CN107767023A - A kind of road traffic service level evaluation method based on the matching of fuzzy KNN space characteristics - Google Patents
A kind of road traffic service level evaluation method based on the matching of fuzzy KNN space characteristics Download PDFInfo
- Publication number
- CN107767023A CN107767023A CN201710850376.5A CN201710850376A CN107767023A CN 107767023 A CN107767023 A CN 107767023A CN 201710850376 A CN201710850376 A CN 201710850376A CN 107767023 A CN107767023 A CN 107767023A
- Authority
- CN
- China
- Prior art keywords
- mrow
- msub
- data
- road traffic
- service level
- 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
Landscapes
- Traffic Control Systems (AREA)
Abstract
A kind of road traffic service level evaluation method based on the matching of fuzzy KNN space characteristics, first, road traffic features reference sequences are established, obtained under different modalities, road traffic reference data spatially;Secondly, road traffic training data and test data under same mode, spatially is extracted, based on reference data, obtains road traffic difference data, handled by thresholding, obtain the feature of training data and test data;By road traffic service level Fuzzy processing, combined training data characteristics, the structure of knowledge base is completed;Finally, by KNN space characteristics matching algorithms, k feature nearest with test data characteristic distance in knowledge base is chosen, and obtains the service level of corresponding blurring;The service level of k groups blurring is added, draws service level corresponding to maximum probability, the as affiliated service level of current signature, completes the road traffic evaluation of running status based on space characteristics matching.
Description
Technical field
The invention belongs to road traffic service level to evaluate field, is related to the A+E of highway traffic data, is one
The road traffic service level evaluation method of kind space characteristics matching.
Background technology
With the development of the social economy, the recoverable amount of automobile constantly increases, road traffic problem becomes more acute, must
Road traffic service level must correctly be evaluated, it can be handled before traffic problems generation, avoid handing over
The generation of the events such as logical congestion, accident, formulate decision-making for vehicle supervision department and foundation is provided.
At present, the research major part of road traffic service level evaluation is for congested in traffic expansion.The friendship of comparative maturity
Logical service level evaluation index has connective, Travel Time Reliability and Road Network Capacity reliability.Other level of service are commented
Valency research also has traffic congestion evaluation weighted model, road network entirety Adaptability Analysis method, space saturation degree index, road network dynamic
Traffic flow modes estimation etc..The studies above is mainly studied in road network aspect, does not account for the traffic in wall scroll section
Service level, and implementation process is complex.
The content of the invention
In order to overcome the shortcomings of existing road traffic service level evaluation method, the present invention provides a kind of simplified algorithm, base
In the road traffic service level evaluation method of fuzzy KNN space characteristics matching.
The technical solution adopted for the present invention to solve the technical problems is:
It is a kind of based on fuzzy KNN space characteristics matching road traffic service level evaluation method, methods described include with
Lower step:
1) highway traffic data for obtaining different sections of highway under same mode, spatially establishes road traffic features reference
Sequence, based on spatial Correlation Analysis, benchmark section is selected, and using its data as road traffic reference data spatially;
2) extract under same mode, the historical data in spatially other sections, as training data, based on same mode
Under, road traffic reference data spatially, obtain difference data, handled by thresholding, obtain the feature of training data;
3) divided by existing highway traffic data and road traffic service level, the feature of combined training data is complete
Into the structure of knowledge base;
4) obtain under same mode, the test data in spatially other sections, as test data, based on same mode
Under, road traffic reference data spatially, obtain road traffic difference data spatially, handled by thresholding, obtain
The feature of test data;
5) by KNN Feature Correspondence Algorithms, k feature nearest with test data characteristic distance in knowledge base is chosen, and
The service level of blurring corresponding to obtaining;
6) service level by the blurring of k groups is added, and draws service level corresponding to maximum probability, as current signature
Affiliated service level, complete the road traffic service level evaluation of space characteristics matching.
Further, in the step 1), the highway traffic data for obtaining different sections of highway under same mode, spatially is built
Vertical road traffic features reference sequences, based on spatial Correlation Analysis, benchmark section is selected, and using its data as spatially
Road traffic reference data, comprise the following steps:
1.1) division of road traffic operational modal
The division of road traffic operational modal is divided into two levels:Road network layer and section layer, the traffic fortune of setting road network layer
The traffic circulation mode of road is divided into g seed mode by the division mark of row mode, and the traffic circulation mode of section layer is drawn
Minute mark is known is divided into h seed mode by the traffic circulation mode of road, then the traffic circulation mode of road is divided into g × h altogether
Kind, it is designated as set Mode={ M11, M12..., Mgh, the division mark of wherein g and h value traffic circulation mode selected by
It is determined that;
1.2) structure of road traffic features reference sequences is designed
The collection period for setting road traffic state data is Δ t;
As shown in Table 1 and Table 2, table 1 is road traffic features reference sequences to the sheet format of road traffic features reference sequences
Information table, table 2 are that road traffic features reference sequences describe table:
Table 1
Table 2
P bars are selected in setting altogether has the section of spatial correlation characteristic, is designated as:
L=[L1L2…Lp] (1)
Wherein, p represents the section bar number on path space;LiRepresent i-th 1≤i of section≤p;L represents the tool of selection
There is the set in spatial correlation characteristic section.
Further, in the step 2), extract under same mode, the historical data in spatially other sections, as instruction
Practice data, based on road traffic reference data under same mode, spatially, obtain difference data, handled by thresholding,
The feature of training data is obtained, its general expression is as follows:
Si(m* Δs t, Mgh)=STi(m* Δs t, Mgh)-SB (m* Δs t, Mgh) (2)
ei(m, Mgh)=[Si(Δ t, Mgh)Si(2* Δs t, Mgh)…Si(m* Δs t, Mgh)] (3)
Wherein, Δ t is the collection period of road traffic state data;(m* Δs t) is m-th of road traffic state data
Collection period, 0≤m≤N, N represent the quantity of the transport information gathered daily;I represents i-th section, 1≤i≤p;STi(m*
Δ t, Mgh) represent mode MghUnder, (the highway traffic data in m* Δ t) moment i sections;SB (m* Δs t, Mgh) represent mode Mgh
Under, (the reference data in m* Δ t) moment benchmark section;Si(m* Δs t, Mgh) represent mode MghUnder, (m* Δ t) moment i sections
Training data and benchmark section reference data difference data;ei(m, Mgh) represent mode MghUnder, Δ t is to (during m* Δ t)
The training data and the difference data of the reference data in benchmark section in section i sections;Ei(m, Mgh) represent mode MghUnder, Δ t arrives
(the threshold value that m* Δ t) period i sections are chosen;hei(m, Mgh) represent mode MghUnder, Δ t to (m* Δ t) period thresholds processing after
Difference data, as training data feature;Represent the mapping rule of traffic state data and feature.
