CN106846806A - Urban highway traffic method for detecting abnormality based on Isolation Forest - Google Patents
Urban highway traffic method for detecting abnormality based on Isolation Forest Download PDFInfo
- Publication number
- CN106846806A CN106846806A CN201710131835.4A CN201710131835A CN106846806A CN 106846806 A CN106846806 A CN 106846806A CN 201710131835 A CN201710131835 A CN 201710131835A CN 106846806 A CN106846806 A CN 106846806A
- Authority
- CN
- China
- Prior art keywords
- road
- speed
- record
- data
- isolation forest
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Classifications
-
- G—PHYSICS
- G08—SIGNALLING
- G08G—TRAFFIC CONTROL SYSTEMS
- G08G1/00—Traffic control systems for road vehicles
- G08G1/01—Detecting movement of traffic to be counted or controlled
- G08G1/0104—Measuring and analyzing of parameters relative to traffic conditions
- G08G1/0108—Measuring and analyzing of parameters relative to traffic conditions based on the source of data
- G08G1/0112—Measuring and analyzing of parameters relative to traffic conditions based on the source of data from the vehicle, e.g. floating car data [FCD]
-
- G—PHYSICS
- G08—SIGNALLING
- G08G—TRAFFIC CONTROL SYSTEMS
- G08G1/00—Traffic control systems for road vehicles
- G08G1/01—Detecting movement of traffic to be counted or controlled
- G08G1/0104—Measuring and analyzing of parameters relative to traffic conditions
- G08G1/0125—Traffic data processing
-
- G—PHYSICS
- G08—SIGNALLING
- G08G—TRAFFIC CONTROL SYSTEMS
- G08G1/00—Traffic control systems for road vehicles
- G08G1/01—Detecting movement of traffic to be counted or controlled
- G08G1/0104—Measuring and analyzing of parameters relative to traffic conditions
- G08G1/0137—Measuring and analyzing of parameters relative to traffic conditions for specific applications
Landscapes
- Chemical & Material Sciences (AREA)
- Analytical Chemistry (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Traffic Control Systems (AREA)
Abstract
The present invention discloses a kind of urban highway traffic method for detecting abnormality based on Isolation Forest, with road as detection object, average running speed according to road in different periods divides different classes of data set, one Isolation Forest is trained based on each data set, by detecting whether road speeds are abnormal to judge road to the distance of root node in Isolation Forest.It is small with amount of calculation using technical scheme, it is not necessary to the characteristics of all calculating every road its distribution.
Description
Technical field
The invention belongs to intelligent transportation field, more particularly to a kind of urban highway traffic based on Isolation Forest
Method for detecting abnormality.
Background technology
Traffic abnormity refers to that the actual motion feature of road deviate from its desired operation characteristic, and this is in urban road network
In widely exist.When it there is major traffic accidents, rally, construction and traffic control, road network traffic flow can become
Obtain abnormal, influence travel time, efficiency and pollute, therefore for vehicle supervision department, detect city road network
In exception it is particularly significant.But the complexity based on city road network, at present most anomaly algorithm research is also based only on
Highway and through street and launch, for urban highway traffic abnormality detection still in weak tendency.
The abnormal detection of urban highway traffic is carried out based on the traffic data for collecting, be substantially a pattern-recognition or
Classification problem.According to the difference of the data source for being used, existing accident detection method can be divided into two classes：Based on fixation
The detection method of monitor and the detection method based on Floating Car.
The accident detection algorithm based on stationary monitoring device being widely used has：Jia Lifuni based on pattern-recognition
Sub- algorithm, the standard deviation based on statistical analysis etc..California algorithm is by comparing occupying between adjacent inspection stations
Rate data, differentiate to traffic abnormity that may be present.Whether standard deviation is big by judging the rate of change of traffic parameter
The differentiation to traffic abnormity is realized in specified threshold value.Stationary monitoring device is laid on road mostly, data precision compared with
It is good, but the transport information of specified point can only be detected, it is difficult to detect the transport information in section.
