CN105160866A - Traffic flow prediction method based on deep learning nerve network structure - Google Patents
Traffic flow prediction method based on deep learning nerve network structure Download PDFInfo
- Publication number
- CN105160866A CN105160866A CN201510478215.9A CN201510478215A CN105160866A CN 105160866 A CN105160866 A CN 105160866A CN 201510478215 A CN201510478215 A CN 201510478215A CN 105160866 A CN105160866 A CN 105160866A
- Authority
- CN
- China
- Prior art keywords
- traffic flow
- data
- degree
- flow data
- traffic
- 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
The invention discloses a traffic flow prediction method based on a deep learning nerve network structure. Various kinds of traffic flow data are collected, a depth automatic encoder model is utilized for training the collected various kinds of traffic flow data, the depth automatic encoder model is adjusted in the training process, and finally, the adjusted depth automatic encoder model is used for predicting a short-period traffic flow. By adopting the method, deeper excavation analysis is carried out on traffic flow data, so that the short-period prediction of the traffic flow is more accurate, and the performance is better.
Description
Technical field
The present invention relates to a kind of traffic flow forecasting method based on degree of deep learning neural network structure, belong to the technical field of traffic forecast.
Background technology
Along with the development of China's automobile industry, city and Expressway Road congestion problems increasingly serious.By deeply excavating traffic flow data, and setting up short-term traffic flow forecasting model on this basis, effectively can predict traffic congestion thus guided vehicle selects reasonable traffic path.
The traffic following on certain section of short-time traffic flow forecast or certain paths, the time interval does not generally exceed 15 minutes.This prediction supvr in order to work out and to implement traffic management scheme, can be regulated and controled traffic flow, with the traffic congestion that may occur during slowing down this and potential safety hazard.Relative to Long-run Forecasting Analysis, short-time traffic flow forecast will use as historical data.
Short-time Traffic Flow Forecasting Methods mainly comprises two parts: the foundation of traffic flow historical data java standard library and forecast model build.The former provides Data safeguard for short-time traffic flow forecast, and the prediction of the latter's future transportation state provides method accurately.Up till now, a series of technology and means have been had to be developed the prediction being applied to Short-term Traffic Flow, but the not satisfied problem of result.
Traditional traffic flow data Forecasting Methodology mainly contains the classification of K-arest neighbors, support vector machine, BP neural network and decision tree etc., and these methods carry out modeling on the traffic flow data basis of certain scale, thus predict, achieve certain prediction effect.But along with the arrival of large data age, this kind of shallow-layer model of such as support vector machine is difficult to effectively to excavate the implicit information under mass data, cannot obtain optimum performance.
Summary of the invention
The object of the invention is to, a kind of traffic flow forecasting method based on degree of deep learning neural network structure is provided.Method of the present invention can carry out more deep mining analysis to traffic flow data, and therefore more accurate when carrying out short-term forecasting to traffic flow, performance is more excellent.
Technical scheme of the present invention: a kind of traffic flow forecasting method based on degree of deep learning neural network structure, be characterized in: by gathering all kinds of traffic flow data, and utilize degree of depth autocoder model to train all kinds of traffic flow data of collection, adjust degree of depth autocoder model in the training process, the degree of depth autocoder model after finally utilizing adjustment is predicted Short-term Traffic Flow.
The above-mentioned traffic flow forecasting method based on degree of deep learning neural network structure, specifically comprises the following steps:
1. the collection of traffic flow data;
2. the pre-service of traffic flow data;
3. degree of depth autocoder model is utilized to train traffic flow data; (degree of depth autocoder model is made up of scrambler, demoder and hidden layer)
4. with supervised learning algorithm, degree of depth autocoder model is finely tuned;
5. the final traffic flow of degree of depth autocoder model to short-term 4. obtained according to step is predicted.
In the aforesaid traffic flow forecasting method based on degree of deep learning neural network structure, described step 1. in traffic flow data comprise:
(1) flow that on highway, installation car flow detector gathers and speed of a motor vehicle traffic flow data is used;
(2) the structural type traffic flow data that video image non-structural data are converted;
(3) road (express highway roadbed, road surface, bridge and tunnel etc.) maintenance data;
(4) after traffic accidents generation, event reports and submits data;
(5) with the similar historical traffic flow data of section chain rate.
In the aforesaid traffic flow forecasting method based on degree of deep learning neural network structure, described step concrete grammar 1.: the traffic flow data of each year is divided into M group, according to respectively the traffic flow data in each sky being classified as 7 classes Monday to Sunday in each group, generate the set of the traffic flow data of 7*M, and the traffic flow data in each set is divided into idle period data and peak hours/period data on a time period.(wherein, described traffic flow data is about seasonal effect in time series data, each set in comprise 5 the above be classified as of a sort original traffic stream time series data.)
In the aforesaid traffic flow forecasting method based on degree of deep learning neural network structure, the traffic flow data in described set is about time interval continuous print data; If the discontinuous situation in time of occurrence interval in peak hours/period data, then regard as original traffic shortage of data, for shortage of data point, adopt average interpolation method polishing, if the traffic flow data of certain time point is beyond threshold value, then regard as misdata, and the carrying out of misdata is rejected, adopt average interpolation method to carry out polishing simultaneously; If the discontinuous situation in time of occurrence interval in idle period data, then regard as original traffic shortage of data, for shortage of data point, the mean value of the traffic flow data of the same time point adopting other to gather carries out polishing, if the traffic flow data of certain time point is beyond threshold value, then regard as misdata, and reject the carrying out of misdata, the mean value of the traffic flow data of the same time point simultaneously adopting other to gather carries out polishing.
In the aforesaid traffic flow forecasting method based on degree of deep learning neural network structure, described step concrete grammar 3. comprises the following steps:
(1) monolayer neuronal unit is successively built, each structure single layer network.
(2) after all layers have constructed, every layer adopts wake-sleep algorithm to carry out tuning, only adjusts one deck at every turn, successively adjust.