Further, in the step 3), drawn by the way that existing highway traffic data and road traffic service level are fuzzy
Divide, the feature of combined training data, complete the structure of knowledge base, its general expression is as follows:
Losm=φ (STi(m, Mgh)) (6)
Wherein, φ represents the mapping relations of current road traffic state and the level of service of blurring, LosmTable
Show the road traffic service level after blurring.
With reference to (5) (6), the relation between traffic circulation state and feature is obtained:
Losm=ω (hei(m, Mgh)) (7)
Wherein, ω represents the mapping rule between the service level and traffic circulation feature of blurring, so as to complete knowledge
The structure in storehouse.
Further, in the step 4), road traffic test data is extracted, based on the road traffic under same mode
Reference data, road traffic difference data is obtained, is handled by thresholding, obtains the feature of test data, its general expression
It is as follows:
MSj(m* Δs t, Mgh)=SMj(m* Δs t, Mgh)-SB (m* Δs t, Mgh) (8)
errj(m, Mgh)=[MSj(Δ t, Mgh)MSj(2* Δs t, Mgh)…MSj(m* Δs t, Mgh)] (9)
Wherein, j represents j-th strip section, 1≤i≤p;SMj(m* Δs t, Mgh) represent mode MghUnder, (m* Δ t) moment j roads
The test data of section;MSj(m* Δs t, Mgh) it is mode MghUnder, (test data in m* Δ t) moment j sections and benchmark section
The difference data of reference data;errj(m, Mgh) it is mode MghUnder, the Δ t to (test data and base in m* Δ t) period j sections
The difference data of the reference data in quasi- section.herrj(m, Mgh) difference data after threshold process, as test data
Feature.
Further, in the step 5) by KNN Feature Correspondence Algorithms, choose in knowledge base with test data feature
K closest feature, and the service level of corresponding blurring is obtained, comprise the following steps:
5.1) distance of the feature of training data and the feature of test data is calculated
Dist (m)=| | Thej(m, Mgh)-hei(m, Mgh) | | (10)
Dp(m)=[dist1(m)dist2(m)…distp(m)] (11)
Wherein, | | | | Euclidean distance is asked in expression, and dist (m) represents Δ t to the (feature of m* Δ t) period test datas
With the distance of feature in knowledge base, p represents the group number of training data, Dp(m) represent that the feature of j sections test data arrives in Δ t
(the characteristic distance set of p group training datas is arrived in the m* Δ t) periods.
5.2) feature corresponding to k minimum distance section is found out, if k feature is s1, s2... sk, according to feature and mould
The mapping relations of the service level of gelatinization, then have
(L1, L2... Lk)=ω (s1, s2... sk) (12)
Wherein, 1≤k≤p, L1, L2... LkS is represented respectively1, s2... skThe service level of corresponding blurring.Wherein,
Lk(m)=[uAk(STk(m, Mgh))uBk(STk(m, Mgh))uCk(STk(m, Mgh))]。uAk(STk(m, Mgh)), uBk(STk(m,
Mgh)), uCk(STk(m, Mgh)) mode M is represented respectivelyghUnder, Δ t is to (unimpeded, the general congestion in m* Δ t) period k sections, tight
The probability of congestion again.
In the step 6), the service level of k groups blurring is added, draws service level corresponding to maximum probability, i.e.,
For the affiliated service level of current road segment feature, the road traffic service level evaluation of space characteristics matching is completed, process is as follows:
The service level probability being blurred corresponding to k feature is added, then had
Wherein, SuAK, SuAK, SuAKCorresponding unimpeded, general congestion after comprehensive k feature, heavy congestion are represented respectively
Probability.SuAK, SuAK, SuAKService level corresponding to middle maximum probable value, the as affiliated service level of current signature.
The present invention technical concept be:Propose a kind of road traffic based on the matching of fuzzy KNN space characteristics and service water
Flat evaluation method, take full advantage of the spatial correlation characteristic that spatially different sections of highway has.Extract the road of spatially different sections of highway
Road traffic data, and it is divided into reference data, training data and test data.To training data and the difference of reference data
Data carry out thresholding processing, obtain the feature of training data.By road traffic service level Fuzzy processing, with reference to current
Traffic circulation state, structure training data feature and the knowledge base of road traffic service level composition.To test data and benchmark
The difference data of data carries out thresholding processing, obtains the feature of test data, asks for the feature and training data of test data
Euclidean distance between feature.Based on KNN algorithms, the service level being blurred corresponding to nearest k minimum distance is chosen.Will
The service level of blurring is added corresponding to k minimum distance, asks for service level corresponding to maximum probability, is current clothes
Business is horizontal, realizes the road traffic service level evaluation of space characteristics matching.
This method only subtracts processing between data, obtains traffic circulation feature, builds knowledge base, passes through feature
With fuzzy KNN algorithms are combined, the road traffic service level evaluation of space characteristics matching is realized.Method is realized simple, it is not necessary to
The data for carrying out large amount of complex calculate, and can effectively improve processing speed.
Beneficial effects of the present invention are mainly manifested in:By by same mode MghRoad traffic space training data and
The difference data of reference data carries out thresholding processing, the feature of training data is obtained, with reference to the road traffic service of blurring
Level, realize the structure of training data knowledge base;The feature of test data under same mode is obtained, is calculated and surveyed by KNN algorithms
Try the distance of data characteristics and training data feature.The service level for choosing the minimum corresponding blurring of k groups distance is carried out
It is added, the service level for choosing maximum probability is current road traffic service level, completes the road of space characteristics matching
The evaluation of level of service.
Brief description of the drawings
Fig. 1 is the schematic diagram of the time format of Traffic Information template.
Fig. 2 is road traffic operational modal division schematic diagram.