Floating car technology is the relatively new type traffic information collection technology of development.Floating Car is the car for being equipped with GPS device
, with application convenience, economy, wide coverage the features such as, and collect be based on section interval data (such as section
Speed), more can really reflect the operation conditions of road.
Detection method based on stationary monitoring device data is the change of direct detection road relevant parameter, and Floating Car is returned
Data be vehicle running state parameter, therefore the detection method based on stationary monitoring device cannot be applied to floating car data.
Urban transportation Outlier Detection Algorithm based on floating car data has two classes：Map partitioning is region by one class algorithm,
With the change of vehicular movement between vector representation region and region, then detect abnormal with methods such as KNN, what these algorithms were obtained
It is the exception between region and region；Another kind of algorithm calculates the exception of detection road, such as normal deviate detection algorithm, the calculation
Method assumes traffic value Normal Distribution of the road within certain time period, if the magnitude of traffic flow of certain time period has exceeded finger
Fixed threshold value, then it is assumed that this road is abnormal within this time period, and the method needs to calculate all roads different
Distribution Value in time period, blocking the way way amount is computationally intensive when more.
Isolation Forest methods initially put forward to be for detecting network intrusions data etc..Its basic thought is
Think that abnormal data is isolated and less in overall data, then when being divided to data, abnormal data is relative
Only needing to little number of times in normal data can just mark off.Isolation Forest methods are relatively low to memory requirements
In the case of can keep linear time complexity because Isolation Forest methods train exception by sampled data set
Tree, therefore also have performance well on the larger data set of data volume and High Dimensional Data Set.
The content of the invention
The present invention provides a kind of urban highway traffic method for detecting abnormality based on Isolation Forest.It is with road
Detection object, the average running speed according to road in different periods divides different classes of data set, based on the training of each data
Practice Isolation Forest, by detecting that road speeds are sentenced in Isolation Forest to the distance of root node
Whether disconnected road is abnormal.Method amount of calculation proposed by the present invention is small, it is not necessary to all calculate every road its distribution.
To realize the above method, the present invention is adopted the following technical scheme that：
A kind of urban highway traffic method for detecting abnormality based on Isolation Forest is comprised the following steps：
Step 1, with road as detection object, the average running speed according to road in different periods divides different classes of number
According to collection；
Step 2, based on each data set train an Isolation Forest, by detecting that road speeds exist
Judge whether road is abnormal in Isolation Forest to the distance of root node.
Preferably, step 1 is specially：
If city road network is represented with a digraph G：
G=(V, E)
V is the point set of digraph, and each point represents the intersection in road network, two tuple tables being made up of longitude and latitude
Show；E is the side collection of digraph, and each edge represents certain road in road network；
Floating Car returns to vehicle-state by the GPS device for carrying, and the form of returned data is as follows：
Record=[car ID, time, latitude, lontitude, speed, angel]
Wherein car ID represent the license plate number of vehicle, and time represents time during return recording, latitude,
The longitude and latitude of vehicle position when lontitude represents return recording, speed represent car speed, and angel represents car
Traveling angle；
First, map match carried out to the data that Floating Car is returned, calculate with Floating Car returned data (latitude,
Lontitude) the point on immediate road, even if the correct position that floating car data is matched with section；
Secondly, the road data after floating car data is matched is classified, and classification has two dimensions of time and speed
Standard：1) to daily 06：00 to 21：00 time period was divided with prefixed time interval；2) by road speed in each time period
Degree is divided into four speed intervals according to urban highway traffic assessment indicator system,
Then it is interval according to time period signature velocity to all roads in road network, i.e.,<e,t,s>Represent in time period t, road
Road e belongs to speed interval s, s ∈ { S1,S2,S3,S4, e ∈ E, wherein s are the averages of the time period t road e speed in n days；
According to the speed interval of pavement marker, all of road data is divided into four data set Di, i ∈ { 1,2,3,4 },
If<e,t,s>Component s ∈ Si, then the record of time period ts of the road e in n days<e,tj,speed>∈Di, j=1,
2,……n}。
Preferably, step 2 specifically includes following steps being：
Step 2.1, based on data set DiBuild Isolation Forest Fi={ Tk| k=1,2 ..., m } for detecting
Urban highway traffic exception, FiIt is comprising the m forest of abnormal tree；
Step 2.2, one road record of detection<e,tj,speed>, i.e. road e is average in the time period t of jth day
Whether speed speed is abnormal, first finds the speed interval S belonging to iti, that is, have found corresponding Isolation Forest Fi,
Then road record is calculated in FiIn mean depth；
Step 2.3, basis are calculated road record in Isolation Forest FiIn mean depth, i.e. road
The desired length of recording distance root node is length, and exceptional value is calculated according to formula (1)
C (ψ) is calculated by formula (2).