In the aforesaid traffic flow forecasting method based on degree of deep learning neural network structure, the concrete grammar of described tuning comprises the following steps:
A, traffic flow data is inputted the scrambler of ground floor, generate a coding, this is encoded to an expression of input, then by this coding input demoder, demoder can export a reconstruction information, again by calculating the residual error of the characteristic sum reconstruction information of the traffic flow data of input, use the weight generation of the scrambler of gradient descent method amendment ground floor;
B, the coding that exported by ground floor, as the input traffic of the second layer, adopt the method identical with steps A to revise the weight generation of the scrambler of the second layer, and repeat this step until the weight generation of scrambler of all layers is revised complete.
Aforementioned process is actual is unsupervised learning.
In the aforesaid traffic flow forecasting method based on degree of deep learning neural network structure, described step concrete grammar 4.:
The coding input support vector machine classifier (SVM) exported by last one deck, has re-used exemplar and has carried out Training, realized the fine setting of the scrambler to each layer.
In the aforesaid traffic flow forecasting method based on degree of deep learning neural network structure, the training of data carries out concurrent operation to realize by the CPU cluster in distributed system, thus can the excavation of expedited data and analysis.
Compared with prior art, the present invention carries out training study by degree of depth autocoder model to traffic flow data, thus define the neural network structure that can carry out degree of depth study to traffic flow data, it is for traditional traffic flow data forecast model, more deep analysis can be carried out to data, and effectively can extract the potential layered characteristic of data, but also eliminate the artificial huge workload extracting data characteristics, improve the efficiency of feature extraction, reduce the dimension of original input.When therefore using forecast model of the present invention to carry out short-term traffic flow forecasting, the precision of prediction can be improved.
And in method of the present invention by adopting the grouping of traffic flow data, at times, and different polishings and error correction method are carried out to the data of Different periods, not only simplify the complicacy of input traffic flow data, reduce difficulty in computation and the calculated amount of model, and also further increase the precision of short-term traffic flow forecasting to a great extent, and the stability of prediction can be improved, not easily there is apparent error.
Along with increasing of vehicle, various traffic flow data comes one after another, process a large amount of traffic flow datas and just need a kind of encoding model that can process mass data, degree of depth autocoder model has clear superiority on the one hand at this, and it can be called modeling tool indispensable in the process of the large data of following process.The present invention by degree of depth autocoder models applying above the process of traffic flow, rationally will go on a journey to Transporting Arrangement and bring great convenience, the present situation of blocking up of road can be alleviated on the one hand, the consumption of vehicle at the process China Petrochemical Industry fuel waited for can be reduced on the other hand, thus play alleviation air-polluting present situation.
Accompanying drawing explanation
Fig. 1 is basic step process flow diagram of the present invention.
Embodiment
Below in conjunction with embodiment, the present invention is further illustrated, but not as the foundation limited the present invention.
Embodiment.Based on a traffic flow forecasting method for degree of deep learning neural network structure, its basic procedure as shown in Figure 1, comprises the following steps:
1. the collection of traffic flow data;
2. the pre-service of traffic flow data;
3. degree of depth autocoder model is utilized to train traffic flow data;
4. with supervised learning algorithm, degree of depth autocoder model is finely tuned;
5. the final traffic flow of degree of depth autocoder model to short-term 4. obtained according to step is predicted.
One, the collection of traffic flow data
Gather all kinds of traffic flow data for successive depths study abundant current data and historical data are provided.Mainly comprise following aspect:
(1) flow that on highway, installation car flow detector gathers and speed of a motor vehicle traffic flow data is used;
(2) the structural type traffic flow data that video image non-structural data are converted;
(3) road maintenance data;
(4) after traffic accidents generation, event reports and submits data;
(5) with the similar historical traffic flow data of section chain rate.
Two, the pre-service of data
Consider the difference of Different periods traffic flow rule in a day in road, the time of one day is divided into peak hours/period and idle period, peak hours/period is generally 6:00 ~ 0:00, and idle period is generally 0:00 ~ 6:00, can set flexibly according to the situation in each section.Adopt diverse ways to process for peak hours/period and idle period in the embodiment of the present invention, be described in detail below.
To divide into groups to it according to certain rule for traffic flow data, form corresponding data set.In units of year, gather original traffic flow data to every day according to the collection period of setting, to obtain the original traffic stream time series data in each sky, original traffic stream time series data gathered once at interval of 5 minutes.Being divided into M=4 group i.e. 4 season according to setting rule by 1 year, according to respectively the original traffic stream time series data in each sky being classified as 7 classes Monday to Sunday in each group, generating the original traffic stream time series data set of 7*M=28.Wherein, each set in comprise 5 the above be classified as of a sort original traffic stream time series data.
Traffic flow data is gathered to obtain the original traffic stream time series data (x of the same day before current point in time according to the collection period of setting
1, x
2, x
3..., x
t, x
n), wherein x
tfor the transport data stream of t time point, x
nfor the traffic flow data of current point.Such as: collection period can be 5 minutes, and current point in time n is 10:00, then x in Traffic Flow Time Series data
n-1for the traffic flow data of 9:55, the like.If gather original traffic flow data from 0:00, then x in Traffic Flow Time Series data
1for the traffic flow data of 0:00; If gather original traffic flow data from 6:00, then x in Traffic Flow Time Series data
1for the traffic flow data of 6:00.
At peak hours/period, if the time interval discontinuous (as 8:00,8:10), be then defined as original traffic shortage of data.For shortage of data point, adopt average interpolation method polishing.Threshold method is adopted to the judgement of original traffic stream time series data mistake, exceeds the original traffic flow data of threshold value, be defined as misdata.For the rejecting of misdata, adopt average interpolation method polishing.
Section at one's leisure, because idle period vehicle is less, therefore there will be continuously the situation of long-time disappearance traffic data, in this case, average interpolation method is difficult to the data of the realistic traffic behavior of polishing, and data prediction flow process is no longer applicable.But the traffic behavior of idle period is all in freestream conditions substantially always, the therefore generation of this period historical standard data of traffic flow, employing gathers many day data and carries out pretreated method again.To in the original traffic stream time series data in each set, the data being divided into idle period calculate the mean value of the traffic flow data of same time point, and replace the traffic flow data of former same time point, obtain average traffic stream time series data.
