landform of the first vehicle parameter
Other formula (6) are as follows:
Wherein, Sig1iIndicate that the first landform of i-th kind of vehicle parameter in the second vehicle parameter group obtained in landform 1 is known
Other data, Y1imaxIndicate the maximum value of the first projection information of i-th kind of vehicle parameter, Y1iminIndicate i-th kind of vehicle parameter
The first projection information minimum value, SDy1iIndicate the variance of the first projection information of i-th kind of vehicle parameter, μy1i_adjIt indicates
The adjustment average of first projection information of i-th kind of vehicle parameter；Illustrated with the form of presentation of the step 1061 to 1063,
The formula can state are as follows: work as Y1imaxTake maximum value, the Y of the first projection information of the first vehicle parameter1iminTake this first
The minimum value of first projection information of kind vehicle parameter, SDy1iTake the first projection information of the first vehicle parameter variance,
μy1i_adjWhen taking the adjustment average of the first projection information of the first vehicle parameter, Sig1iFor the first vehicle parameter
First landform identifies data.
Illustratively, in the formula, subscript i is for indicating vehicle parameter kind corresponding to first landform identification data
Class, m are the total number of vehicle parameter type, for example, the second vehicle parameter group includes steering wheel angle, steering wheel angular speed, indulges
To this five kinds of vehicle parameters of acceleration, side acceleration and yaw rate, then m=5, Sig11Indicate vehicle parameter steering wheel
First landform of corner identifies data, Sig12Indicate that the first landform of vehicle parameter steering wheel angular speed identifies data, with such
It pushes away.
In step 1063, data are identified according to the first landform of the first vehicle parameter, determine the first driving
Parameter corresponding landform under first landform identifies section.
Illustratively, first landform identification data can be obtained by the multiple calculating to first landform identification data
Minimum and maximum threshold value, and the corresponding landform of the first vehicle parameter is formed by the minimum and maximum threshold value and identifies section.
It should be understood that the corresponding landform of each vehicle parameter of each landform identifies section, all rows of the comprehensive landform
The landform of vehicle parameter identifies section, the available landform corresponding landform identification section set.
Fig. 5 be according to the flow chart of terrain category probability determination method shown in embodiment illustrated in fig. 1 a kind of, referring to Fig. 5,
The step 102 may comprise steps of:
In step 1021, the corresponding terrain information projection square of every kind of vehicle parameter in the first driving parameter group is obtained
Battle array.
In step 1022, thrown according to aforementioned every kind of vehicle parameter and the corresponding terrain information of aforementioned every kind of vehicle parameter
Shadow matrix determines aforementioned every kind of vehicle parameter corresponding second landform identification data under landform in this prior.
In step 1023, data are identified by the second landform corresponding under this every kind vehicle parameter in this prior landform,
And aforesaid plurality of landform identification section set, determine the terrain category probability.
Fig. 6 is the flow chart that data capture method is identified according to a kind of landform shown in embodiment illustrated in fig. 5, such as Fig. 6 institute
Show, which may comprise steps of.
Any vehicle parameter being directed in the first driving parameter group:
In step 10221, the peak value and valley that the vehicle parameter includes are obtained, to form the peak value of the vehicle parameter
Group and valley array.
In step 10222, according to the peak value array and the valley array, the second peak averaging of the vehicle parameter is calculated
Value and the second valley average value.
In step 10223, according to the peak value array, the valley array, second peak average value and second valley
Average value constructs the second vehicle parameter matrix of the vehicle parameter.
Illustratively, the calculation of the step 10221 to 10223 and step 1051 are no longer superfluous here to 1053 similar
It states.
In step 10224, square is projected according to the second vehicle parameter matrix terrain information corresponding with the vehicle parameter
Battle array obtains second projection information of the vehicle parameter in this prior under landform.
Illustratively, second projection information of the vehicle parameter in this prior under landform can be obtained by formula (7):
Wherein, YxiIndicate second projection information of i-th kind of vehicle parameter at current landform x,Join for the 1st kind of driving
The corresponding terrain information projection matrix of number, XxiThe the second vehicle parameter matrix for being i-th kind of vehicle parameter at current landform x.
It is in step 10225, the maximum value of second projection information, minimum value, average value, variance and adjustment is average
Inputs of the number as landform identification formula (8), to obtain second landform identification data of the vehicle parameter in this prior under landform.