H (ψ -1)=ln (ψ -1)+0.57721 (Euler's constant)
Wherein, ψ is 256,
If the return value s closely 1 of record, show that record close to root node, just can isolate by little division
Out, then this record may be considered it is abnormal；
If return value is far smaller than 0.5, illustrating recording distance root node farther out could remember this, it is necessary to repeatedly divide
Record is isolated, then this record may be considered normal.
Brief description of the drawings
Fig. 1 is the node structure of Isolation Tree；
Fig. 2 exceptions section number ratio；
The abnormal section that Fig. 3 this method is detected；
Fig. 4 Isolation Forest train consumed time diagram with normal deviate detection algorithm.
The flow chart of Fig. 5 method for detecting abnormality of the present invention.
Specific embodiment
As shown in figure 5, the present invention provides a kind of urban highway traffic abnormality detection based on Isolation Forest calculating
Method.With road as detection object, the average running speed according to road in different periods divides different classes of data set, based on every
An Isolation Forest is practiced in individual data training, by detecting that road speeds arrive root node in Isolation Forest
Distance it is whether abnormal to judge road.Method amount of calculation proposed by the present invention is small, it is not necessary to all calculate its point to every road
Cloth.
City road network is represented with a digraph G：
G=(V, E)
V is the point set of digraph, and each point represents the intersection in road network, two tuples being typically made up of longitude and latitude
Represent.E is the side collection of digraph, and each edge represents certain road in road network.
Floating Car returns to vehicle-state by the GPS device for carrying, and the form of returned data is as follows：
Record=[car ID, time, latitude, lontitude, speed, angel]
Wherein car ID represent the license plate number of vehicle, and time represents time during return recording, latitude,
The longitude and latitude of vehicle position when lontitude represents return recording, speed represent car speed, and angel represents car
Traveling angle.In general, daily 21：00 to next day 6：Floating Car negligible amounts between 00, the data volume of return is inadequate, because
This abnormality detection only considers the same day 06：00 to 21：Road abnormal state detection on 00 time period.
Map match is carried out firstly the need of the data returned to Floating Car, because the latitude and longitude information that Floating Car is returned is present
Deviation, it is necessary to calculate with the point on Floating Car returned data (latitude, lontitude) immediate road, even if float
The correct position that car data is matched with section, then the speed according to Floating Car calculate the average speed of road.
Secondly it is abnormal, it is necessary to the road data after floating car data is matched is divided in order to more accurately detect
Class, classification has two standards of dimension of time and speed：1) to daily 06：00 to 21：00 time period with 20 minutes for be spaced into
Row is divided, i.e., include 45 time periods daily；2) road speeds in each time period are referred to according to urban highway traffic evaluation
Mark system is divided into four speed intervals, for specific speed value speed
if speed>60km/h speed∈S1
else if speed≥40km/h speed∈S2
else if speed≥20km/h speed∈S3
else speed∈S4
Then it is interval according to time period signature velocity to all roads in road network, i.e.,<e,t,s>Represent in time period t, road
Road e belongs to speed interval s, s ∈ { S1,S2,S3,S4, e ∈ E, wherein s are the averages of the time period t road e speed in n days.