Three, training process
Degree of depth autocoder (DAE) is that a kind of utilization supervises successively elasticity through nothing, the multilayered nonlinear network of pre-training and systemic parameter optimization extracts the layered characteristic of higher-dimension complexity input data from unlabeled data, and the degree of deep learning neural network structure that the distributed nature obtaining raw data represents.DAE is made up of scrambler, demoder and hidden layer.Scrambler is the mapping of input x to implicit expression h, is expressed as: h=f (x)=S
f(w+b
n), wherein, S
fbe non-linear activation function, be generally logical function, its expression formula is:
sigmoid(z)＝1/(1+z
-1)
Hidden layer data-mapping is returned reconstruct y by demoder function g (h), is expressed as:
y＝g(h)＝S
g(w'h+b
y)
Wherein, S
gbe the activation function of demoder, be generally linear function or sigmoid function.The process of training DAE finds parameter θ={ W, b on training sample set D
y, b
hminimum reconstructed, the expression formula of reconstructed error is:
Wherein, L is reconstruct error function, generally can with squared error function or cross entropy loss function, and the two is expressed as:
L(x,y)＝||x-y||
2
Wherein, square error is used for linear S
g, cross entropy loss function is used for sigmoid.
The structure of DAE mainly contains 2 steps, and the first step improves the structure of prototype autocoder, namely increases hidden layer and neuronic quantity, the distribution of adjustment hidden layer node and change weights share mode etc., build the basic framework of DAE.Second step is the performance index etc. needed when choosing suitable cost function and optimisation strategy, hidden layer quality factor and systemic parameter optimization according to different task, determines the training program of DAE.
Autocoder is exactly a kind of neural network reappearing input signal as far as possible.In order to realize this reproduction, autocoder just must catch the most important factor that can represent input data, just as principal component analytical method, finds the principal ingredient that can represent prime information.Namely will find the useful component of a large amount of transport data stream in the present invention, detailed process is as follows:
A, given without label data, use unsupervised learning learning characteristic;
In neural network before, the sample of input has label, i.e. (input, target), goes the parameter changing preceding layers like this according to the difference between current output and target (label), until convergence.
The automatic coding model that the present invention adopts, by input traffic flow data stream (x
1, x
2, x
3..., x
t..., x
n) an input encoder scrambler, a code (h will be obtained
1 (1), h
2 (1), h
3 (1), h
t (1)..., h
n (1)), the expression that this code namely inputs, in order to determine to export the expression that code is exactly input, add a decoder demoder, at this time decoder will export an information reconstruction, this information exported and input data input are at the beginning relatively, are almost the same in the ideal case, and we just have reason to believe that this code is the preferably annotation of input input.So every one deck is exactly the parameter by adjustment encoder scrambler and decoder demoder, makes reconstructed error minimum, first that now just obtains input input data stream illustrates, and namely encode code.Because be without label data, so the source of error is exactly obtain with former input compared with input after directly reconstruct.
B, by scrambler produce feature then train lower one deck so successively to train
Code (the h of ground floor transport data stream is obtained by previous step
1 (1), h
2 (1), h
3 (1)..., h
t (1)..., h
n (1)), now reconstructed error is minimum so this code is exactly the good representation of former input data.The principle that the second layer does not have differential utilization identical with the training patterns of ground floor, the code that ground floor exports by we is as the input input data stream of the second layer, same minimum reconstructed, will obtain the parameter of the second layer, and obtains the code (h of second layer input transport data stream
1 (2), h
2 (2), h
3 (2)..., h
t (2)..., h
n (2)), namely former input information (x
1, x
2, x
3..., x
t..., x
n) second have expressed.Other layer of profit uses the same method and trains, and training current layer, the parameter of front layer is all fixing, now no longer needs decoder.
C, have supervision fine setting
Through method above, just plurality of layers can be obtained.Every one deck all can obtain the different expression of original input.Now degree of depth autocoder is that study obtains one and well can represent input transport data stream (x
1, x
2, x
3..., x
t..., x
n) feature, this feature can represent original input signal to the full extent.Therefore in order to realize classification, needing the coding layer pushed up most at autocoder to add a support vector machine classifier (SVM) and then going training by the supervised training method (gradient descent method) of the multilayer neural network of standard.Its basic step is as follows:
1) random selecting has label data sample BP algorithm to train neural network, calculates the output of each layer;
2) reconstructed error of each layer is obtained, and according to error correction weights with biased;
3) whether meet the demands according to performance index decision errors, if fail to meet the demands, repeat step 1) and 2), until whole network exports the requirement that meets the expectation.
4) last degree of depth autocoder model is drawn.
The label data sample that has in the present invention carries out following short-term Real-Time Monitoring to aforesaid five kinds of traffic flow datas, the data of Real-Time Monitoring are compared as there being the short-term data of exemplar and the prediction of degree of depth autocoder Model forecast system, if the result difference of comparison is relatively large, now just need with there being exemplar to have the training of supervision to depth encoder, adjustment scrambler network parameter, until predict the outcome and actual observation result difference drop on reasonable interval.
And for carrying out the 5 class traffic flow datas without supervised training in the application, do not carry out the interval division such as corresponding artificial congestion in road index, therefore need to be learnt by the model of oneself for system, cluster analysis is carried out to them.
Claims (9)
1. the traffic flow forecasting method based on degree of deep learning neural network structure, it is characterized in that: by gathering all kinds of traffic flow data, and utilize degree of depth autocoder model to train all kinds of traffic flow data of collection, adjust degree of depth autocoder model in the training process, the degree of depth autocoder model after finally utilizing adjustment is predicted Short-term Traffic Flow.
2. the traffic flow forecasting method based on degree of deep learning neural network structure according to claim 1, is characterized in that, specifically comprise the following steps:
1. the collection of traffic flow data;
2. the pre-service of traffic flow data;
3. degree of depth autocoder model is utilized to train traffic flow data;
4. with supervised learning algorithm, degree of depth autocoder model is finely tuned;
5. the final traffic flow of degree of depth autocoder model to short-term 4. obtained according to step is predicted.
3. the traffic flow forecasting method based on degree of deep learning neural network structure according to claim 2, is characterized in that, described step 1. in traffic flow data comprise:
(1) flow that on highway, installation car flow detector gathers and speed of a motor vehicle traffic flow data is used;
(2) the structural type traffic flow data that video image non-structural data are converted;
(3) road maintenance data;
(4) after traffic accidents generation, event reports and submits data;
(5) with the similar historical traffic flow data of section chain rate.