Fig. 3 is the road traffic service level evaluation rubric figure based on the matching of fuzzy KNN space characteristics.
Embodiment
The invention will be further described below in conjunction with the accompanying drawings.
1~Fig. 3 of reference picture, a kind of road traffic service level evaluation method based on the matching of fuzzy KNN space characteristics,
Comprise the following steps:
1) highway traffic data for obtaining different sections of highway under same mode, spatially establishes road traffic features reference
Sequence, based on spatial Correlation Analysis, select benchmark section, and using its data as road traffic reference data spatially,
Comprise the following steps:
1.1) division of road traffic operational modal
The division of road traffic operational modal is divided into two levels:Road network layer and section layer, the traffic fortune of setting road network layer
The traffic circulation mode of road is divided into g seed mode by the division mark of row mode, and the traffic circulation mode of section layer is drawn
Minute mark is known is divided into h seed mode by the traffic circulation mode of road, then the traffic circulation mode of road is divided into g × h altogether
Kind, it is designated as set Mode={ M11, M12..., Mgh, the division mark of wherein g and h value traffic circulation mode selected by
It is determined that;
1.2) structure of road traffic features reference sequences is designed
The collection period for setting road traffic state data is Δ t;
As shown in Table 1 and Table 2, table 1 is road traffic features reference sequences to the sheet format of road traffic features reference sequences
Information table, table 2 are that road traffic features reference sequences describe table:
Table 1
Table 2
P bars are selected in setting altogether has the section of spatial correlation characteristic, is designated as:
L=[L1L2…Lp] (1)
Wherein, p represents the section bar number on path space;LiRepresent i-th section, 1≤i≤p;L represents the tool of selection
There is the set in spatial correlation characteristic section;
2) extract under same mode, the historical data in spatially other sections, as training data, based on same mode
Under, road traffic reference data spatially, obtain difference data, handled by thresholding, obtain the feature of training data,
Its general expression is as follows:
Si(m* Δs t, Mgh)=STi(m* Δs t, Mgh)-SB (m* Δs t, Mgh) (2)
ei(m, Mgh)=[Si(Δ t, Mgh)Si(2* Δs t, Mgh)…Si(m* Δs t, Mgh)] (3)
Wherein, Δ t is the collection period of road traffic state data;(m* Δs t) is m-th of road traffic state data
Collection period, 0≤m≤N, N represent the quantity of the transport information gathered daily;I represents i-th section, 1≤i≤p;STi(m*
Δ t, Mgh) represent mode MghUnder, (the highway traffic data in m* Δ t) moment i sections;SB (m* Δs t, Mgh) represent mode Mgh
Under, (the reference data in m* Δ t) moment benchmark section;Si(m* Δs t, Mgh) represent mode MghUnder, (m* Δ t) moment i sections
Training data and benchmark section reference data difference data;ei(m, Mgh) represent mode MghUnder, Δ t is to (during m* Δ t)
The training data and the difference data of the reference data in benchmark section in section i sections;Ei(m, Mgh) represent mode MghUnder, Δ t arrives
(the threshold value that m* Δ t) period i sections are chosen;hei(m, Mgh) represent mode MghUnder, Δ t to (m* Δ t) period thresholds processing after
Difference data, as training data feature;Represent the mapping rule of traffic state data and feature.
3) existing highway traffic data and road traffic service level fuzzy division, the spy of combined training data are passed through
Sign, completes the structure of knowledge base, and its general expression is as follows:
Losm=φ (STi(m, Mgh)) (6)
Wherein, φ represents the mapping relations of current road traffic state and the level of service of blurring, LosmTable
Show the road traffic service level after blurring.
With reference to (5) (6), the relation between traffic circulation state and feature is obtained:
Losm=ω (hei(m, Mgh)) (7)
Wherein, ω represents the mapping rule between the service level and traffic circulation feature of blurring, so as to complete knowledge
The structure in storehouse.
4) road traffic test data is extracted, based on the road traffic reference data under same mode, obtains road traffic
Difference data, handled by thresholding, obtain the feature of test data, its general expression is as follows:
MSj(m* Δs t, Mgh)=SMj(m* Δs t, Mgh)-SB (m* Δs t, Mgh) (8)
errj(m, Mgh)=[MSj(Δ t, Mgh)MSj(2* Δs t, Mgh)…MSj(m* Δs t, Mgh)] (9)
Wherein, j represents j-th strip section, 1≤i≤p;SMj(m* Δs t, Mgh) represent mode MghUnder, (m* Δ t) moment j roads
The test data of section;MSj(m* Δs t, Mgh) it is mode MghUnder, (test data in m* Δ t) moment j sections and benchmark section
The difference data of reference data;errj(m, Mgh) it is mode MghUnder, the Δ t to (test data and base in m* Δ t) period j sections
The difference data of the reference data in quasi- section.herrj(m, Mgh) difference data after threshold process, as test data
Feature.
5) by KNN Feature Correspondence Algorithms, k feature nearest with test data characteristic distance in knowledge base is chosen, and
The service level of blurring, comprises the following steps corresponding to obtaining:5.1) feature of training data and the spy of test data are calculated
The distance of sign
Dist (m)=| | Thej(m, Mgh)-hei(m, Mgh) | | (10)
Dp(m)=[dist1(m)dist2(m)…distp(m)] (11)
Wherein, | | | | Euclidean distance is asked in expression, and dist (m) represents Δ t to the (feature of m* Δ t) period test datas
With the distance of feature in knowledge base, p represents the group number of training data, Dp(m) represent that the feature of j sections test data arrives in Δ t
(the characteristic distance set of p group training datas is arrived in the m* Δ t) periods.
5.2) feature corresponding to k minimum distance section is found out, if k feature is s1, s2... sk, according to feature and mould
The mapping relations of the service level of gelatinization, then have
(L1, L2... Lk)=ω (s1, s2... sk) (12)
Wherein, 1≤k≤p, L1, L2... LkS is represented respectively1, s2... skThe service level of corresponding blurring.Wherein,
Lk(m)=[uAk(STk(m, Mgh))uBk(STk(m, Mgh))uCk(STk(m, Mgh))]。uAk(STk(m, Mgh)), uBk(STk(m,
Mgh)), uCk(STk(m, Mgh)) mode M is represented respectivelyghUnder, Δ t is to (unimpeded, the general congestion in m* Δ t) period k sections, tight
The probability of congestion again.