Wherein, landform identification formula (8) are as follows:
Wherein, SigxiIndicate that the second landform of i-th kind of vehicle parameter in the obtain on landform x first driving parameter group is known
Other data, YximaxIndicate the maximum value of the second projection information of i-th kind of vehicle parameter, YximinIndicate i-th kind of vehicle parameter
The second projection information minimum value, SDyxiIndicate the variance of the second projection information of i-th kind of vehicle parameter, μyxi_adjIt indicates
The adjustment average of second projection information of i-th kind of vehicle parameter；For the form of presentation of the step 10221 to 10225
Bright, which can state are as follows: work as YximaxTake maximum value, the Y of the second projection information of the vehicle parameterximinThe driving is taken to join
The minimum value of the second several projection informations, SDyxiTake variance, the μ of the second projection information of the vehicle parameteryxi_adjTake the driving
When the adjustment average of the second projection information of parameter, Sig1iData are identified for the second landform of the vehicle parameter.In addition, subscript
It is consistent in the effect and abovementioned steps 1062 in formulae express of i and parameter m, it repeats no more.
Fig. 7 is the flow chart according to a kind of terrain category method for calculating probability shown in embodiment illustrated in fig. 5, such as Fig. 7 institute
Show, which includes the following steps.
In step 10231, any vehicle parameter being directed in the first driving parameter group determines that the driving is joined
It is right respectively under the different terrain whether second landform identification data of the number under the current landform belong to the vehicle parameter
Any landform identification section answered.
It should be noted that having respectively corresponded landform knowledge under each landform of the every kind of vehicle parameter in the different terrain
Other section forms the landform cog region of the landform between the corresponding whole landform cog regions of whole vehicle parameters under same landform
Between gather.
In step 10232, when second landform identification data of the vehicle parameter under the current landform belong to the row
Vehicle parameter when corresponding any landform identification section, determines that affiliated landform identification section is corresponding under the different terrain
The likelihood probability of landform be 1, and determine the corresponding landform in other landform identification section in addition to affiliated landform identifies section
Likelihood probability be 0.
Illustratively, it is obtained under which is used to indicate to obtain under the current landform the vehicle parameter and different terrain
Vehicle parameter of the same race similarity degree, also, the likelihood probability can further react the current landform and different terrain
Similarity degree under the vehicle parameter.When likelihood probability is 1, can determine in current landform and different terrain a certainly
Shape is completely the same under the vehicle parameter, when likelihood probability be 0 when, can determine current landform with it is a certain in different terrain
Landform is complete inconsistent under the vehicle parameter.
In step 10233, when second landform identification data of the vehicle parameter under the current landform are not belonging to this
Vehicle parameter under the different terrain within corresponding any landform identification section when, identified according to second landform
At a distance from each of data and the landform identification section set landform identification section, calculates second landform and identify number
According to the vehicle parameter under the different terrain the corresponding landform in corresponding each landform identification section it is similar general
Rate.
Illustratively, when determining that second landform identification data of the vehicle parameter in this prior under landform are not at the landform and know
When within any landform identification section in other section set, second landform identification data and the landform cog region are calculated first
Between each of gather the distance in landform identification section, in following equation (9), with corresponding second ground of i-th kind of vehicle parameter
Shape identifies data, the calculating pair at a distance from the jth kind landform corresponding landform identification section in the set of landform identification section
The calculating process of the step 10234 is described:
Wherein, DsigxjiIndicate that the corresponding second landform identification data of i-th kind of vehicle parameter and the landform identify Interval Set
The distance in the jth kind landform corresponding landform identification section in conjunction, can be aforementioned SigxiWith the SigxiLandform identifies section
Threshold value constructs valued space of second terrain data on an XY axial plane, SigxXiFor SigxiCoordinate in X-axis,
Sig_thrjXiFor SigxiCoordinate of the threshold value in X-axis, SigxYiFor SigxiCoordinate in Y-axis, Sig_thrjYiFor SigxiThreshold
It is worth the coordinate in Y-axis.
In step 10234, the corresponding whole phases of whole vehicle parameters that the first driving parameter group includes are being obtained
After probability, by being averaged for the corresponding whole likelihood probabilities of the whole vehicle parameters for corresponding to same landform in the different terrain
Value belongs to the terrain category probability of the landform as the current landform.
Illustratively, it after the corresponding multiple likelihood probabilities of the every kind of vehicle parameter obtained under landform in this prior, needs
Calculate every kind of vehicle parameter in this prior under landform corresponding multiple likelihood probabilities average value, for example, under current landform
Collected first parameter group includes: steering wheel angle, steering wheel angular speed and longitudinal acceleration these three vehicle parameters, and
Calculate separately out driving parametric direction disk corner, steering wheel angular speed and longitudinal acceleration in this prior under landform with certain
Three of one landform likelihood probabilities, and then the average value of these three likelihood probabilities can be calculated, obtain the current landform with
The likelihood probability of a certain landform, as the terrain category probability.