4. the traffic flow forecasting method based on degree of deep learning neural network structure according to claim 2, it is characterized in that, described step concrete grammar 1.: the traffic flow data of each year is divided into M group, according to respectively the traffic flow data in each sky being classified as 7 classes Monday to Sunday in each group, generate the set of the traffic flow data of 7 × M, and the traffic flow data in each set is divided into idle period data and peak hours/period data on a time period.
5. the traffic flow forecasting method based on degree of deep learning neural network structure according to claim 4, is characterized in that: the traffic flow data in described set is about time interval continuous print data; If the discontinuous situation in time of occurrence interval in peak hours/period data, then regard as original traffic shortage of data, for shortage of data point, adopt average interpolation method polishing, if the traffic flow data of certain time point is beyond threshold value, then regard as misdata, and the carrying out of misdata is rejected, adopt average interpolation method to carry out polishing simultaneously; If the discontinuous situation in time of occurrence interval in idle period data, then regard as original traffic shortage of data, for shortage of data point, the mean value of the traffic flow data of the same time point adopting other to gather carries out polishing, if the traffic flow data of certain time point is beyond threshold value, then regard as misdata, and reject the carrying out of misdata, the mean value of the traffic flow data of the same time point simultaneously adopting other to gather carries out polishing.
6. the traffic flow forecasting method based on degree of deep learning neural network structure according to claim 2, is characterized in that, described step concrete grammar 3. comprises the following steps:
(1) monolayer neuronal unit is successively built, each structure single layer network.
(2) after all layers have constructed, every layer adopts wake-sleep algorithm to carry out tuning, only adjusts one deck at every turn, successively adjust.
7. the traffic flow forecasting method based on degree of deep learning neural network structure according to claim 6, is characterized in that, the concrete grammar of described tuning comprises the following steps:
A, traffic flow data is inputted the scrambler of ground floor, generate a coding, this is encoded to an expression of input, then by this coding input demoder, demoder can export a reconstruction information, again by calculating the residual error of the characteristic sum reconstruction information of the traffic flow data of input, use the weight generation of the scrambler of gradient descent method amendment ground floor;
B, the coding that exported by ground floor, as the input traffic of the second layer, adopt the method identical with steps A to revise the weight generation of the scrambler of the second layer, and repeat this step until the weight generation of scrambler of all layers is revised complete.
8. the traffic flow forecasting method based on degree of deep learning neural network structure according to claim 2, is characterized in that, described step concrete grammar 4.:
The coding input support vector machine classifier exported by last one deck, has re-used exemplar and has carried out Training, realized the fine setting of the scrambler to each layer.
9. the traffic flow forecasting method based on degree of deep learning neural network structure according to claim 2, is characterized in that; The training of data carries out concurrent operation to realize by the CPU cluster in distributed system.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510478215.9A CN105160866A (en) | 2015-08-07 | 2015-08-07 | Traffic flow prediction method based on deep learning nerve network structure |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510478215.9A CN105160866A (en) | 2015-08-07 | 2015-08-07 | Traffic flow prediction method based on deep learning nerve network structure |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN105160866A true CN105160866A (en) | 2015-12-16 |
Family
ID=54801705
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201510478215.9A Pending CN105160866A (en) | 2015-08-07 | 2015-08-07 | Traffic flow prediction method based on deep learning nerve network structure |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN105160866A (en) |
Cited By (53)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105654729A (en) * | 2016-03-28 | 2016-06-08 | 南京邮电大学 | Short-term traffic flow prediction method based on convolutional neural network |
| CN105922583A (en) * | 2016-05-04 | 2016-09-07 | 杭州研智科技有限公司 | Remotely controlled 3D printer and control method thereof |
| CN105930955A (en) * | 2016-04-07 | 2016-09-07 | 浙江万马新能源有限公司 | Deep learning-based charging network operation situation analysis method and apparatus |
| CN105946229A (en) * | 2016-05-04 | 2016-09-21 | 杭州研智科技有限公司 | 3d printer and control method thereof |
| CN105965886A (en) * | 2016-05-04 | 2016-09-28 | 杭州研智科技有限公司 | Error control self-learning 3D printer and control method thereof |
| CN106027438A (en) * | 2016-07-04 | 2016-10-12 | 东南大学 | Anti-aliasing modulation demodulator for analog-digital hybrid amplitude modulation broadcasting system |
| CN106003708A (en) * | 2016-05-04 | 2016-10-12 | 杭州研智科技有限公司 | 3D printing experiment platform and control method thereof |
| CN106161209A (en) * | 2016-07-21 | 2016-11-23 | 康佳集团股份有限公司 | A kind of method for filtering spam short messages based on degree of depth self study and system |
| CN106205126A (en) * | 2016-08-12 | 2016-12-07 | 北京航空航天大学 | Large-scale Traffic Network based on convolutional neural networks is blocked up Forecasting Methodology and device |