According to the speed interval of pavement marker, all of road data is divided into four data set Di, i ∈ { 1,2,3,4 }.
If<e,t,s>Component s ∈ Si, then the record of time period ts of the road e in n days<e,tj,speed>∈Di, j=1,
2,……n}。
Next it is based on data set DiBuild Isolation Forest Fi={ Tk| k=1,2 ..., m } for detecting city
City's road traffic exception, FiIt is comprising the m forest of abnormal tree.
The algorithm for building Isolation Forest F based on data set D is as follows：
The developing algorithm of abnormal tree is as follows：
One road record of detection<e,tj,speed>, i.e. average speed speeds of the road e in the time period t of jth day
It is whether abnormal, first find the speed interval S belonging to iti, that is, have found corresponding IsolationForest Fi, then calculate
Road is recorded in FiIn mean depth.
Calculate mean depth algorithm of the road record in Isolation Forest as follows：
Record x can be calculated in Isolation Forest F according to Algorithm 3iIn mean depth, that is, remember
Record x is length apart from the desired length of root node, and exceptional value is calculated according to formula (1)
C (ψ) is calculated by formula (2).
H (ψ -1)=ln (ψ -1)+0.57721 (Euler's constant)
ψ values 256 in the present invention.
S is on length monotone decreasings, if the return value s of record closely 1, shows record from root node close to, logical
Crossing little division just can isolate, then this record may be considered abnormal.
If return value is far smaller than 0.5, illustrating recording distance root node farther out could remember this, it is necessary to repeatedly divide
Record is isolated, then this record may be considered normal.
The present invention is tested to the above method, and has obtained obvious effect.Experimental Area is at Beijing five rings
It is interior.We build the road network of Beijing with OpenStreeMap (OSM) data, weed out the impassable road of some vehicles
Afterwards, 39951 roads are had.Car data is hired out from June 14,1 day to 2013 June in 2013, wrong data is weeded out and is not inconsistent
After closing desired data, daily data volume is about more than 10,000,000.
Method according to us calculate 14 days in daily exception section number account for overall ratio, as shown in Figure 2, it can be seen that
In on June 8th, 2013 and the exception of June 9 section number showed increased.And the Beijing issued according to Beijing Transportation Research Center
Pointed out in city's road traffic operating analysis report, work of not restricted driving before on June 8th, 2013 and June 9 happen the vacation Dragon Boat Festival
Day, Beijing's college entrance examination and the factor of rainfall three superposition, road traffic congestion situation are protruded, particularly the peak traffic index of June 9
Reach June peak 8.4.
And the data of Beijing Communication committee issue display that June 8 and the road traffic index on June 9 be substantially higher in
Other several days.Table 1 is the road traffic index of Beijing Communication committee issue.
Table in June, 1 2013 Beijing Communication index
Fig. 3 is 8 days 18 June in 2013：20：Traffic conditions when 00.Wherein black line part represents what this method was detected
Abnormal section.
And Isolation Forest are also better than the normal deviate detection algorithm based on road in time.Both are calculated
Whether the time that exception is consumed is all in 1ms or so in one record of detection for method, but normal deviate algorithm in the training process
The required time is linear with road quantity, and road quantity is more, and the time of its training need is more long.And Isolation
Forest algorithms are because be to build abnormal tree based on sampled data, therefore train the consumed time unrelated with road quantity.
The time that Fig. 4 is consumed by two kinds of algorithms in training.It can be seen that when road quantity is more, Isolation
The time that Forest is consumed will be considerably less than normal deviate detection algorithm.