After abovementioned steps 1022 and 1023, abovementioned steps 103 may include: that the terrain category probability value is maximum
Landform is determined as landform belonging to the current landform.
In conclusion the recognition methods of landform locating for the vehicle that the disclosure provides, can obtain vehicle in current landform
Traveling first driving parameter group, this first driving parameter group include steering wheel angle, steering wheel angular speed, longitudinal acceleration,
At least one of side acceleration and yaw rate vehicle parameter；It is and right respectively according to the first driving parameter group
It should identify that section is gathered in multiple landform of different terrain, determine that the current landform belongs to the landform of each landform in different terrain
Class probability；The terrain category probability for belonging to each landform in different terrain according to the current landform determines the current landform institute
The landform of category.Can by vehicle part collected in vehicle driving and vehicle are acted relevant vehicle parameter calculating and
The precise classification of processing and the landform being likely to be to vehicle identifies, the present invention is without adopting to landform locating for vehicle
Collect Image Acquisition, avoids environmental factor on influence caused by landform identification, recognition result is more acurrate, is conducive to further preferably
Vehicle is controlled, while also improving the intelligence degree of vehicle.
Block diagram of the Fig. 8 for the identification device of landform locating for a kind of vehicle described in the embodiment of the present invention, the locating ground of the vehicle
The identification device 800 of shape can be used for executing method described in Fig. 1.Referring to Fig. 8, which may include:
Parameter acquisition module 810, the first driving parameter group sailed for obtaining vehicle in current landform uplink, the first row
Vehicle parameter group include in steering wheel angle, steering wheel angular speed, longitudinal acceleration, side acceleration and yaw rate extremely
A kind of few vehicle parameter；
Probability determination module 820 is used for according to the first driving parameter group, and corresponds respectively to the multiple of different terrain
Landform identifies section set, determines that the current landform belongs to the terrain category probability of each landform in aforementioned different terrain；
Landform determining module 830, for belonging to the landform class of each landform in aforementioned different terrain according to the current landform
Other probability determines landform belonging to the current landform.
Further, Fig. 9 is the frame of the identification device of the landform according to locating for another vehicle shown in embodiment illustrated in fig. 8
Scheme, the identification device 800 of landform locating for the vehicle can be used for executing any method of Fig. 2 to Fig. 4.Referring to Fig. 9, the dress
Setting 800 can also include:
Parameter collection module 840, multiple groups the second vehicle parameter group travelled in different terrain respectively for collecting vehicle,
The type for the vehicle parameter that every group of vehicle parameter group in aforementioned the second vehicle parameter of multiple groups group is included and first driving are joined
The type for the vehicle parameter that array includes is consistent；
Projection matrix determining module 850, for determining every kind of vehicle parameter pair according to aforementioned the second vehicle parameter of multiple groups group
The terrain information projection matrix answered；
Section obtains module 860, for being corresponded to according to every kind of vehicle parameter for any landform in aforementioned different terrain
Terrain information projection matrix, obtain every kind of vehicle parameter corresponding landform under the landform and identify section, wherein in the landform
Under the corresponding whole landform cog regions of whole vehicle parameters between form the landform corresponding landform identification section set.
Further, the projection matrix determining module 850, is used for:
For any vehicle parameter obtained in any landform in aforementioned different terrain:
The peak value and valley that the vehicle parameter includes are obtained, to form the peak value array and valley array of the vehicle parameter；
According to the peak value array and the valley array, the first peak average value and the first valley for calculating the vehicle parameter are flat
Mean value；
According to the peak value array, the valley array, first peak average value and the first valley average value, construction should
First driving parameter matrix of vehicle parameter；
According to first peak average value, the first valley average value and the first driving parameter matrix, obtain this
The discrete matrix of one driving parameter matrix；
After obtaining the corresponding discrete matrix of the vehicle parameter under every kind of landform, according to whole discrete squares of acquisition
Battle array, determines the corresponding terrain information projection matrix of the vehicle parameter.