| CN106295906A (en) * | 2016-08-22 | 2017-01-04 | 万马联合新能源投资有限公司 | A kind of urban public bus lines querying method based on degree of depth study |
| CN106326634A (en) * | 2016-08-09 | 2017-01-11 | 浙江工业大学 | Hospital outpatient quantity predicting method based on deep belief network |
| CN106529574A (en) * | 2016-10-17 | 2017-03-22 | 北京工业大学 | Image classification method based on sparse automatic encoder and support vector machine |
| CN106548645A (en) * | 2016-11-03 | 2017-03-29 | 济南博图信息技术有限公司 | Vehicle route optimization method and system based on deep learning |
| CN106650913A (en) * | 2016-12-31 | 2017-05-10 | 中国科学技术大学 | Deep convolution neural network-based traffic flow density estimation method |
| CN106781465A (en) * | 2016-12-06 | 2017-05-31 | 广州市科恩电脑有限公司 | A kind of road traffic Forecasting Methodology |
| CN106846816A (en) * | 2017-04-12 | 2017-06-13 | 山东理工大学 | A kind of discretization traffic state judging method based on deep learning |
| CN106951872A (en) * | 2017-03-24 | 2017-07-14 | 江苏大学 | A kind of recognition methods again of the pedestrian based on unsupervised depth model and hierarchy attributes |
| CN106971548A (en) * | 2017-05-18 | 2017-07-21 | 福州大学 | The Short-time Traffic Flow Forecasting Methods of optimizable multi-core adaptive SVMs |
| WO2017124336A1 (en) * | 2016-01-20 | 2017-07-27 | Sensetime Group Limited | Method and system for adapting deep model for object representation from source domain to target domain |
| CN107093216A (en) * | 2017-04-14 | 2017-08-25 | 广州地理研究所 | Urban traffic blocking charging method and device based on vehicle electron identifying |
| CN107103758A (en) * | 2017-06-08 | 2017-08-29 | 厦门大学 | A kind of city area-traffic method for predicting based on deep learning |
| CN107154150A (en) * | 2017-07-25 | 2017-09-12 | 北京航空航天大学 | A kind of traffic flow forecasting method clustered based on road with double-layer double-direction LSTM |
| CN107170235A (en) * | 2017-06-13 | 2017-09-15 | 广东石油化工学院 | The traffic flow forecasting timing method of network is extracted based on depth characteristic |
| CN107169598A (en) * | 2017-05-05 | 2017-09-15 | 上海斐讯数据通信技术有限公司 | A kind of weather prediction method and system based on deep learning |
| CN107451599A (en) * | 2017-06-28 | 2017-12-08 | 青岛科技大学 | A kind of traffic behavior Forecasting Methodology of the collective model based on machine learning |
| CN107463953A (en) * | 2017-07-21 | 2017-12-12 | 上海交通大学 | Image classification method and system based on quality insertion in the case of label is noisy |
| CN107480807A (en) * | 2017-07-06 | 2017-12-15 | 中山大学 | Shared bicycle destination Forecasting Methodology and device based on space-time layered perception neural networks |
| CN107945534A (en) * | 2017-12-13 | 2018-04-20 | 浙江大学城市学院 | A kind of special bus method for predicting based on GMDH neutral nets |
| CN108198415A (en) * | 2017-12-28 | 2018-06-22 | 同济大学 | A kind of city expressway accident forecast method based on deep learning |
| CN108256052A (en) * | 2018-01-15 | 2018-07-06 | 成都初联创智软件有限公司 | Automobile industry potential customers' recognition methods based on tri-training |
| CN108629503A (en) * | 2018-04-28 | 2018-10-09 | 南通大学 | A kind of taxi based on deep learning is got on the bus the prediction technique of demand |
| CN108629976A (en) * | 2018-05-17 | 2018-10-09 | 同济大学 | Urban traffic blocking predetermined depth learning method based on GPS |
| CN108629618A (en) * | 2017-03-22 | 2018-10-09 | 董泽平 | Product sales prediction method and system without model speculation foundation |
| CN109146194A (en) * | 2018-09-05 | 2019-01-04 | 重庆工商大学 | A kind of Runoff forestry method encoding interconnection vector machine certainly based on change mode convolution |
| CN109166309A (en) * | 2018-08-06 | 2019-01-08 | 重庆邮电大学 | A kind of missing data on flows restoration methods towards complicated urban traffic network |
| CN109272157A (en) * | 2018-09-12 | 2019-01-25 | 中交信息技术国家工程实验室有限公司 | A kind of freeway traffic flow parameter prediction method and system based on gate neural network |
| CN109461311A (en) * | 2018-12-19 | 2019-03-12 | 哈尔滨工业大学（深圳） | A kind of road network traffic flow spatio-temporal prediction method towards intelligent transportation and intelligent driving |
| CN109658695A (en) * | 2019-01-02 | 2019-04-19 | 华南理工大学 | A kind of multifactor Short-time Traffic Flow Forecasting Methods |
| CN109887290A (en) * | 2019-03-30 | 2019-06-14 | 西安电子科技大学 | Traffic flow forecasting method based on equilibrium index exponential smoothing and stack self-encoding encoder |
| TWI676397B (en) * | 2018-10-18 | 2019-11-01 | 中華電信股份有限公司 | Artificial intelligence traffic estimation system using mobile network signaling data and method thereof |
| CN110443503A (en) * | 2019-08-07 | 2019-11-12 | 成都九鼎瑞信科技股份有限公司 | The training method and related system of water utilities system industrial gross output value analysis model |
| CN110459050A (en) * | 2019-05-27 | 2019-11-15 | 华南理工大学 | A kind of short-term bus passenger flow prediction technique based on hybrid decision tree |
| CN110517488A (en) * | 2019-08-19 | 2019-11-29 | 南京理工大学 | The Short-time Traffic Flow Forecasting Methods with Recognition with Recurrent Neural Network are decomposed based on timing |
| CN110619427A (en) * | 2019-08-26 | 2019-12-27 | 北京建筑大学 | Traffic index prediction method and device based on sequence-to-sequence learning model |
| CN110633707A (en) * | 2018-06-21 | 2019-12-31 | 北京京东尚科信息技术有限公司 | Method and device for predicting speed |
| CN111582251A (en) * | 2020-06-15 | 2020-08-25 | 江苏航天大为科技股份有限公司 | Method for detecting passenger crowding degree of urban rail transit based on convolutional neural network |
| CN111598222A (en) * | 2020-05-14 | 2020-08-28 | 哈尔滨工业大学（威海） | Re-optimization depth automatic encoder and engine automatic detection system |
| CN111724596A (en) * | 2020-06-23 | 2020-09-29 | 上海电科智能系统股份有限公司 | Method for intelligently, accurately and automatically identifying and prejudging bottleneck region of expressway |