In the step 6), the service level of k groups blurring is added, draws service level corresponding to maximum probability, i.e.,
For the affiliated service level of current road segment feature, the road traffic service level evaluation of space characteristics matching is completed, process is as follows:
The service level probability being blurred corresponding to k feature is added, then had
Wherein, SuAK, SuAK, SuAKCorresponding unimpeded, general congestion after comprehensive k feature, heavy congestion are represented respectively
Probability.SuAK, SuAK, SuAKService level corresponding to middle maximum probable value, the as affiliated service level of current signature.
Example:A kind of road traffic service level evaluation method based on the matching of fuzzy KNN space characteristics, including it is following
Step:
1) highway traffic data for obtaining different sections of highway under same mode, spatially establishes road traffic features reference
Sequence, based on spatial Correlation Analysis, benchmark section is selected, and using its data as road traffic reference data spatially;
Typical 6 sections in Beijing are selected to carry out analysis of cases, specific section title is as shown in table 3 below.
Road section ID | Section title |
HI2075a | Central Conservatory of Music-Western Informal Gate bridge |
HI7000d | Side street bridge east -- Dongzhimen Qiao Bei entrances |
HI7008a | Bai Qiao street-Wide Channel Gate bridge |
HI7035d | Ponding Tan Qiao -- the Desheng raft of pontoons |
HI7060a | Wide straight raft of pontoons north entrance-side street bridge east |
HI7051a | The Temple of Moon north bridge-the Temple of Moon south bridge |
Table 3
Using HI3006b sections as benchmark section, the highway traffic data of collection is as reference data;By HI7000d,
The highway traffic data of the same mode of HI7008a, HI7035d, HI7060a section collection is as training data;Will
The highway traffic data of the same mode of HI7051a sections collection carries out proof of algorithm as real time data.
Further, in the step 2), extract under same mode, the historical data in spatially other sections, as instruction
Practice data, based on road traffic reference data under same mode, spatially, obtain difference data, handled by thresholding,
The feature of training data is obtained, its general expression is as follows:
Si(m* Δs t, Mgh)=STi(m* Δs t, Mgh)-SB (m* Δs t, Mgh) (1)
ei(m, Mgh)=[Si(Δ t, Mgh)Si(2* Δs t, Mgh)…Si(m* Δs t, Mgh)] (2)
Wherein, Δ t is the collection period of road traffic state data;(m* Δs t) is m-th of road traffic state data
Collection period, 0≤m≤N, N represent the quantity of the transport information gathered daily;I represents i-th section, 1≤i≤p;STi(m*
Δ t, Mgh) represent mode MghUnder, (the highway traffic data in m* Δ t) moment i sections;SB (m* Δs t, Mgh) represent mode Mgh
Under, (the reference data in m* Δ t) moment benchmark section;Si(m* Δs t, Mgh) represent mode MghUnder, (m* Δ t) moment i sections
Training data and benchmark section reference data difference data;ei(m, Mgh) represent mode MghUnder, Δ t is to (during m* Δ t)
The training data and the difference data of the reference data in benchmark section in section i sections;Ei(m, Mgh) represent mode MghUnder, Δ t arrives
(the threshold value that m* Δ t) period i sections are chosen;hei(m, Mgh) represent mode MghUnder, Δ t to (m* Δ t) period thresholds processing after
Difference data, as training data feature;Represent the mapping rule of traffic state data and feature.
Further, in the step 3), drawn by the way that existing highway traffic data and road traffic service level are fuzzy
Divide, the feature of combined training data, complete the structure of knowledge base, its general expression is as follows:
Losm=φ (STi(m, Mgh)) (5)
Wherein, φ represents the mapping relations of current road traffic state and the level of service of blurring, LosmTable
Show the road traffic service level after blurring.
With reference to (5) (6), the relation between traffic circulation state and feature is obtained:
Losm=ω (hei(m, Mgh)) (6)
Wherein, ω represents the mapping rule between the service level and traffic circulation feature of blurring, so as to complete knowledge
The structure in storehouse.
Further, in the step 4), road traffic test data is extracted, based on the road traffic under same mode
Reference data, road traffic difference data is obtained, is handled by thresholding, obtains the feature of test data, its general expression
It is as follows:
MSj(m* Δs t, Mgh)=SMj(m* Δs t, Mgh)-SB (m* Δs t, Mgh) (7)
errj(m, Mgh)=[MSj(Δ t, Mgh)MSj(2* Δs t, Mgh)…MSj(m* Δs t, Mgh)] (8)
Wherein, j (1≤i≤p) represents j-th strip section;SMj(m* Δs t, Mgh) represent mode MghUnder, (m* Δ t) moment j
The test data in section;MSj(m* Δs t, Mgh) it is mode MghUnder, (test data in m* Δ t) moment j sections and benchmark section
Reference data difference data;errj(m, Mgh) it is mode MghUnder, Δ t to (test data in m* Δ t) period j sections with
The difference data of the reference data in benchmark section.herrj(m, Mgh) difference data after threshold process, as test number
According to feature.
Further, in the step 5) by KNN Feature Correspondence Algorithms, choose in knowledge base with test data feature
K closest feature, and the service level of corresponding blurring is obtained, comprise the following steps:
5.1) distance of the feature of training data and the feature of test data is calculated
Dist (m)=| | Thej(m, Mgh)-hei(m, Mgh) | | (10)
Dp(m)=[dist1(m)dist2(m)…distp(m)] (11)
Wherein, | | | | Euclidean distance is asked in expression, and dist (m) represents Δ t to the (feature of m* Δ t) period test datas
With the distance of feature in knowledge base, p represents the group number of training data, Dp(m) represent that the feature of j sections test data arrives in Δ t
(the characteristic distance set of p group training datas is arrived in the m* Δ t) periods.
5.2) feature corresponding to k minimum distance section is found out, if k feature is s1, s2... sk, according to feature and mould
The mapping relations of the service level of gelatinization, then have
(L1, L2... Lk)=ω (s1, s2... sk) (12)
Wherein, 1≤k≤p, L1, L2... LkS is represented respectively1, s2... skThe service level of corresponding blurring.Wherein,
Lk(m)=[uAk(STk(m, Mgh))uBk(STk(m, Mgh))uCk(STk(m, Mgh))]。uAk(STk(m, Mgh)), uBk(STk(m,
Mgh)), uCk(STk(m, Mgh)) mode M is represented respectivelyghUnder, Δ t is to (unimpeded, the general congestion in m* Δ t) period k sections, tight
The probability of congestion again.