Further, the first landform is any landform in aforementioned different terrain, the first vehicle parameter is second row
Any vehicle parameter in vehicle parameter group, the section obtain module 860, are used for:
According to the corresponding terrain information projection matrix of the first vehicle parameter, determine the first vehicle parameter this
The first projection information under one landform, wherein first projection information is that the terrain information of the first vehicle parameter projects square
The product of the transposed matrix of battle array the first driving parameter matrix corresponding with the first vehicle parameter；
It is identified using the maximum value of first projection information, minimum value, average value, variance and adjustment average as landform
The input of formula identifies that data, the landform identify formula to obtain the first landform of the first vehicle parameter are as follows:
Wherein, Sig1iIndicate that the first landform of i-th kind of vehicle parameter in the second vehicle parameter group obtained in landform 1 is known
Other data, Y1imaxIndicate the maximum value of the first projection information of i-th kind of vehicle parameter, Y1iminIndicate i-th kind of vehicle parameter
The first projection information minimum value, SDy1iIndicate the variance of the first projection information of i-th kind of vehicle parameter, μy1i_adjIt indicates
The adjustment average of first projection information of i-th kind of vehicle parameter；Work as Y1imaxThe first of the first vehicle parameter is taken to project
The maximum value of information, Y1iminTake minimum value, the SD of the first projection information of the first vehicle parametery1iTake the first driving
The variance of first projection information of parameter, μy1i_adjTake the adjustment average of the first projection information of the first vehicle parameter
When, Sig1iData are identified for the first landform of the first vehicle parameter；
Data are identified according to the first landform of the first vehicle parameter, determine the first vehicle parameter on first ground
Corresponding landform identifies section under shape.
Optionally, Figure 10 is the block diagram according to a kind of probability determination module shown in embodiment illustrated in fig. 8, the determine the probability
Module 820 can be used for executing any method of Fig. 4 to Fig. 7.Referring to Figure 10, the probability determination module 820, comprising:
Projection matrix acquisition submodule 821, it is corresponding for obtaining every kind of vehicle parameter in the first driving parameter group
Terrain information projection matrix；
Identification data determine submodule 822, for according to aforementioned every kind of vehicle parameter and aforementioned every kind of vehicle parameter pair
The terrain information projection matrix answered determines aforementioned every kind of vehicle parameter corresponding second landform identification number under landform in this prior
According to；
Determine the probability submodule 823, for passing through corresponding second ground under aforementioned every kind of vehicle parameter in this prior landform
Shape identifies data and aforesaid plurality of landform identification section set, determines the terrain category probability.
Further, which determines submodule 822, is used for:
Any vehicle parameter being directed in the first driving parameter group:
The peak value and valley that the vehicle parameter includes are obtained, to form the peak value array and valley array of the vehicle parameter；
According to the peak value array and the valley array, the second peak average value and the second valley for calculating the vehicle parameter are flat
Mean value；
According to the peak value array, the valley array, second peak average value and the second valley average value, construction should
Second vehicle parameter matrix of vehicle parameter；
According to the second vehicle parameter matrix terrain information projection matrix corresponding with the vehicle parameter, driving ginseng is obtained
Count the second projection information under landform in this prior；
It is identified using the maximum value of second projection information, minimum value, average value, variance and adjustment average as landform
The input of formula identifies that data, the landform identify formula to obtain second landform of the vehicle parameter in this prior under landform are as follows:
Wherein, SigxiIndicate that the second landform of i-th kind of vehicle parameter in the obtain on landform x first driving parameter group is known
Other data, YximaxIndicate the maximum value of the second projection information of i-th kind of vehicle parameter, YximinIndicate i-th kind of vehicle parameter
The second projection information minimum value, SDyxiIndicate the variance of the second projection information of i-th kind of vehicle parameter, μyxi_adjIt indicates
The adjustment average of second projection information of i-th kind of vehicle parameter；Work as YximaxTake the second projection information of the vehicle parameter
Maximum value, YximinTake minimum value, the SD of the second projection information of the vehicle parameteryxiTake the second projection information of the vehicle parameter
Variance, μyxi_adjWhen taking the adjustment average of the second projection information of the vehicle parameter, Sig1iIt is the second of the vehicle parameter
Landform identifies data.