| CN111862585A (en) * | 2019-07-23 | 2020-10-30 | 北京嘀嘀无限科技发展有限公司 | System and method for traffic prediction |
| WO2020220439A1 (en) * | 2019-04-29 | 2020-11-05 | 东北大学 | Highway traffic flow state recognition method based on deep neural network |
| CN113450566A (en) * | 2021-06-22 | 2021-09-28 | 中科曙光(南京)计算技术有限公司 | Urban traffic flow prediction method |
| JP2021170410A (en) * | 2020-12-16 | 2021-10-28 | アポロ インテリジェント コネクティビティ (ベイジン) テクノロジー カンパニー リミテッドApollo Intelligent Connectivity (Beijing) Technology Co., Ltd. | Method of analyzing traffic data, apparatus therefor, device therefor, vehicle, recording medium, and program |
| CN116363878A (en) * | 2023-05-26 | 2023-06-30 | 云南大学 | Traffic flow prediction system and method based on continuous dynamic ordinary differential equation |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2006018435A (en) * | 2004-06-30 | 2006-01-19 | Matsushita Electric Ind Co Ltd | Traffic flow data prediction device and traffic flow data prediction method |
| CN101188002A (en) * | 2007-12-24 | 2008-05-28 | 北京大学 | A city traffic dynamic prediction system and method with real time and continuous feature |
| CN102033892A (en) * | 2009-09-30 | 2011-04-27 | 北京四通智能交通系统集成有限公司 | Method and system for generating historical standard data of traffic flow |
| CN102521671A (en) * | 2011-11-29 | 2012-06-27 | 华北电力大学 | Ultrashort-term wind power prediction method |
| CN103489039A (en) * | 2013-09-12 | 2014-01-01 | 重庆大学 | Expressway traffic flow fusing and forecasting method with online self-tuning and optimizing function |
| CN104134351A (en) * | 2014-08-14 | 2014-11-05 | 中国科学院自动化研究所 | Short-term traffic flow predicting method |
-
2015
- 2015-08-07 CN CN201510478215.9A patent/CN105160866A/en active Pending
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2006018435A (en) * | 2004-06-30 | 2006-01-19 | Matsushita Electric Ind Co Ltd | Traffic flow data prediction device and traffic flow data prediction method |
| CN101188002A (en) * | 2007-12-24 | 2008-05-28 | 北京大学 | A city traffic dynamic prediction system and method with real time and continuous feature |
| CN102033892A (en) * | 2009-09-30 | 2011-04-27 | 北京四通智能交通系统集成有限公司 | Method and system for generating historical standard data of traffic flow |
| CN102521671A (en) * | 2011-11-29 | 2012-06-27 | 华北电力大学 | Ultrashort-term wind power prediction method |
| CN103489039A (en) * | 2013-09-12 | 2014-01-01 | 重庆大学 | Expressway traffic flow fusing and forecasting method with online self-tuning and optimizing function |
| CN104134351A (en) * | 2014-08-14 | 2014-11-05 | 中国科学院自动化研究所 | Short-term traffic flow predicting method |
Non-Patent Citations (1)
| Title |
|---|
| 曲建岭； 杜辰飞；邸亚洲；高峰；郭超然: "深度自动编码器的研究与展望", 《计算机与现代化》 * |
Cited By (79)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108604304A (en) * | 2016-01-20 | 2018-09-28 | 商汤集团有限公司 | For adapting the depth model indicated for object from source domain to the method and system of aiming field |
| WO2017124336A1 (en) * | 2016-01-20 | 2017-07-27 | Sensetime Group Limited | Method and system for adapting deep model for object representation from source domain to target domain |
| CN105654729A (en) * | 2016-03-28 | 2016-06-08 | 南京邮电大学 | Short-term traffic flow prediction method based on convolutional neural network |
| CN105654729B (en) * | 2016-03-28 | 2018-01-02 | 南京邮电大学 | A kind of short-term traffic flow forecast method based on convolutional neural networks |
| CN105930955A (en) * | 2016-04-07 | 2016-09-07 | 浙江万马新能源有限公司 | Deep learning-based charging network operation situation analysis method and apparatus |
| CN105946229A (en) * | 2016-05-04 | 2016-09-21 | 杭州研智科技有限公司 | 3d printer and control method thereof |
| CN106003708A (en) * | 2016-05-04 | 2016-10-12 | 杭州研智科技有限公司 | 3D printing experiment platform and control method thereof |
| CN105965886A (en) * | 2016-05-04 | 2016-09-28 | 杭州研智科技有限公司 | Error control self-learning 3D printer and control method thereof |
| CN105922583A (en) * | 2016-05-04 | 2016-09-07 | 杭州研智科技有限公司 | Remotely controlled 3D printer and control method thereof |
| CN106027438A (en) * | 2016-07-04 | 2016-10-12 | 东南大学 | Anti-aliasing modulation demodulator for analog-digital hybrid amplitude modulation broadcasting system |
| CN106027438B (en) * | 2016-07-04 | 2020-04-07 | 东南大学 | Anti-aliasing modem for analog-digital hybrid amplitude modulation broadcasting system |
| CN106161209A (en) * | 2016-07-21 | 2016-11-23 | 康佳集团股份有限公司 | A kind of method for filtering spam short messages based on degree of depth self study and system |
| CN106161209B (en) * | 2016-07-21 | 2019-09-20 | 康佳集团股份有限公司 | A kind of method for filtering spam short messages and system based on depth self study |
| CN106326634A (en) * | 2016-08-09 | 2017-01-11 | 浙江工业大学 | Hospital outpatient quantity predicting method based on deep belief network |
| CN106205126A (en) * | 2016-08-12 | 2016-12-07 | 北京航空航天大学 | Large-scale Traffic Network based on convolutional neural networks is blocked up Forecasting Methodology and device |
| CN106295906A (en) * | 2016-08-22 | 2017-01-04 | 万马联合新能源投资有限公司 | A kind of urban public bus lines querying method based on degree of depth study |
| CN106529574A (en) * | 2016-10-17 | 2017-03-22 | 北京工业大学 | Image classification method based on sparse automatic encoder and support vector machine |
| CN106529574B (en) * | 2016-10-17 | 2019-09-27 | 北京工业大学 | Image classification method based on sparse autocoder and support vector machines |
| CN106548645B (en) * | 2016-11-03 | 2019-07-12 | 济南博图信息技术有限公司 | Vehicle route optimization method and system based on deep learning |
| CN106548645A (en) * | 2016-11-03 | 2017-03-29 | 济南博图信息技术有限公司 | Vehicle route optimization method and system based on deep learning |
| CN106781465A (en) * | 2016-12-06 | 2017-05-31 | 广州市科恩电脑有限公司 | A kind of road traffic Forecasting Methodology |
| CN106650913A (en) * | 2016-12-31 | 2017-05-10 | 中国科学技术大学 | Deep convolution neural network-based traffic flow density estimation method |
| CN106650913B (en) * | 2016-12-31 | 2018-08-03 | 中国科学技术大学 | A kind of vehicle density method of estimation based on depth convolutional neural networks |
| CN108629618A (en) * | 2017-03-22 | 2018-10-09 | 董泽平 | Product sales prediction method and system without model speculation foundation |
| CN106951872B (en) * | 2017-03-24 | 2020-11-06 | 江苏大学 | Pedestrian re-identification method based on unsupervised depth model and hierarchical attributes |
| CN106951872A (en) * | 2017-03-24 | 2017-07-14 | 江苏大学 | A kind of recognition methods again of the pedestrian based on unsupervised depth model and hierarchy attributes |
| CN106846816A (en) * | 2017-04-12 | 2017-06-13 | 山东理工大学 | A kind of discretization traffic state judging method based on deep learning |
| CN106846816B (en) * | 2017-04-12 | 2019-09-17 | 山东理工大学 | A kind of discretization traffic state judging method based on deep learning |
| CN107093216A (en) * | 2017-04-14 | 2017-08-25 | 广州地理研究所 | Urban traffic blocking charging method and device based on vehicle electron identifying |
| CN107169598A (en) * | 2017-05-05 | 2017-09-15 | 上海斐讯数据通信技术有限公司 | A kind of weather prediction method and system based on deep learning |
| CN106971548A (en) * | 2017-05-18 | 2017-07-21 | 福州大学 | The Short-time Traffic Flow Forecasting Methods of optimizable multi-core adaptive SVMs |
| CN106971548B (en) * | 2017-05-18 | 2019-06-07 | 福州大学 | The Short-time Traffic Flow Forecasting Methods of optimizable multi-core adaptive support vector machines |
| CN107103758B (en) * | 2017-06-08 | 2019-06-21 | 厦门大学 | A kind of city area-traffic method for predicting based on deep learning |
| CN107103758A (en) * | 2017-06-08 | 2017-08-29 | 厦门大学 | A kind of city area-traffic method for predicting based on deep learning |
| CN107170235B (en) * | 2017-06-13 | 2020-03-03 | 广东石油化工学院 | Traffic flow prediction time sequence method based on depth feature extraction network |
| CN107170235A (en) * | 2017-06-13 | 2017-09-15 | 广东石油化工学院 | The traffic flow forecasting timing method of network is extracted based on depth characteristic |
| CN107451599A (en) * | 2017-06-28 | 2017-12-08 | 青岛科技大学 | A kind of traffic behavior Forecasting Methodology of the collective model based on machine learning |
| CN107480807A (en) * | 2017-07-06 | 2017-12-15 | 中山大学 | Shared bicycle destination Forecasting Methodology and device based on space-time layered perception neural networks |
| CN107463953B (en) * | 2017-07-21 | 2019-11-19 | 上海媒智科技有限公司 | Image classification method and system based on quality insertion in the noisy situation of label |
| CN107463953A (en) * | 2017-07-21 | 2017-12-12 | 上海交通大学 | Image classification method and system based on quality insertion in the case of label is noisy |
| CN107154150B (en) * | 2017-07-25 | 2019-07-02 | 北京航空航天大学 | A kind of traffic flow forecasting method based on road cluster and double-layer double-direction LSTM |
| CN107154150A (en) * | 2017-07-25 | 2017-09-12 | 北京航空航天大学 | A kind of traffic flow forecasting method clustered based on road with double-layer double-direction LSTM |
| CN107945534A (en) * | 2017-12-13 | 2018-04-20 | 浙江大学城市学院 | A kind of special bus method for predicting based on GMDH neutral nets |
| CN108198415A (en) * | 2017-12-28 | 2018-06-22 | 同济大学 | A kind of city expressway accident forecast method based on deep learning |
| CN108256052A (en) * | 2018-01-15 | 2018-07-06 | 成都初联创智软件有限公司 | Automobile industry potential customers' recognition methods based on tri-training |
| CN108629503A (en) * | 2018-04-28 | 2018-10-09 | 南通大学 | A kind of taxi based on deep learning is got on the bus the prediction technique of demand |
| CN108629503B (en) * | 2018-04-28 | 2020-11-03 | 南通大学 | Prediction method for taxi getting-on demand based on deep learning |
| CN108629976A (en) * | 2018-05-17 | 2018-10-09 | 同济大学 | Urban traffic blocking predetermined depth learning method based on GPS |
| CN110633707A (en) * | 2018-06-21 | 2019-12-31 | 北京京东尚科信息技术有限公司 | Method and device for predicting speed |
| CN109166309B (en) * | 2018-08-06 | 2021-03-19 | 重庆邮电大学 | Missing traffic data recovery method for complex urban traffic network |
| CN109166309A (en) * | 2018-08-06 | 2019-01-08 | 重庆邮电大学 | A kind of missing data on flows restoration methods towards complicated urban traffic network |
| CN109146194A (en) * | 2018-09-05 | 2019-01-04 | 重庆工商大学 | A kind of Runoff forestry method encoding interconnection vector machine certainly based on change mode convolution |
| CN109272157A (en) * | 2018-09-12 | 2019-01-25 | 中交信息技术国家工程实验室有限公司 | A kind of freeway traffic flow parameter prediction method and system based on gate neural network |
| TWI676397B (en) * | 2018-10-18 | 2019-11-01 | 中華電信股份有限公司 | Artificial intelligence traffic estimation system using mobile network signaling data and method thereof |
| CN111161529A (en) * | 2018-10-18 | 2020-05-15 | 中华电信股份有限公司 | Artificial intelligent traffic flow estimation system and method using mobile network signaling data |
| CN111161529B (en) * | 2018-10-18 | 2021-06-08 | 中华电信股份有限公司 | Artificial intelligent traffic flow estimation system and method using mobile network signaling data |
| CN109461311A (en) * | 2018-12-19 | 2019-03-12 | 哈尔滨工业大学（深圳） | A kind of road network traffic flow spatio-temporal prediction method towards intelligent transportation and intelligent driving |
| CN109658695B (en) * | 2019-01-02 | 2020-09-22 | 华南理工大学 | Multi-factor short-term traffic flow prediction method |
| CN109658695A (en) * | 2019-01-02 | 2019-04-19 | 华南理工大学 | A kind of multifactor Short-time Traffic Flow Forecasting Methods |
| CN109887290A (en) * | 2019-03-30 | 2019-06-14 | 西安电子科技大学 | Traffic flow forecasting method based on equilibrium index exponential smoothing and stack self-encoding encoder |
| CN109887290B (en) * | 2019-03-30 | 2021-03-23 | 西安电子科技大学 | Traffic flow prediction method based on balance index smoothing method and stack type self-encoder |
| WO2020220439A1 (en) * | 2019-04-29 | 2020-11-05 | 东北大学 | Highway traffic flow state recognition method based on deep neural network |
| CN110459050A (en) * | 2019-05-27 | 2019-11-15 | 华南理工大学 | A kind of short-term bus passenger flow prediction technique based on hybrid decision tree |
| CN110459050B (en) * | 2019-05-27 | 2021-07-20 | 华南理工大学 | Short-term bus passenger flow prediction method based on mixed decision tree |
| CN111862585A (en) * | 2019-07-23 | 2020-10-30 | 北京嘀嘀无限科技发展有限公司 | System and method for traffic prediction |
| CN111862585B (en) * | 2019-07-23 | 2021-11-02 | 北京嘀嘀无限科技发展有限公司 | System and method for traffic prediction |
| CN110443503A (en) * | 2019-08-07 | 2019-11-12 | 成都九鼎瑞信科技股份有限公司 | The training method and related system of water utilities system industrial gross output value analysis model |
| CN110517488A (en) * | 2019-08-19 | 2019-11-29 | 南京理工大学 | The Short-time Traffic Flow Forecasting Methods with Recognition with Recurrent Neural Network are decomposed based on timing |
| CN110619427A (en) * | 2019-08-26 | 2019-12-27 | 北京建筑大学 | Traffic index prediction method and device based on sequence-to-sequence learning model |
| CN111598222B (en) * | 2020-05-14 | 2023-04-28 | 哈尔滨工业大学（威海） | Re-optimized depth automatic encoder and engine automatic detection system |
| CN111598222A (en) * | 2020-05-14 | 2020-08-28 | 哈尔滨工业大学（威海） | Re-optimization depth automatic encoder and engine automatic detection system |
| CN111582251A (en) * | 2020-06-15 | 2020-08-25 | 江苏航天大为科技股份有限公司 | Method for detecting passenger crowding degree of urban rail transit based on convolutional neural network |
| CN111724596B (en) * | 2020-06-23 | 2022-11-11 | 上海电科智能系统股份有限公司 | Method for intelligently, accurately and automatically identifying and prejudging bottleneck region of expressway |
| CN111724596A (en) * | 2020-06-23 | 2020-09-29 | 上海电科智能系统股份有限公司 | Method for intelligently, accurately and automatically identifying and prejudging bottleneck region of expressway |
| JP2021170410A (en) * | 2020-12-16 | 2021-10-28 | アポロ インテリジェント コネクティビティ (ベイジン) テクノロジー カンパニー リミテッドApollo Intelligent Connectivity (Beijing) Technology Co., Ltd. | Method of analyzing traffic data, apparatus therefor, device therefor, vehicle, recording medium, and program |
| JP7227320B2 (en) | 2020-12-16 | 2023-02-21 | 阿波▲羅▼智▲聯▼（北京）科技有限公司 | TRAFFIC DATA ANALYSIS METHOD, APPARATUS, DEVICE, VEHICLE, STORAGE MEDIUM AND PROGRAM |
| CN113450566A (en) * | 2021-06-22 | 2021-09-28 | 中科曙光(南京)计算技术有限公司 | Urban traffic flow prediction method |
| CN116363878A (en) * | 2023-05-26 | 2023-06-30 | 云南大学 | Traffic flow prediction system and method based on continuous dynamic ordinary differential equation |
| CN116363878B (en) * | 2023-05-26 | 2023-08-11 | 云南大学 | Traffic flow prediction system and method based on continuous dynamic ordinary differential equation |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN105160866A (en) | Traffic flow prediction method based on deep learning nerve network structure | |
| Essien et al. | Improving urban traffic speed prediction using data source fusion and deep learning | |
| CN109272157A (en) | A kind of freeway traffic flow parameter prediction method and system based on gate neural network | |
| CN104900063B (en) | A kind of short distance running time Forecasting Methodology | |
| CN102346964A (en) | Real-time jam prediction and intelligent management system for road traffic network area | |
| CN113487066A (en) | Long-time-sequence freight volume prediction method based on multi-attribute enhanced graph convolution-Informer model | |
| CN104992244A (en) | Airport freight traffic prediction analysis method based on SARIMA and RBF neural network integration combination model | |
| CN104318757B (en) | Bus bus or train route section Forecasting Methodology working time on a kind of public transportation lane | |
| CN104798043A (en) | Data processing method and computer system | |
| CN105096614A (en) | Newly established crossing traffic flow prediction method based on generating type deep belief network | |
| Xiao-chen | Application of neural network in the cost estimation of highway engineering | |
| CN107544251A (en) | A kind of minimum based on Robust distributed model always drags the Single Machine Scheduling method of phase | |
| CN108665093A (en) | Traffic accidents Severity forecasting method based on deep learning | |
| CN115440032A (en) | Long-term and short-term public traffic flow prediction method | |
| CN111815046A (en) | Traffic flow prediction method based on deep learning | |
| CN108053052A (en) | A kind of oil truck oil and gas leakage speed intelligent monitor system | |
| CN114565187A (en) | Traffic network data prediction method based on graph space-time self-coding network | |
| CN109102016A (en) | A kind of test method for bridge technology situation | |
| CN103456163A (en) | City expressway interchange traffic capacity and running status discrimination method and system | |
| CN105160434A (en) | Wind power ramp event prediction method by adopting SVM to select forecasting model | |
| Zhang et al. | Overview of traffic flow hybrid ANN forecasting algorithm study | |
| CN110490365A (en) | A method of based on the pre- survey grid of multisource data fusion about vehicle order volume | |
| CN117078191A (en) | Data-driven multi-system rail transit emergency collaborative decision-making method and device | |
| CN107423811A (en) | The streamflow change attribution recognition methods combined based on BP artificial neural networks and Scene Simulation | |
| CN115482666B (en) | Multi-graph convolution neural network traffic prediction method based on data fusion |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20151216 |
|
| RJ01 | Rejection of invention patent application after publication |