Service level in the step 6) by the blurring of k groups is added, and draws service level corresponding to maximum probability, i.e.,
For the affiliated service level of current road segment feature, the road traffic service level evaluation of space characteristics matching is completed, process is as follows:
The service level probability being blurred corresponding to k feature is added, then had
Wherein, SuAK, SuAK, SuAKCorresponding unimpeded, general congestion after comprehensive k feature, heavy congestion are represented respectively
Probability.SuAK, SuAK, SuAKService level corresponding to middle maximum probable value, the as affiliated service level of current signature.
6) the road traffic service level evaluating based on the matching of fuzzy KNN space characteristics determines
In the space road traffic service level evaluation procedure based on fuzzy KNN characteristic matchings, it is related to having following several
Individual parameter:Ei(m,Mgh)、Ej(m,Mgh), k, wherein, Ei(m,Mgh) can be by STi(m* Δs t, Mgh) and SB (m* Δs t, Mgh) really
It is fixed, Ej(m,Mgh) value choose and Ei(m,Mgh) equal, k values typically take 5.Here the parameter setting done is simply to obscuring KNN
The general impact analysis of the road traffic service level evaluation of character matching.
Because these parameters respectively have an impact to the precision of algorithm, influence of each parameter to arithmetic accuracy is individually analyzed simultaneously
The optimal of algorithm is cannot ensure, therefore should consider that all parameters service water to the road traffic simultaneously when carrying out Algorithm Analysis
The influence of flat evaluation.
The road traffic service level evaluation index that deviation ratio matches as space characteristics is introduced, its calculation formula is as follows:
Wherein, PE represents deviation ratio, Nsr≠smRepresent test and true different level of service quantity, NsrRepresent real
The level of service number of border test.
I.e. for different (Ei(m,Mgh), Ej(m,Mgh), k), corresponding NMAE be present.Therefore following equation be present:
NMAE=f ((Ei(m,Mgh), Ej(m,Mgh), k)) (15)
That is (Ei(m,Mgh), Ej(m,Mgh), k) with NMAE certain distribution relation f be present, it is corresponding during searching NMAE minimums
((Ei(m,Mgh), Ej(m,Mgh), k)), as optimized parameter sets process.Therefore it can obtain such as drag:
Min f((Ei(m,Mgh), Ej(m,Mgh), k)) (16)
Finally ((Ei(m,Mgh), Ej(m,Mgh), k)) value can pass through road traffic reference data and training data
Training determine.
7) experimental result
Road traffic reference data and training data based on same mode, obtain optimized parameter (Ei(m,Mgh), Ej(m,
Mgh), k).This experimental result carries out road traffic service level evaluation mainly for the vehicle velocity value in section.Extract road traffic
Test data, the evaluation of test mode is realized based on KNN algorithms.
Deviation ratio is chosen as road traffic service level evaluation index, it is calculated as shown in formula (13).Test section
June 25 in 2011,26,27 days, the deviation statistics of velocity amplitude were analyzed as follows shown in table 4.
Road section ID | Time | Deviation ratio |
HI7051a | 2011-6-25 | 5.00% |
HI7051a | 2011-6-26 | 3.33% |
HI7051a | 2011-6-27 | 10.00% |
Table 4.
Claims (7)
- A kind of 1. road traffic service level evaluation method based on the matching of fuzzy KNN space characteristics, it is characterised in that:The side Method comprises the following steps:1) highway traffic data for obtaining different sections of highway under same mode, spatially establishes road traffic features reference sequences, Based on spatial Correlation Analysis, benchmark section is selected, and using its data as road traffic reference data spatially;2) extract under same mode, the historical data in spatially other sections, as training data, based under same mode, it is empty Between on road traffic reference data, obtain difference data, handled by thresholding, obtain the feature of training data;3) divided by existing highway traffic data and road traffic service level, the feature of combined training data, complete to know Know the structure in storehouse;4) obtain under same mode, the test data in spatially other sections, as test data, based under same mode, it is empty Between on road traffic reference data, obtain road traffic difference data spatially, handled by thresholding, obtain test number According to feature;5) by KNN Feature Correspondence Algorithms, k feature nearest with test data characteristic distance in knowledge base is chosen, and obtain The service level of corresponding blurring;6) service level by the blurring of k groups is added, and is drawn service level corresponding to maximum probability, is as taken belonging to current signature Business is horizontal, completes the road traffic service level evaluation of space characteristics matching.
- 2. a kind of road traffic service level evaluation method based on the matching of fuzzy KNN space characteristics as claimed in claim 1, It is characterized in that:In the step 1), the highway traffic data for obtaining different sections of highway under same mode, spatially establishes road Traffic characteristic reference sequences, based on spatial Correlation Analysis, benchmark section is selected, and hand over its data as road spatially Logical reference data, comprises the following steps:1.1) division of road traffic operational modalThe division of road traffic operational modal is divided into two levels:Road network layer and section layer, set the traffic circulation mould of road network layer The traffic circulation mode of road is divided into g seed mode, the division mark of the traffic circulation mode of section layer by the division mark of state Know and the traffic circulation mode of road is divided into h seed mode, then the traffic circulation mode of road is divided into g × h kinds altogether, remembers For set Mode={ M11, M12..., Mgh, the division mark of wherein g and h value traffic circulation mode selected by determines;1.2) structure of road traffic features reference sequences is designedThe collection period for setting road traffic state data is Δ t;As shown in Table 1 and Table 2, table 1 is road traffic features reference sequences information to the sheet format of road traffic features reference sequences Table, table 2 are that road traffic features reference sequences describe table:Table 1Table 2P bars are selected in setting altogether has the section of spatial correlation characteristic, is designated as:L=[L1L2…Lp] (1)Wherein, p represents the section bar number on path space;LiRepresent i-th section, 1≤i≤p;What L represented selection has space The set in correlation properties section.