Further, landform cog region has been respectively corresponded under each landform of the every kind of vehicle parameter in aforementioned different terrain
Between, the landform identification Interval Set of the landform is formed between the corresponding whole landform cog regions of whole vehicle parameters under same landform
It closes, which is used for:
Any vehicle parameter being directed in the first driving parameter group, determines the vehicle parameter in this prior under landform
The second landform identification data whether belong to the vehicle parameter corresponding any landform cog region under aforementioned different terrain
Between；
When the second landform identification data under the vehicle parameter in this prior landform belong to the vehicle parameter in aforementioned difference
Under landform when corresponding any landform identification section, the likelihood probability of the affiliated corresponding landform in landform identification section is determined
It is 1, and determines that the likelihood probability of the corresponding landform in other landform identification section in addition to affiliated landform identifies section is 0；Or
Person,
When the second landform identification data under the vehicle parameter in this prior landform are not belonging to the vehicle parameter in the difference
When under landform within corresponding any landform identification section, identify that data and the landform identify section according to second landform
Set each of landform identification section distance, calculate second landform identification data and the vehicle parameter this differently
The likelihood probability of the corresponding landform in corresponding each landform identification section under shape；
After obtaining the corresponding whole likelihood probabilities of whole vehicle parameters that the first driving parameter group includes, it will correspond to
The average value of the corresponding whole likelihood probabilities of whole vehicle parameters of same landform is as the current landform in aforementioned different terrain
Belong to the terrain category probability of the landform.
Further, the landform determining module 830, is used for:
The maximum landform of terrain category probability value is determined as landform belonging to the current landform.
In conclusion the identification device of landform locating for the vehicle that the disclosure provides, can obtain vehicle in current landform
Traveling first driving parameter group, this first driving parameter group include steering wheel angle, steering wheel angular speed, longitudinal acceleration,
At least one of side acceleration and yaw rate vehicle parameter；It is and right respectively according to the first driving parameter group
It should identify that section is gathered in multiple landform of different terrain, determine that the current landform belongs to the landform of each landform in different terrain
Class probability；The terrain category probability for belonging to each landform in different terrain according to the current landform determines the current landform institute
The landform of category.Can by vehicle part collected in vehicle driving and vehicle are acted relevant vehicle parameter calculating and
Processing and the classification for the landform being likely to be to vehicle identify, the present invention is without acquiring figure to landform locating for vehicle
As acquisition, avoid environmental factor on influence caused by landform identification, recognition result is more acurrate, is conducive to further preferably to vehicle
It is controlled, while also improving the intelligence degree of vehicle.
The foregoing is merely illustrative of the preferred embodiments of the present invention, is not intended to limit the invention, all in essence of the invention
Within mind and principle, any modification, equivalent replacement, improvement and so on be should all be included in the protection scope of the present invention.
Claims (10)
1. a kind of recognition methods of landform locating for vehicle, which is characterized in that the described method includes:
The first driving parameter group that vehicle is sailed in current landform uplink is obtained, the first driving parameter group includes that steering wheel turns
At least one of angle, steering wheel angular speed, longitudinal acceleration, side acceleration and yaw rate vehicle parameter；
According to the first driving parameter group, and multiple landform identification section set of different terrain is corresponded respectively to, determined
The current landform belongs to the terrain category probability of each landform in the different terrain；
The terrain category probability for belonging to each landform in the different terrain according to the current landform determines the current landform
Affiliated landform.
2. the method according to claim 1, wherein the first driving sailed in acquisition vehicle in current landform uplink
Before parameter group, the method also includes:
Multiple groups the second vehicle parameter group for travelling in different terrain respectively of vehicle is collected, in the second vehicle parameter of multiple groups group
Every group of vehicle parameter group vehicle parameter for being included type and the first driving parameter group vehicle parameter for including kind
Class is consistent；
According to the second vehicle parameter of multiple groups group, the corresponding terrain information projection matrix of every kind of vehicle parameter is determined；
It is obtained for any landform in the different terrain according to the corresponding terrain information projection matrix of every kind of vehicle parameter
Every kind of vehicle parameter corresponding landform under the landform identifies section, wherein whole vehicle parameters under the landform are corresponding
The landform corresponding landform identification section set is formed between whole landform cog regions.
3. according to the method described in claim 2, it is characterized in that, described according to the second vehicle parameter of multiple groups group, determination
The corresponding terrain information projection matrix of every kind of vehicle parameter, comprising:
For any vehicle parameter obtained in any landform in the different terrain:
The peak value and valley that the vehicle parameter includes are obtained, to form the peak value array and valley array of the vehicle parameter；
According to the peak value array and the valley array, the first peak average value and the first valley for calculating the vehicle parameter are flat
Mean value；
According to the peak value array, the valley array, first peak average value and the first valley average value, structure
Make the first driving parameter matrix of the vehicle parameter；
According to first peak average value, the first valley average value and the first driving parameter matrix, institute is obtained
State the discrete matrix of the first driving parameter matrix；
After obtaining the corresponding discrete matrix of the vehicle parameter under every kind of landform, according to whole discrete matrix of acquisition, really
The fixed corresponding terrain information projection matrix of the vehicle parameter.