Claims (3)
1. a kind of urban highway traffic method for detecting abnormality based on Isolation Forest, it is characterised in that including following
Step：
Step 1, with road as detection object, the average running speed according to road in different periods divides different classes of data
Collection；
Step 2, based on each data set train an Isolation Forest, by detecting road speeds in Isolation
Judge whether road is abnormal in Forest to the distance of root node.
2. the urban highway traffic method for detecting abnormality of Isolation Forest, its feature are based on as claimed in claim 1
It is that step 1 is specially：
If city road network is represented with a digraph G：
G=(V, E)
V is the point set of digraph, and each point represents the intersection in road network, two element group representations being made up of longitude and latitude；E is
The side collection of digraph, each edge represents certain road in road network；
Floating Car returns to vehicle-state by the GPS device for carrying, and the form of returned data is as follows：
Record=[carID, time, latitude, lontitude, speed, angel]
Wherein carID represents the license plate number of vehicle, and time represents time during return recording, and latitude, lontitude are represented
The longitude and latitude of vehicle position during return recording, speed represent car speed, and angel represents that vehicle travels angle；
First, map match carried out to the data that Floating Car is returned, calculate with Floating Car returned data (latitude,
Lontitude) the point on immediate road, even if the correct position that floating car data is matched with section；
Secondly, the road data after floating car data is matched is classified, and classification has two standards of dimension of time and speed：
1) to daily 06：00 to 21：00 time period was divided with prefixed time interval；2) by road speeds in each time period according to
Urban highway traffic assessment indicator system is divided into four speed intervals,
Then it is interval according to time period signature velocity to all roads in road network, i.e.,<e,t,s>Represent in time period t, road e category
In speed interval s, s ∈ { S1,S2,S3,S4, e ∈ E, wherein s are the averages of the time period t road e speed in n days；
According to the speed interval of pavement marker, all of road data is divided into four data set Di, i ∈ { 1,2,3,4 }, if<
e,t,s>Component s ∈ Si, then the record of time period ts of the road e in n days<e,tj,speed>∈Di, j=1,2 ...
n}。
3. the urban highway traffic method for detecting abnormality of Isolation Forest, its feature are based on as claimed in claim 1
It is that step 2 specifically includes following steps and is：
Step 2.1, based on data set DiBuild Isolation Forest Fi={ Tk| k=1,2 ..., m } for detecting city
Road traffic exception, FiIt is comprising the m forest of abnormal tree；
Step 2.2, one road record of detection<e,tj,speed>, i.e. average speeds of the road e in the time period t of jth day
Whether speed is abnormal, first finds the speed interval S belonging to iti, that is, have found corresponding Isolation Forest Fi, then
Road record is calculated in FiIn mean depth；
Step 2.3, basis are calculated road record in Isolation Forest FiIn mean depth, i.e., road record away from
It is length from the desired length of root node, exceptional value is calculated according to formula (1)
C (ψ) is calculated by formula (2).
H (ψ -1)=ln (ψ -1)+0.57721 (Euler's constant)
Wherein, ψ is 256,
If the return value s closely 1 of record, show that record close to root node, just can be isolated by little division,
Then this record may be considered abnormal；
If return value is far smaller than 0.5, illustrate recording distance root node farther out, it is necessary to repeatedly divide could by this record every
Separate out and, then this record may be considered normal.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710131835.4A CN106846806A (en) | 2017-03-07 | 2017-03-07 | Urban highway traffic method for detecting abnormality based on Isolation Forest |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710131835.4A CN106846806A (en) | 2017-03-07 | 2017-03-07 | Urban highway traffic method for detecting abnormality based on Isolation Forest |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN106846806A true CN106846806A (en) | 2017-06-13 |
Family
ID=59138180
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710131835.4A Pending CN106846806A (en) | 2017-03-07 | 2017-03-07 | Urban highway traffic method for detecting abnormality based on Isolation Forest |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN106846806A (en) |
Cited By (14)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107992741A (en) * | 2017-10-24 | 2018-05-04 | 阿里巴巴集团控股有限公司 | A kind of model training method, the method and device for detecting URL |
| CN108154029A (en) * | 2017-10-25 | 2018-06-12 | 上海观安信息技术股份有限公司 | Intrusion detection method, electronic equipment and computer storage media |
| CN108900476A (en) * | 2018-06-07 | 2018-11-27 | 桂林电子科技大学 | Based on Spark and the parallel network flow method for detecting abnormality that forest is isolated |
| CN108921440A (en) * | 2018-07-11 | 2018-11-30 | 平安科技（深圳）有限公司 | Pollutant method for monitoring abnormality, system, computer equipment and storage medium |
| CN109325691A (en) * | 2018-09-27 | 2019-02-12 | 上海观安信息技术股份有限公司 | Abnormal behaviour analysis method, electronic equipment and computer program product |
| CN109561052A (en) * | 2017-09-26 | 2019-04-02 | 北京国双科技有限公司 | The detection method and device of website abnormal flow |
| CN109886724A (en) * | 2018-12-29 | 2019-06-14 | 中南大学 | A kind of resident trip track recognizing method of robustness |
| CN110659280A (en) * | 2019-08-01 | 2020-01-07 | 北京百度网讯科技有限公司 | Road blocking abnormity detection method and device, computer equipment and storage medium |
| CN110705635A (en) * | 2019-09-29 | 2020-01-17 | 京东城市(北京)数字科技有限公司 | Method and apparatus for generating an isolated forest |
| CN111081016A (en) * | 2019-12-18 | 2020-04-28 | 北京航空航天大学 | Urban traffic abnormity identification method based on complex network theory |
| CN112541981A (en) * | 2020-11-03 | 2021-03-23 | 山东中创软件商用中间件股份有限公司 | ETC portal system early warning method, device, equipment and medium |
| US11824877B2 (en) | 2020-11-10 | 2023-11-21 | Armis Security Ltd. | System and method for anomaly detection interpretation |
| CN117874653A (en) * | 2024-03-11 | 2024-04-12 | 武汉佳华创新电气有限公司 | Power system safety monitoring method and system based on multi-source data |
| CN117874653B (en) * | 2024-03-11 | 2024-05-31 | 武汉佳华创新电气有限公司 | Power system safety monitoring method and system based on multi-source data |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102231235A (en) * | 2011-04-29 | 2011-11-02 | 陈伟 | Method for detecting and positioning traffic flow abnormal point |
| CN102779420A (en) * | 2012-07-31 | 2012-11-14 | 哈尔滨工业大学 | Road traffic event automatic detection method based on real-time vehicle-mounted GPS (global position system) data |
| CN103748621A (en) * | 2011-06-29 | 2014-04-23 | 因瑞克斯控股有限公司 | Method and system for collecting traffic data |
| CN103761430A (en) * | 2014-01-10 | 2014-04-30 | 安徽科力信息产业有限责任公司 | Method for identifying peak periods of road networks on basis of floating cars |
| CN104809878A (en) * | 2015-05-14 | 2015-07-29 | 重庆大学 | Method for detecting abnormal condition of urban road traffic by utilizing GPS (Global Positioning System) data of public buses |
-
2017
- 2017-03-07 CN CN201710131835.4A patent/CN106846806A/en active Pending
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102231235A (en) * | 2011-04-29 | 2011-11-02 | 陈伟 | Method for detecting and positioning traffic flow abnormal point |
| CN103748621A (en) * | 2011-06-29 | 2014-04-23 | 因瑞克斯控股有限公司 | Method and system for collecting traffic data |
| CN102779420A (en) * | 2012-07-31 | 2012-11-14 | 哈尔滨工业大学 | Road traffic event automatic detection method based on real-time vehicle-mounted GPS (global position system) data |
| CN103761430A (en) * | 2014-01-10 | 2014-04-30 | 安徽科力信息产业有限责任公司 | Method for identifying peak periods of road networks on basis of floating cars |
| CN104809878A (en) * | 2015-05-14 | 2015-07-29 | 重庆大学 | Method for detecting abnormal condition of urban road traffic by utilizing GPS (Global Positioning System) data of public buses |
Non-Patent Citations (5)
| Title |
|---|
| DONG XU 等: "An Improved Data Anomaly Detection Method Based on Isolation Forest", 《2017 10TH INTERNATIONAL SYMPOSIUM ON COMPUTATIONAL INTELLIGENCE AND DESIGN (ISCID)》 * |
| FEI TONY LIU 等: "Isolation Forest", 《2008 EIGHTH IEEE INTERNATIONAL CONFERENCE ON DATA MINING》 * |
| XIAOLEI LI 等: "Temporal Outlier Detection in Vehicle Traffic Data", 《2009 IEEE 25TH INTERNATIONAL CONFERENCE ON DATA ENGINEERING》 * |
| 侯泳旭 等: "基于Isolation Forest的城市道路交通异常检测", 《计算机工程与科学》 * |
| 牛爽: "基于浮动车的城市道路交通异常事件检测的研究", 《中国优秀硕士学位论文全文数据库》 * |
Cited By (24)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109561052A (en) * | 2017-09-26 | 2019-04-02 | 北京国双科技有限公司 | The detection method and device of website abnormal flow |
| CN109561052B (en) * | 2017-09-26 | 2022-01-28 | 北京国双科技有限公司 | Method and device for detecting abnormal flow of website |
| CN107992741A (en) * | 2017-10-24 | 2018-05-04 | 阿里巴巴集团控股有限公司 | A kind of model training method, the method and device for detecting URL |
| CN112182578A (en) * | 2017-10-24 | 2021-01-05 | 创新先进技术有限公司 | Model training method, URL detection method and device |
| CN108154029A (en) * | 2017-10-25 | 2018-06-12 | 上海观安信息技术股份有限公司 | Intrusion detection method, electronic equipment and computer storage media |
| CN108900476A (en) * | 2018-06-07 | 2018-11-27 | 桂林电子科技大学 | Based on Spark and the parallel network flow method for detecting abnormality that forest is isolated |
| CN108900476B (en) * | 2018-06-07 | 2021-05-11 | 桂林电子科技大学 | Parallel network flow abnormity detection method based on Spark and forest isolation |
| CN108921440A (en) * | 2018-07-11 | 2018-11-30 | 平安科技（深圳）有限公司 | Pollutant method for monitoring abnormality, system, computer equipment and storage medium |
| WO2020010701A1 (en) * | 2018-07-11 | 2020-01-16 | 平安科技（深圳）有限公司 | Pollutant anomaly monitoring method and system, computer device, and storage medium |
| CN109325691B (en) * | 2018-09-27 | 2020-10-16 | 上海观安信息技术股份有限公司 | Abnormal behavior analysis method, electronic device and computer program product |
| CN109325691A (en) * | 2018-09-27 | 2019-02-12 | 上海观安信息技术股份有限公司 | Abnormal behaviour analysis method, electronic equipment and computer program product |
| CN109886724A (en) * | 2018-12-29 | 2019-06-14 | 中南大学 | A kind of resident trip track recognizing method of robustness |
| CN109886724B (en) * | 2018-12-29 | 2021-02-12 | 中南大学 | Robust resident travel track identification method |
| CN110659280A (en) * | 2019-08-01 | 2020-01-07 | 北京百度网讯科技有限公司 | Road blocking abnormity detection method and device, computer equipment and storage medium |
| CN110659280B (en) * | 2019-08-01 | 2022-05-27 | 北京百度网讯科技有限公司 | Road blocking abnormity detection method and device, computer equipment and storage medium |
| CN110705635B (en) * | 2019-09-29 | 2020-11-03 | 京东城市(北京)数字科技有限公司 | Method and apparatus for generating an isolated forest |
| CN110705635A (en) * | 2019-09-29 | 2020-01-17 | 京东城市(北京)数字科技有限公司 | Method and apparatus for generating an isolated forest |
| CN111081016A (en) * | 2019-12-18 | 2020-04-28 | 北京航空航天大学 | Urban traffic abnormity identification method based on complex network theory |
| CN111081016B (en) * | 2019-12-18 | 2021-07-06 | 北京航空航天大学 | Urban traffic abnormity identification method based on complex network theory |
| CN112541981A (en) * | 2020-11-03 | 2021-03-23 | 山东中创软件商用中间件股份有限公司 | ETC portal system early warning method, device, equipment and medium |
| CN112541981B (en) * | 2020-11-03 | 2022-07-22 | 山东中创软件商用中间件股份有限公司 | ETC portal system early warning method, device, equipment and medium |
| US11824877B2 (en) | 2020-11-10 | 2023-11-21 | Armis Security Ltd. | System and method for anomaly detection interpretation |
| CN117874653A (en) * | 2024-03-11 | 2024-04-12 | 武汉佳华创新电气有限公司 | Power system safety monitoring method and system based on multi-source data |
| CN117874653B (en) * | 2024-03-11 | 2024-05-31 | 武汉佳华创新电气有限公司 | Power system safety monitoring method and system based on multi-source data |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN106846806A (en) | Urban highway traffic method for detecting abnormality based on Isolation Forest | |
| Li et al. | Link travel time estimation using single GPS equipped probe vehicle | |
| CA2811879C (en) | Driver profiling system and method | |
| CN100589143C (en) | Method and appaatus for judging the traveling state of a floating vehicle | |
| CN111612670B (en) | Method, device and computer equipment for constructing emission list of motor vehicle | |
| CN105869402B (en) | Express highway section speed modification method based on polymorphic type floating car data | |
| CN102508949A (en) | Digital analyzing method of city tunnel motor vehicle emission | |
| CN105810006A (en) | Method and system for recognizing roadside parking place | |
| Kumar et al. | A model based approach to predict stream travel time using public transit as probes | |
| Liu et al. | Evaluation of floating car technologies for travel time estimation | |
| CN107240264A (en) | A kind of non-effective driving trace recognition methods of vehicle and urban road facility planing method | |
| CN104900057A (en) | City expressway main and auxiliary road floating vehicle map matching method | |
| CN107730892A (en) | A kind of traffic congestion index number evaluation method merged based on FCD with internet data | |
| CN103942952B (en) | A kind of road network functional hierarchy state grade appraisal procedure | |
| Hao et al. | Evaluating the environmental impact of traffic congestion based on sparse mobile crowd-sourced data | |
| Gore et al. | Exploring credentials of Wi‐Fi sensors as a complementary transport data: an Indian experience | |
| Shang et al. | Analyzing the effects of road type and rainy weather on fuel consumption and emissions: A mesoscopic model based on big traffic data | |
| TW201327458A (en) | Transportation route network generation method using vehicle detection data | |
| Scora et al. | Real-time roadway emissions estimation using visual traffic measurements | |
| CN116092037B (en) | Vehicle type identification method integrating track space-semantic features | |
| Hao et al. | Modal activity-based vehicle energy/emissions estimation using sparse mobile sensor data | |
| Vasantha Kumar et al. | Urban arterial travel time estimation using buses as probes | |
| Dinh et al. | Real-time queue-end detection on freeways with floating car data: Practice-ready algorithm | |
| Habtie et al. | Cellular network based real-time urban road traffic state estimation framework using neural network model estimation | |
| CN113611130B (en) | Method, system and storage medium for acquiring traffic flow of local and transit trucks |
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: 20170613 |