- 3. a kind of road traffic service level evaluation method based on the matching of fuzzy KNN space characteristics as claimed in claim 2, It is characterized in that:In the step 2), extract under same mode, the historical data in spatially other sections, as training data, Based on road traffic reference data under same mode, spatially, difference data is obtained, is handled by thresholding, obtain training The feature of data, its general expression are as follows:Si(m* Δs t, Mgh)=STi(m* Δs t, Mgh)-SB (m* Δs t, Mgh) (2)ei(m, Mgh)=[Si(Δ t, Mgh)Si(2* Δs t, Mgh)…Si(m* Δs t, Mgh)] (3)<mrow> <msub> <mi>he</mi> <mi>i</mi> </msub> <mrow> <mo>(</mo> <mi>m</mi> <mo>,</mo> <msub> <mi>M</mi> <mrow> <mi>g</mi> <mi>h</mi> </mrow> </msub> <mo>)</mo> </mrow> <mo>=</mo> <mfenced open = "{" close = ""> <mtable> <mtr> <mtd> <mrow> <mn>0</mn> <mo>,</mo> </mrow> </mtd> <mtd> <mrow> <msub> <mi>e</mi> <mi>i</mi> </msub> <mrow> <mo>(</mo> <mi>m</mi> <mo>,</mo> <msub> <mi>M</mi> <mrow> <mi>g</mi> <mi>h</mi> </mrow> </msub> <mo>)</mo> </mrow> <mo><</mo> <msub> <mi>E</mi> <mi>i</mi> </msub> <mrow> <mo>(</mo> <mi>m</mi> <mo>,</mo> <msub> <mi>M</mi> <mrow> <mi>g</mi> <mi>h</mi> </mrow> </msub> <mo>)</mo> </mrow> </mrow> </mtd> </mtr> <mtr> <mtd> <mrow> <msub> <mi>e</mi> <mi>i</mi> </msub> <mrow> <mo>(</mo> <mi>m</mi> <mo>,</mo> <msub> <mi>M</mi> <mrow> <mi>g</mi> <mi>h</mi> </mrow> </msub> <mo>)</mo> </mrow> <mo>,</mo> </mrow> </mtd> <mtd> <mrow> <msub> <mi>e</mi> <mi>i</mi> </msub> <mrow> <mo>(</mo> <mi>m</mi> <mo>,</mo> <msub> <mi>M</mi> <mrow> <mi>g</mi> <mi>h</mi> </mrow> </msub> <mo>)</mo> </mrow> <mo>></mo> <msub> <mi>E</mi> <mi>i</mi> </msub> <mrow> <mo>(</mo> <mi>m</mi> <mo>,</mo> <msub> <mi>M</mi> <mrow> <mi>g</mi> <mi>h</mi> </mrow> </msub> <mo>)</mo> </mrow> </mrow> </mtd> </mtr> </mtable> </mfenced> <mo>-</mo> <mo>-</mo> <mo>-</mo> <mrow> <mo>(</mo> <mn>4</mn> <mo>)</mo> </mrow> </mrow>Wherein, Δ t is the collection period of road traffic state data;(m* Δs t) is m-th of road traffic state data acquisition week Phase, 0≤m≤N, N represent the quantity of the transport information gathered daily;I represents i-th section, 1≤i≤p;STi(m* Δs t, Mgh) Represent mode MghUnder, (the highway traffic data in m* Δ t) moment i sections;SB (m* Δs t, Mgh) represent mode MghUnder, (m* Δs t) The reference data in moment benchmark section;Si(m* Δs t, Mgh) represent mode MghUnder, (training data in m* Δ t) moment i sections with The difference data of the reference data in benchmark section;ei(m, Mgh) represent mode MghUnder, the Δ t to (training in m* Δ t) period i sections The difference data of data and the reference data in benchmark section;Ei(m, Mgh) represent mode MghUnder, Δ t to (m* Δ t) period i sections The threshold value of selection;hei(m, Mgh) represent mode MghUnder, Δ t to (m* Δ t) period thresholds processing after difference data, as instruct Practice the feature of data;Represent the mapping rule of traffic state data and feature.
- 4. a kind of road traffic service level evaluation method based on the matching of fuzzy KNN space characteristics as claimed in claim 3, It is characterized in that:In the step 3), by existing highway traffic data and road traffic service level fuzzy division, with reference to The feature of training data, the structure of knowledge base is completed, its general expression is as follows:Wherein,Represent the mapping relations of current road traffic state and the level of service of blurring, LosmRepresent fuzzy Road traffic service level after change.With reference to (5) (6), the relation between traffic circulation state and feature is obtained:Losm=ω (hei(m, Mgh)) (7)Wherein, ω represents the mapping rule between the service level and traffic circulation feature of blurring, so as to complete knowledge base Structure.
- 5. a kind of space road traffic service level evaluation method based on fuzzy KNN characteristic matchings as claimed in claim 4, It is characterized in that:In the step 4), road traffic test data is extracted, based on the road traffic base value under same mode According to, obtain road traffic difference data, handled by thresholding, obtain the feature of test data, its general expression is as follows:MSj(m* Δs t, Mgh)=SMj(m* Δs t, Mgh)-SB (m* Δs t, Mgh) (8)errj(m, Mgh)=[MSj(Δ t, Mgh)MSj(2* Δs t, Mgh)…MSj(m* Δs t, Mgh)] (9)<mrow> <msub> <mi>herr</mi> <mi>j</mi> </msub> <mrow> <mo>(</mo> <mrow> <mi>m</mi> <mo>,</mo> <msub> <mi>M</mi> <mrow> <mi>g</mi> <mi>h</mi> </mrow> </msub> </mrow> <mo>)</mo> </mrow> <mo>=</mo> <mfenced open = "{" close = ""> <mtable> <mtr> <mtd> <mrow> <mn>0</mn> <mo>,</mo> </mrow> </mtd> <mtd> <mrow> <msub> <mi>err</mi> <mi>j</mi> </msub> <mrow> <mo>(</mo> <mrow> <mi>m</mi> <mo>,</mo> <msub> <mi>M</mi> <mrow> <mi>g</mi> <mi>h</mi> </mrow> </msub> </mrow> <mo>)</mo> </mrow> <mo><</mo> <msub> <mi>E</mi> <mi>j</mi> </msub> <mrow> <mo>(</mo> <mrow> <mi>m</mi> <mo>,</mo> <msub> <mi>M</mi> <mrow> <mi>g</mi> <mi>h</mi> </mrow> </msub> </mrow> <mo>)</mo> </mrow> </mrow> </mtd> </mtr> <mtr> <mtd> <mrow> <msub> <mi>err</mi> <mi>j</mi> </msub> <mrow> <mo>(</mo> <mrow> <mi>m</mi> <mo>,</mo> <msub> <mi>M</mi> <mrow> <mi>g</mi> <mi>h</mi> </mrow> </msub> </mrow> <mo>)</mo> </mrow> <mo>,</mo> </mrow> </mtd> <mtd> <mrow> <msub> <mi>err</mi> <mi>j</mi> </msub> <mrow> <mo>(</mo> <mrow> <mi>m</mi> <mo>,</mo> <msub> <mi>M</mi> <mrow> <mi>g</mi> <mi>h</mi> </mrow> </msub> </mrow> <mo>)</mo> </mrow> <mo>></mo> <msub> <mi>E</mi> <mi>j</mi> </msub> <mrow> <mo>(</mo> <mrow> <mi>m</mi> <mo>,</mo> <msub> <mi>M</mi> <mrow> <mi>g</mi> <mi>h</mi> </mrow> </msub> </mrow> <mo>)</mo> </mrow> </mrow> </mtd> </mtr> </mtable> </mfenced> <mo>-</mo> <mo>-</mo> <mo>-</mo> <mrow> <mo>(</mo> <mn>10</mn> <mo>)</mo> </mrow> </mrow>Wherein, j represents j-th strip section, 1≤i≤p;SMj(m* Δs t, Mgh) represent mode MghUnder, (m* Δ t) moment j sections Test data;MSj(m* Δs t, Mgh) it is mode MghUnder, (test data in m* Δ t) moment j sections and the base value in benchmark section According to difference data;errj(m, Mgh) it is mode MghUnder, Δ t to (test data in m* Δ t) period j sections and benchmark section The difference data of reference data.herrj(m, Mgh) difference data, as test data after threshold process feature.
- 6. a kind of road traffic service level evaluation method based on the matching of fuzzy KNN space characteristics as claimed in claim 5, It is characterized in that:In the step 5), by KNN Feature Correspondence Algorithms, choose in knowledge base with test data characteristic distance most K near feature, and the service level of corresponding blurring is obtained, comprise the following steps:6.1) distance of the feature of training data and the feature of test data is calculatedDist (m)=| | Thej(m, Mgh)-hei(m, Mgh) | | (11)Dp(m)=[dist1(m)dist2(m)…distp(m)] (12)Wherein, | | | | Euclidean distance is asked in expression, and dist (m) represents Δ t to the (feature and knowledge of m* Δ t) period test datas The distance of feature in storehouse, p represent the group number of training data, Dp(m) represent the feature of j sections test data in Δ t to (m* Δs t) The characteristic distance set of p group training datas is arrived in period;6.2) feature corresponding to k minimum distance section is found out, if k feature is s1, s2... sk, according to feature and blurring The mapping relations of service level, then have(L1, L2... Lk)=ω (s1, s2... sk) (13)Wherein, 1≤k≤p, L1, L2... LkS is represented respectively1, s2... skThe service level of corresponding blurring.Wherein, Lk(m)= [uAk(STk(m, Mgh))uBk(STk(m, Mgh))uCk(STk(m, Mgh))]。uAk(STk(m, Mgh)), uBk(STk(m, Mgh)), uCk (STk(m, Mgh)) mode M is represented respectivelyghUnder, Δ t to (unimpeded, the general congestion in m* Δ t) period k sections, heavy congestion it is general Rate.
- 7. a kind of road traffic service level evaluation method based on the matching of fuzzy KNN space characteristics as claimed in claim 6, It is characterized in that:In the step 6), the service level of k groups blurring is added, draws service level corresponding to maximum probability, The as affiliated service level of current road segment feature, complete the road traffic service level evaluation based on the matching of KNN space characteristics, mistake Journey is as follows:The service level probability being blurred corresponding to k feature is added, then had<mrow> <mfenced open = "[" close = "]"> <mtable> <mtr> <mtd> <mrow> <msub> <mi>Su</mi> <mrow> <mi>A</mi> <mi>k</mi> </mrow> </msub> </mrow> </mtd> <mtd> <mrow> <msub> <mi>Su</mi> <mrow> <mi>B</mi> <mi>k</mi> </mrow> </msub> </mrow> </mtd> <mtd> <mrow> <msub> <mi>Su</mi> <mrow> <mi>C</mi> <mi>k</mi> </mrow> </msub> </mrow> </mtd> </mtr> </mtable> </mfenced> <mo>=</mo> <mfenced open = "[" close = "]"> <mtable> <mtr> <mtd> <mrow> <munderover> <mi>&Sigma;</mi> <mrow> <mi>i</mi> <mo>=</mo> <mn>1</mn> </mrow> <mi>k</mi> </munderover> <msub> <mi>u</mi> <mrow> <mi>A</mi> <mi>i</mi> </mrow> </msub> <mrow> <mo>(</mo> <mrow> <msub> <mi>ST</mi> <msub> <mi>M</mi> <mrow> <mi>g</mi> <mi>h</mi> </mrow> </msub> </msub> <mrow> <mo>(</mo> <mi>m</mi> <mo>)</mo> </mrow> </mrow> <mo>)</mo> </mrow> </mrow> </mtd> <mtd> <mrow> <munderover> <mi>&Sigma;</mi> <mrow> <mi>i</mi> <mo>=</mo> <mn>1</mn> </mrow> <mi>k</mi> </munderover> <msub> <mi>u</mi> <mrow> <mi>B</mi> <mi>i</mi> </mrow> </msub> <mrow> <mo>(</mo> <mrow> <msub> <mi>ST</mi> <msub> <mi>M</mi> <mrow> <mi>g</mi> <mi>h</mi> </mrow> </msub> </msub> <mrow> <mo>(</mo> <mi>m</mi> <mo>)</mo> </mrow> </mrow> <mo>)</mo> </mrow> </mrow> </mtd> <mtd> <mrow> <munderover> <mi>&Sigma;</mi> <mrow> <mi>i</mi> <mo>=</mo> <mn>1</mn> </mrow> <mi>k</mi> </munderover> <msub> <mi>u</mi> <mrow> <mi>C</mi> <mi>i</mi> </mrow> </msub> <mrow> <mo>(</mo> <mrow> <msub> <mi>ST</mi> <msub> <mi>M</mi> <mrow> <mi>g</mi> <mi>h</mi> </mrow> </msub> </msub> <mrow> <mo>(</mo> <mi>m</mi> <mo>)</mo> </mrow> </mrow> <mo>)</mo> </mrow> </mrow> </mtd> </mtr> </mtable> </mfenced> <mo>-</mo> <mo>-</mo> <mo>-</mo> <mrow> <mo>(</mo> <mn>14</mn> <mo>)</mo> </mrow> </mrow>Wherein, SuAK, SuAK, SuAKCorresponding unimpeded, general congestion, the probability of heavy congestion after the comprehensive k feature of expression respectively. SuAK, SuAK, SuAKService level corresponding to middle maximum probable value, the as affiliated service level of current signature.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201710850376.5A CN107767023A (en) | 2017-09-20 | 2017-09-20 | A kind of road traffic service level evaluation method based on the matching of fuzzy KNN space characteristics |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201710850376.5A CN107767023A (en) | 2017-09-20 | 2017-09-20 | A kind of road traffic service level evaluation method based on the matching of fuzzy KNN space characteristics |
Publications (1)
Publication Number | Publication Date |
---|---|
CN107767023A true CN107767023A (en) | 2018-03-06 |
Family
ID=61266051
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201710850376.5A Pending CN107767023A (en) | 2017-09-20 | 2017-09-20 | A kind of road traffic service level evaluation method based on the matching of fuzzy KNN space characteristics |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN107767023A (en) |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110021178A (en) * | 2019-04-18 | 2019-07-16 | 浙江工业大学 | A kind of signal time distributing conception recommended method based on crossing similarity |
CN113377828A (en) * | 2021-05-07 | 2021-09-10 | 浙江工业大学 | Congestion-aware road network moving object continuous K nearest neighbor query method |
Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101866143A (en) * | 2009-04-14 | 2010-10-20 | 北京宏德信智源信息技术有限公司 | Road traffic service level prediction method based on space-time characteristic aggregation |
CN102982081A (en) * | 2012-10-31 | 2013-03-20 | 公安部道路交通安全研究中心 | Method and system for screening potential traffic safety hazard road sections |
CN106651083A (en) * | 2016-06-29 | 2017-05-10 | 东南大学 | Pedestrian-non-motor vehicle isolation facility arrangement method for urban road segment |
CN107067727A (en) * | 2017-06-07 | 2017-08-18 | 浙江工业大学 | A kind of road traffic service level evaluation method based on fuzzy KNN characteristic matchings |
-
2017
- 2017-09-20 CN CN201710850376.5A patent/CN107767023A/en active Pending
Patent Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101866143A (en) * | 2009-04-14 | 2010-10-20 | 北京宏德信智源信息技术有限公司 | Road traffic service level prediction method based on space-time characteristic aggregation |
CN102982081A (en) * | 2012-10-31 | 2013-03-20 | 公安部道路交通安全研究中心 | Method and system for screening potential traffic safety hazard road sections |
CN106651083A (en) * | 2016-06-29 | 2017-05-10 | 东南大学 | Pedestrian-non-motor vehicle isolation facility arrangement method for urban road segment |
CN107067727A (en) * | 2017-06-07 | 2017-08-18 | 浙江工业大学 | A kind of road traffic service level evaluation method based on fuzzy KNN characteristic matchings |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110021178A (en) * | 2019-04-18 | 2019-07-16 | 浙江工业大学 | A kind of signal time distributing conception recommended method based on crossing similarity |
CN113377828A (en) * | 2021-05-07 | 2021-09-10 | 浙江工业大学 | Congestion-aware road network moving object continuous K nearest neighbor query method |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
Li et al. | Prediction of urban human mobility using large-scale taxi traces and its applications | |
CN102799897B (en) | Computer recognition method of GPS (Global Positioning System) positioning-based transportation mode combined travelling | |
CN102693633B (en) | Short-term traffic flow weighted combination prediction method | |
CN104021671B (en) | The determination methods of the road real-time road that a kind of svm combines with fuzzy Judgment | |
CN103208034B (en) | A kind of track traffic for passenger flow forecast of distribution model is set up and Forecasting Methodology | |
CN104573116B (en) | The traffic abnormity recognition methods excavated based on GPS data from taxi | |
CN108171993B (en) | Highway vehicle speed calculation method based on mobile phone signaling big data | |
CN109410577A (en) | Adaptive traffic control sub-area division method based on Spatial Data Mining | |
CN103903441B (en) | Road traffic state distinguishing method based on semi-supervised learning | |
CN106781479A (en) | A kind of method for obtaining highway running status in real time based on mobile phone signaling data | |
CN106710208A (en) | Traffic state acquisition method and device | |
CN107195180A (en) | A kind of traffic trip track extraction method and device based on the alert data of electricity | |
CN106228808A (en) | City expressway travel time prediction method based on Floating Car space-time grid data | |
CN103198104A (en) | Bus station origin-destination (OD) obtaining method based on urban advanced public transportation system | |
CN108597227A (en) | Road traffic flow forecasting method under freeway toll station | |
CN104900057B (en) | A kind of Floating Car map-matching method in the major-minor road of city expressway | |
CN105427001A (en) | Optimal route of school bus of regional middle and primary school | |
CN107134137A (en) | A kind of Dynamic User-Optimal Route Choice method for considering real time information | |
CN106777169A (en) | A kind of user's trip hobby analysis method based on car networking data | |
CN111063204A (en) | Expressway vehicle speed prediction model training method based on toll station flow | |
Zhou | Attention based stack resnet for citywide traffic accident prediction | |
CN107067727B (en) | A kind of road traffic service level evaluation method based on fuzzy KNN characteristic matching | |
CN107230350A (en) | A kind of urban transportation amount acquisition methods based on bayonet socket Yu mobile phone flow call bill data | |
Zhu et al. | Green routing fuel saving opportunity assessment: A case study using large-scale real-world travel data | |
CN106709833A (en) | School bus path optimization method based on big data |
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: 20180306 |