4. according to the method described in claim 3, it is characterized in that, the first landform be the different terrain in any landform,
The first vehicle parameter is any vehicle parameter in the second vehicle parameter group, according to the first described vehicle parameter pair
The terrain information projection matrix answered obtains the first vehicle parameter corresponding landform cog region under first landform
Between, comprising:
According to the corresponding terrain information projection matrix of the first described vehicle parameter, determine the first described vehicle parameter described
The first projection information under first landform, wherein first projection information is the terrain information of the first vehicle parameter
The product of the transposed matrix of projection matrix the first driving parameter matrix corresponding with the first described vehicle parameter；
It is identified using the maximum value of first projection information, minimum value, average value, variance and adjustment average as landform public
The input of formula identifies that data, the landform identify formula to obtain the first landform of the first vehicle parameter are as follows:
Wherein, Sig1iIndicate the first landform identification number of i-th kind of vehicle parameter in the second vehicle parameter group obtained in landform 1
According to Y1imaxIndicate the maximum value of the first projection information of i-th kind of vehicle parameter, Y1iminIndicate i-th kind of vehicle parameter
The first projection information minimum value, SDy1iIndicate the variance of the first projection information of i-th kind of vehicle parameter, μy1i_adjTable
Show the adjustment average of the first projection information of i-th kind of vehicle parameter；Work as Y1imaxTake the of the first vehicle parameter
The maximum value of one projection information, Y1iminTake minimum value, the SD of the first projection information of the first vehicle parametery1iIt takes described
The variance of first projection information of the first vehicle parameter, μy1i_adjTake the first projection information of the first vehicle parameter
When adjusting average, Sig1iData are identified for the first landform of the first vehicle parameter；
Data are identified according to the first landform of the first vehicle parameter, determine the first described vehicle parameter described first
Corresponding landform identifies section under landform.
5. the method according to claim 1, wherein described according to the first driving parameter group, and difference
Corresponding to multiple landform identification section set of different terrain, determine that the current landform belongs in the different terrain eachly
The terrain category probability of shape, comprising:
Obtain the corresponding terrain information projection matrix of every kind of vehicle parameter in the first driving parameter group；
According to every kind of vehicle parameter and the corresponding terrain information projection matrix of every kind of vehicle parameter, determine described every
Kind vehicle parameter corresponding second landform under the current landform identifies data；
By every kind of vehicle parameter, corresponding second landform identifies data and the multiple under the current landform
Shape identifies section set, determines the terrain category probability.
6. according to the method described in claim 5, it is characterized in that, described according to every kind of vehicle parameter and every kind described
The corresponding terrain information projection matrix of vehicle parameter determines every kind of vehicle parameter corresponding second under the current landform
Landform identifies data, comprising:
Any vehicle parameter being directed in the first driving parameter group:
The peak value and valley that the vehicle parameter includes are obtained, to form the peak value array and valley array of the vehicle parameter；
According to the peak value array and the valley array, the second peak average value and the second valley for calculating the vehicle parameter are flat
Mean value；
According to the peak value array, the valley array, second peak average value and the second valley average value, structure
Make the second vehicle parameter matrix of the vehicle parameter；
According to the second vehicle parameter matrix terrain information projection matrix corresponding with the vehicle parameter, the vehicle parameter is obtained
The second projection information under the current landform；
It is identified using the maximum value of second projection information, minimum value, average value, variance and adjustment average as landform public
The input of formula identifies that data, the landform identify formula to obtain second landform of the vehicle parameter under the current landform
Are as follows:
Wherein, SigxiIndicate the second landform identification number of i-th kind of vehicle parameter in the obtain on landform x first driving parameter group
According to YximaxIndicate the maximum value of the second projection information of i-th kind of vehicle parameter, YximinIndicate i-th kind of vehicle parameter
The second projection information minimum value, SDyxiIndicate the variance of the second projection information of i-th kind of vehicle parameter, μyxi_adjTable
Show the adjustment average of the second projection information of i-th kind of vehicle parameter；Work as YximaxThe second of the vehicle parameter is taken to project letter
The maximum value of breath, YximinTake minimum value, the SD of the second projection information of the vehicle parameteryxiThe second of the vehicle parameter is taken to project
The variance of information, μyxi_adjWhen taking the adjustment average of the second projection information of the vehicle parameter, Sig1iFor the vehicle parameter
Second landform identifies data.
7. according to the method described in claim 5, it is characterized in that, every kind of vehicle parameter in each of described different terrain
Landform identification section is respectively corresponded under shape, group between the corresponding whole landform cog regions of whole vehicle parameters under same landform
Gather at the landform identification section of the landform,
By every kind of vehicle parameter, corresponding second landform identifies data and the multiple under the current landform
Shape identifies section set, determines the terrain category probability, comprising:
Any vehicle parameter being directed in the first driving parameter group, determines the vehicle parameter under the current landform
The second landform identification data whether belong to the vehicle parameter corresponding any landform cog region under the different terrain
Between；
When the vehicle parameter under the current landform the second landform identification data belong to the vehicle parameter it is described differently
Under shape when corresponding any landform identification section, determine that the likelihood probability of the affiliated corresponding landform in landform identification section is
1, and determine that the likelihood probability of the corresponding landform in other landform identification section in addition to affiliated landform identifies section is 0；Alternatively,
When second landform identification data of the vehicle parameter under the current landform are not belonging to the vehicle parameter in the difference
When under landform within corresponding any landform identification section, identify that data and the landform identify according to second landform
The distance in each of section set landform identification section, calculates the second landform identification data and the vehicle parameter in institute
State the likelihood probability of the corresponding landform in corresponding each landform identification section under different terrain；
After obtaining the corresponding whole likelihood probabilities of whole vehicle parameters that the first driving parameter group includes, institute will be corresponded to
The average value of the corresponding whole likelihood probabilities of whole vehicle parameters of same landform in different terrain is stated as the current landform
Belong to the terrain category probability of the landform.
8. the method according to the description of claim 7 is characterized in that described belong to the different terrain according to the current landform
In each landform terrain category probability, determine landform belonging to the current landform, comprising:
The maximum landform of terrain category probability value is determined as landform belonging to the current landform.
9. a kind of identification device of landform locating for vehicle, which is characterized in that described device includes:
Parameter acquisition module, the first driving parameter group sailed for obtaining vehicle in current landform uplink, the first driving ginseng
Array includes at least one in steering wheel angle, steering wheel angular speed, longitudinal acceleration, side acceleration and yaw rate
Kind vehicle parameter；
Probability determination module is used for according to the first driving parameter group, and corresponds respectively to multiple landform of different terrain
It identifies section set, determines that the current landform belongs to the terrain category probability of each landform in the different terrain；
Landform determining module, the terrain category for belonging to each landform in the different terrain according to the current landform are general
Rate determines landform belonging to the current landform.
10. a kind of vehicle, which is characterized in that the identification device including landform locating for vehicle as claimed in claim 9.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201711040120.4A CN109727334B (en) | 2017-10-30 | 2017-10-30 | Method and device for identifying terrain where vehicle is located and vehicle |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201711040120.4A CN109727334B (en) | 2017-10-30 | 2017-10-30 | Method and device for identifying terrain where vehicle is located and vehicle |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109727334A true CN109727334A (en) | 2019-05-07 |
| CN109727334B CN109727334B (en) | 2021-03-26 |
Family
ID=66292854
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201711040120.4A Active CN109727334B (en) | 2017-10-30 | 2017-10-30 | Method and device for identifying terrain where vehicle is located and vehicle |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109727334B (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113190015A (en) * | 2021-05-20 | 2021-07-30 | 中国第一汽车股份有限公司 | Terrain recognition method and device, computer equipment and storage medium |
| CN113492861A (en) * | 2021-07-26 | 2021-10-12 | 中汽创智科技有限公司 | Vehicle control method and device |
Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2008158598A (en) * | 2006-12-21 | 2008-07-10 | Hitachi Software Eng Co Ltd | Road surface circumstance management system |
| CN101689296A (en) * | 2007-04-19 | 2010-03-31 | 电子地图有限公司 | Method of and apparatus for producing road information |
| WO2013011992A1 (en) * | 2011-07-20 | 2013-01-24 | 株式会社ブリヂストン | Road surface condition estimation method, and road surface condition estimation device |
| CN103764445A (en) * | 2011-08-23 | 2014-04-30 | 罗伯特·博世有限公司 | Method, control unit, and computer program product for setting a range of a headlamp of a vehicle |
| JP2015063228A (en) * | 2013-09-25 | 2015-04-09 | 富士重工業株式会社 | Control device of vehicle |
| CN104540717A (en) * | 2012-08-09 | 2015-04-22 | 株式会社普利司通 | Method and device for determining state of road surface |
| CN104835327A (en) * | 2015-06-01 | 2015-08-12 | 北京经纬恒润科技有限公司 | Method and device for identifying pavement information |
| CN105740793A (en) * | 2016-01-26 | 2016-07-06 | 哈尔滨工业大学深圳研究生院 | Road bump condition and road type identification based automatic speed adjustment method and system |
| CN106092600A (en) * | 2016-05-31 | 2016-11-09 | 东南大学 | A kind of pavement identification method for proving ground strengthening road |
-
2017
- 2017-10-30 CN CN201711040120.4A patent/CN109727334B/en active Active
Patent Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2008158598A (en) * | 2006-12-21 | 2008-07-10 | Hitachi Software Eng Co Ltd | Road surface circumstance management system |
| CN101689296A (en) * | 2007-04-19 | 2010-03-31 | 电子地图有限公司 | Method of and apparatus for producing road information |
| WO2013011992A1 (en) * | 2011-07-20 | 2013-01-24 | 株式会社ブリヂストン | Road surface condition estimation method, and road surface condition estimation device |
| CN103764445A (en) * | 2011-08-23 | 2014-04-30 | 罗伯特·博世有限公司 | Method, control unit, and computer program product for setting a range of a headlamp of a vehicle |
| CN104540717A (en) * | 2012-08-09 | 2015-04-22 | 株式会社普利司通 | Method and device for determining state of road surface |
| JP2015063228A (en) * | 2013-09-25 | 2015-04-09 | 富士重工業株式会社 | Control device of vehicle |
| CN104835327A (en) * | 2015-06-01 | 2015-08-12 | 北京经纬恒润科技有限公司 | Method and device for identifying pavement information |
| CN105740793A (en) * | 2016-01-26 | 2016-07-06 | 哈尔滨工业大学深圳研究生院 | Road bump condition and road type identification based automatic speed adjustment method and system |
| CN106092600A (en) * | 2016-05-31 | 2016-11-09 | 东南大学 | A kind of pavement identification method for proving ground strengthening road |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113190015A (en) * | 2021-05-20 | 2021-07-30 | 中国第一汽车股份有限公司 | Terrain recognition method and device, computer equipment and storage medium |
| CN113492861A (en) * | 2021-07-26 | 2021-10-12 | 中汽创智科技有限公司 | Vehicle control method and device |
Also Published As
| Publication number | Publication date |
|---|---|
| CN109727334B (en) | 2021-03-26 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107492251B (en) | Driver identity recognition and driving state monitoring method based on machine learning and deep learning | |
| CN104573646B (en) | Chinese herbaceous peony pedestrian detection method and system based on laser radar and binocular camera | |
| CN102222236B (en) | Image processing system and position measuring system | |
| US8487991B2 (en) | Clear path detection using a vanishing point | |
| CN108280415A (en) | Driving behavior recognition methods based on intelligent mobile terminal | |
| CN104508719B (en) | Drive assist system and driving assistance method | |
| CN105373783A (en) | Seat belt not-wearing detection method based on mixed multi-scale deformable component model | |
| CN109460023A (en) | Driver's lane-changing intention recognition methods based on Hidden Markov Model | |
| CN105892471A (en) | Automatic automobile driving method and device | |
| CN104933708A (en) | Barrier detection method in vegetation environment based on multispectral and 3D feature fusion | |
| DE102008059822A1 (en) | A position range setting device, control method, and control for a mobile-mounted device and a vehicle air conditioner, and control methods therefor | |
| CN105718868A (en) | Face detection system and method for multi-pose faces | |
| DE102014104663B4 (en) | Road-based localization | |
| DE102008008446A1 (en) | Vehicle air conditioning and method and apparatus for controlling vehicle air conditioning | |
| DE102012222497A1 (en) | System and method for detecting parking space markings for vehicles | |
| CN109050535A (en) | A kind of quick landform industry and mining city method based on vehicle attitude | |
| CN106250816A (en) | A kind of Lane detection method and system based on dual camera | |
| CN105740782A (en) | Monocular vision based driver lane-changing process quantization method | |
| CN103530865A (en) | Apparatus and methods for use in detecting three-dimensional object around the image of the vehicle | |
| US20170057495A1 (en) | Map of the Surroundings for Driving Areas with Random Altitude Profile | |
| CN109727334A (en) | Recognition methods, device and the vehicle of landform locating for vehicle | |
| CN103679205A (en) | Preceding car detection method based on shadow hypothesis and layered HOG (histogram of oriented gradient) symmetric characteristic verification | |
| CN105590087B (en) | A kind of roads recognition method and device | |
| CN113859226B (en) | Motion planning and automatic parking method based on reinforcement learning | |
| CN110263836B (en) | Bad driving state identification method based on multi-feature convolutional neural network |
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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |