CN114495498B - Traffic data distribution effectiveness judging method and device - Google Patents
Traffic data distribution effectiveness judging method and device Download PDFInfo
- Publication number
- CN114495498B CN114495498B CN202210063786.6A CN202210063786A CN114495498B CN 114495498 B CN114495498 B CN 114495498B CN 202210063786 A CN202210063786 A CN 202210063786A CN 114495498 B CN114495498 B CN 114495498B
- Authority
- CN
- China
- Prior art keywords
- traffic flow
- day
- flow data
- data
- parameter
- 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.)
- Active
Links
Images
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
- G08G1/0125—Traffic data processing
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/23—Clustering techniques
- G06F18/232—Non-hierarchical techniques
- G06F18/2321—Non-hierarchical techniques using statistics or function optimisation, e.g. modelling of probability density functions
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
- G06F18/2415—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on parametric or probabilistic models, e.g. based on likelihood ratio or false acceptance rate versus a false rejection rate
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/047—Probabilistic or stochastic networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- 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
- G08G1/0133—Traffic data processing for classifying traffic situation
-
- G—PHYSICS
- G08—SIGNALLING
- G08G—TRAFFIC CONTROL SYSTEMS
- G08G1/00—Traffic control systems for road vehicles
- G08G1/065—Traffic control systems for road vehicles by counting the vehicles in a section of the road or in a parking area, i.e. comparing incoming count with outgoing count
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Evolutionary Computation (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- General Engineering & Computer Science (AREA)
- Probability & Statistics with Applications (AREA)
- Biomedical Technology (AREA)
- Software Systems (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- Computational Linguistics (AREA)
- Biophysics (AREA)
- Mathematical Physics (AREA)
- General Health & Medical Sciences (AREA)
- Health & Medical Sciences (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Evolutionary Biology (AREA)
- Chemical & Material Sciences (AREA)
- Analytical Chemistry (AREA)
- Traffic Control Systems (AREA)
Abstract
The application discloses a method and a device for judging traffic data distribution effectiveness. Acquiring single-day traffic flow data collected by traffic flow detection equipment with data distribution effectiveness to be identified; preprocessing the single-day traffic flow data; performing parameter distribution characteristic extraction on the preprocessed single-day traffic flow data to obtain the distribution characteristics of the single-day traffic flow data; based on the trained neural network classifier, obtaining the probability that the traffic flow data of the single day belongs to each category according to the distribution characteristics of the traffic flow data of the single day; determining whether the distribution characteristics of the traffic flow data on the single day are valid according to the probability that the traffic flow data on the single day belongs to each category, and determining the category to which the traffic flow data on the single day is determined to be valid.
Description
Technical Field
The application relates to the technical field of intelligent traffic, in particular to a method and a device for judging traffic data distribution effectiveness.
Background
With the continuous progress of informatization and intellectualization of urban traffic systems, a large number of traffic flow detection devices are arranged in urban areas to support diversified applications such as traffic management, control and planning data. However, due to the influence of factors such as untimely operation and maintenance, complex scene, algorithm defects, and the like, the data collected by the traffic flow detection device may be abnormal, which results in insufficient data availability and reliability, and further affects traffic management, control, planning, and the like, so that the validity of traffic data distribution needs to be determined.
Disclosure of Invention
The exemplary embodiment of the application provides a method and a device for judging the validity of traffic data distribution, which are used for judging the validity of the traffic data distribution.
In a first aspect, a method for determining traffic data distribution validity is provided, including:
acquiring single-day traffic flow data collected by traffic flow detection equipment with data distribution effectiveness to be identified, wherein the single-day traffic flow data comprises at least one traffic flow parameter collected aiming at least one lane;
preprocessing the single-day traffic flow data;
performing parameter distribution characteristic extraction on the preprocessed single-day traffic flow data to obtain the distribution characteristics of the single-day traffic flow data;
based on the trained neural network classifier, obtaining the probability that the traffic flow data of the single day belongs to each category according to the distribution characteristics of the traffic flow data of the single day;
determining whether the distribution characteristics of the traffic flow data on the single day are valid according to the probability that the traffic flow data on the single day belongs to each category, and determining the category to which the traffic flow data on the single day is determined to be valid.
Optionally, the preprocessing the single-day traffic flow data includes: carrying out normalization processing on the single-day traffic flow data, wherein any traffic flow parameter after normalization processing meets the following formula:
wherein X "represents a traffic flow parameter after normalization, X 'represents a traffic flow parameter before normalization, and X' max Represents the maximum value X 'of the traffic flow parameter in the single-day traffic flow data' min And the minimum value of the traffic flow parameter in the traffic flow data of the single day is represented.
Optionally, before performing the normalization processing on the single-day traffic flow data, the method further includes: and according to a set aggregation period, aggregating the single-day traffic flow data into the single-day traffic flow data of the aggregation period.
Optionally, if the single-day traffic flow data includes at least one of flow, speed, and occupancy, then:
the flow after polymerization satisfies the following formula:
the velocity after polymerization satisfies the following formula:
the occupancy after polymerization satisfies the following formula:
wherein, q' k Denotes the flow rate of the k-th aggregation period, q i Representing the flow of the ith acquisition cycle; v' k Denotes the velocity, v, of the k-th polymerization cycle i Representing the speed of the ith acquisition cycle; o 'to' k Denotes the occupancy of the k-th polymerization cycle, o i Representing the occupancy of the ith acquisition cycle; t is 0 Denotes the aggregation period and T denotes the acquisition period.
Optionally, before performing normalization processing on the traffic flow data for a single day, the method further includes: and eliminating abnormal values in the traffic flow data of the single day by adopting a box type graph method.
Optionally, the single-day traffic flow data includes a single traffic flow parameter; the parameter distribution characteristic extraction of the preprocessed traffic flow data on a single day comprises the following steps: uniformly dividing the value intervals of the single-day traffic flow parameters into sub-intervals with set number; and counting the frequency of the single-day traffic flow parameters falling into each subinterval.
Optionally, the single-day traffic flow data includes a first traffic flow parameter and a second traffic flow parameter, and the first traffic flow parameter and the second traffic flow parameter are different types of traffic flow parameters; the parameter distribution characteristic extraction of the preprocessed traffic flow data on a single day comprises the following steps: dividing the grids according to the set grid size, the value interval of the first traffic flow parameter and the value interval of the second traffic flow parameter, wherein the horizontal dimension of each divided grid is the first traffic flow parameter, and the vertical dimension is the second traffic flow parameter; and counting the frequency of the single-day traffic flow parameters falling into each grid.
Optionally, the training process of the neural network classifier includes: acquiring single-day traffic flow data acquired by traffic flow detection equipment with data distribution effectiveness identified as a sample for training, wherein the single-day traffic flow data comprises at least one traffic flow parameter acquired aiming at least one lane; pre-treating the sample; performing parameter distribution characteristic extraction on the preprocessed sample to obtain the distribution characteristic of the preprocessed sample; determining the distance between the distribution characteristics of each sample in the preprocessed samples based on a relative entropy function, and clustering the preprocessed samples according to the distance between the distribution characteristics of each sample in the preprocessed samples by adopting a density-based clustering algorithm to obtain the category of each sample in the preprocessed samples; training the neural network classifier using class-labeled samples.
In a second aspect, a traffic data distribution validity determination apparatus is provided, including:
the acquisition module can be configured to acquire single-day traffic flow data acquired by traffic flow detection equipment for validity of distribution of data to be identified, wherein the single-day traffic flow data comprises at least one traffic flow parameter acquired for at least one lane;
a preprocessing module configured to preprocess the single-day traffic flow data;
the characteristic extraction module is configured to perform parameter distribution characteristic extraction on the preprocessed single-day traffic flow data to obtain the distribution characteristics of the single-day traffic flow data;
the classification module is configured to obtain the probability that the single-day traffic flow data belongs to each category according to the distribution characteristics of the single-day traffic flow data based on the trained neural network classifier;
and the validity judging module is configured to determine whether the distribution characteristics of the traffic flow data on the single day are valid according to the probability that the traffic flow data on the single day belongs to each class, and determine the class to which the traffic flow data on the single day belongs according to the probability that the traffic flow data on the single day belongs.
In a third aspect, a communications apparatus is provided that includes a memory and a processor; the memory storing computer instructions; the processor is configured to read the computer instructions to perform the method according to any one of the above first aspects.
In a fourth aspect, a computer-readable storage medium is provided. A computer-readable storage medium stores computer-executable instructions for causing a computer to perform the method of any one of the first aspects described above.
In a fifth aspect, there is provided a computer program product which, when invoked by a computer, causes the computer to perform the method of any of the first aspects above.
In the embodiment of the application, after the distribution characteristics of the traffic flow data of a single day are extracted, the probability that the traffic flow data of the single day belongs to each category is obtained according to the distribution characteristics of the traffic flow data of the single day based on the trained neural network classifier, whether the distribution characteristics of the traffic flow data of the single day are effective or not is determined according to the probability that the traffic flow data of the single day belongs to each category, the category to which the traffic flow data of the single day is determined to be effective is determined, and therefore the judgment on the effectiveness of the traffic data distribution is achieved.
On the basis of the common knowledge in the field, the above preferred conditions can be combined randomly to obtain the preferred embodiments of the application.
Drawings
In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings needed to be used in the description of the embodiments or the prior art will be briefly introduced below, and it is obvious that the drawings in the following description are some embodiments of the present application, and for those skilled in the art, other drawings can be obtained according to these drawings without inventive exercise.
Fig. 1 schematically illustrates a flow chart of a traffic data distribution validity determination method provided in an embodiment of the present application;
FIG. 2 is a schematic diagram illustrating a box graph algorithm in an embodiment of the present application;
fig. 3a and 3b respectively illustrate a velocity distribution histogram in the embodiment of the present application;
FIG. 4a is a scatter plot of traffic-occupancy relationships in an embodiment of the present application;
FIG. 4b is a schematic diagram illustrating a traffic-occupancy relationship scattergram gridding in an embodiment of the present application;
FIG. 5 is a schematic diagram illustrating a training flow of a neural network classifier provided by an embodiment of the present application;
fig. 6 is a schematic structural diagram illustrating an example of a traffic data distribution validity determination apparatus provided in an embodiment of the present application;
fig. 7 schematically illustrates a structure of a communication apparatus according to an embodiment of the present application.
Detailed Description
The technical solutions in the embodiments of the present application will be described in detail and clearly with reference to the accompanying drawings. In the description of the embodiments herein, "/" means "or" unless otherwise specified, for example, a/B may mean a or B; "and/or" in the text is only an association relationship describing an associated object, and means that three relationships may exist, for example, a and/or B may mean: three cases of a alone, a and B both, and B alone exist, and in addition, "a plurality" means two or more than two in the description of the embodiments of the present application.
In the following, the terms "first", "second" are used for descriptive purposes only and are not to be construed as implying or implying relative importance or implicitly indicating the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of that feature, and in the description of embodiments of the application, unless stated otherwise, "plurality" means two or more.
The embodiment of the application extracts features (such as extracting features of typical single-parameter distribution or double-parameter combined distribution of traffic flow) of traffic flow data acquired by each traffic flow detection device based on traffic big data, classifies the distribution of the traffic flow data by adopting a neural network classifier according to the extracted features, and can judge whether the distribution of the traffic flow data is effective or not according to the classification result, so that the technical feasibility challenge brought by classifying traffic facilities and traffic states in advance is avoided. Furthermore, the embodiment of the application can automatically divide the positions with similar traffic states, and can provide support for subsequent specific decision-making processes such as traffic management.
Referring to fig. 1, a flow chart of a method for determining traffic data distribution validity according to an embodiment of the present application is schematically shown. The flow can be executed by a traffic data distribution effectiveness judging device, and the device can be realized by a software mode, a hardware mode or a mode of combining software and hardware.
As shown, the process may include the following steps:
s101: acquiring single-day traffic flow data collected by traffic flow detection equipment for distribution effectiveness of data to be identified, wherein the single-day traffic flow data comprises at least one traffic flow parameter collected aiming at least one lane.
The traffic flow detection device in the embodiment of the application may include: the embodiment of the present application does not limit the devices such as a flow geomagnetic detector and a traffic flow detector.
In the embodiment of the present application, the acquired traffic flow data may include a single type of traffic flow parameter, such as a flow rate, or may include multiple types of traffic flow parameters, such as a flow rate, a speed, an occupancy, and the like.
In some embodiments, the traffic flow data collected by each traffic flow detection device may be obtained from a database, and the single-day traffic flow parameters (flow Q, speed V, occupancy O) of the lane level may be obtained. For example, for a certain lane, the collection cycle of the traffic flow detection device is T (unit is minute), and the traffic flow data corresponding to the lane can be represented as a parameter X (X = Q, V, O), where Q represents flow, V represents speed, and O represents occupancy. The parameter X of the k-th acquisition cycle of a single day is denoted X k The traffic flow data for the lane may be expressed as { x } on a single day k }，k＝1，2，...，1440/T。
S102: and preprocessing the single-day traffic flow data.
In some embodiments, the pre-processing of the single-day traffic flow data may include normalization; in other embodiments, the normalization process may further include a data aggregation process, for example, the preprocessing process may include a data aggregation process and a normalization process; in other embodiments, outlier rejection may be included prior to normalization, for example, the preprocessing may include data aggregation, outlier rejection, and normalization. The data aggregation, outlier culling, and normalization processes are described in detail below, respectively.
(1) And (5) data aggregation processing.
In order to ensure that the distribution validity is determined at the same time granularity due to the fact that the traffic flow data acquisition periods of different traffic flow detection devices may be inconsistent, in the embodiment of the present application, the traffic flow data X of a single day may be aggregated into the traffic flow data X' of a single day in the same period (referred to as an aggregation period herein) according to a set aggregation period.
The acquisition period of the traffic flow data of a single day is T, and the aggregation period is T 0 As an example, the traffic flow data of the k aggregation cycle after aggregation can be represented as:
the flow after polymerization satisfies the following formula:
wherein, q' k Denotes the flow rate of the k-th aggregation period, q i Indicating the flow for the ith acquisition cycle.
The velocity after polymerization satisfies the following formula:
wherein, v' k Denotes the velocity, v, of the k-th polymerization cycle i Representing the speed of the ith acquisition cycle.
The occupancy after polymerization satisfies the following formula:
wherein, o' k Denotes the occupancy of the k-th polymerization cycle, o i Representing the occupancy of the ith acquisition cycle.
For example, for the traffic flow parameter of flow, the original collection period is 1 minute, and 1440 sets of flow data { q ] are collected together on the same day 1 ，q 2 ，...，q 1440 }. When the polymerization cycle is 5 minutes, the polymerization flow rate in the 1 st polymerization cycle after the polymerization
Polymerization flow rate of 2 nd polymerization cycle
By analogy, the flow rate after polymerization is represented by { q' 1 ，q′ 2 ，...，q′ 288 }。
(2) And removing abnormal values.
Alternatively, a box diagram (also called box and whisker diagram) method can be adopted to remove abnormal values in the traffic flow data of a single day. Of course, other methods may be adopted to eliminate the abnormal value in the traffic flow data of a single day, which is not limited in the embodiment of the present application.
In order to eliminate the influence of the abnormal values on data normalization by using a box type graph method as an example, the abnormal values in the traffic flow data of a single day are eliminated by using the box type graph method in the embodiment of the application. Fig. 2 illustrates the principle of the box plot, taking the minimum value in the box plot as an example, adding 1.5 times the inner distance (IQR) from the upper quartile point, the maximum value of the variable in this range is called the maximum value, and the values of the inner distance exceeding 1.5 times are called outliers (outliers). The following describes the elimination of abnormal values in traffic flow data of a single day by using a box diagram with reference to fig. 2, where the process may include the following steps: sequencing the traffic flow data of each aggregation period in a single day according to the sequence of the aggregation periods from front to back; calculating the lower quartile point Q1 and the upper quartile point Q3 of the time values in the data; inner distances IQR for calculating time values in these data: IQR = Q3-Q1; the lower edge LB of the time values in these data is calculated: LB = Q1-1.5 × iqr; calculate the upper edge UB of the time value in these data: UB = Q3+1.5 by iqr; from the time values in these data, the time value between the upper edge UB and the lower edge LB is selected, and data outside this range is discarded.
(3) And (6) normalization processing.
In order to focus on the distribution trend of parameters in a reasonable range and reduce the influence of value difference caused by factors such as different road types, the embodiment of the application can perform normalization processing on single-day traffic flow data.
Optionally, for any one traffic flow parameter, the traffic flow parameter after normalization processing satisfies the following formula:
wherein X "represents a traffic flow parameter after normalization, X 'represents a traffic flow parameter before normalization, and X' max Represents the maximum value, X 'of the traffic flow parameter' min Representing the minimum value of the traffic flow parameter.
For example, taking a lane with a road speed limit of 80 km/h as an example, after normalization processing is performed in the above manner, the vehicle speed distribution of the lane is a speed distribution histogram shown in fig. 3 a; taking the lane with the road speed limit of 60 km/h as an example, after normalization processing is performed in the above manner, the vehicle speed distribution of the lane is a speed distribution histogram as shown in fig. 3 b.
According to the speed distribution histogram, although the value intervals of the speeds of the two lanes are different under the condition of different road speed limits, the general distribution trends of the speeds of the two lanes are consistent, the two lanes comprise two peak values, one peak is located in a low-speed area, and the other peak is close to the speed limit. Thus, by the normalization process, the comparison of the parameter distributions can be performed at the same scale.
S103: and performing parameter distribution characteristic extraction on the preprocessed traffic flow data of the single day to obtain the distribution characteristics of the traffic flow data of the single day.
The embodiment of the application can be used for extracting the features of single parameters and double parameters. The following describes the feature extraction of single and double parameters in detail.
(1) And extracting the single-parameter distribution characteristics.
Optionally, if the single-day traffic flow data includes a single traffic flow parameter, the following method may be adopted for feature extraction: and uniformly dividing the value interval of the single-day traffic flow parameter into a set number of sub-intervals, and counting the frequency of the single-day traffic flow parameter falling into each sub-interval.
By adopting the method, the histogram of the single-day traffic flow data can be obtained. If the histogram is a graphical representation of the distribution of numerical data, then the histogram can be used to characterize the distribution for the normalized (i.e., normalized) parameter X ". For example, in fig. 3a, the range of the vehicle speed from 0 to 100 km/h is divided into 20 sections with 0.05 (i.e., 5 km/h) as a set pitch, and the frequency P = { P (1), P (2),.., P (20) } at which the vehicle speed falls in each section is calculated.
(2) And (5) extracting the double-parameter joint distribution characteristics.
Optionally, if the single-day traffic flow data includes the first traffic flow parameter and the second traffic flow parameter, the following method may be adopted for feature extraction: dividing the grids according to the set grid size, the value interval of the first traffic flow parameter and the value interval of the second traffic flow parameter, wherein the horizontal dimension of each divided grid is the first traffic flow parameter, and the vertical dimension is the second traffic flow parameter; and counting the frequency of the single-day traffic flow parameters falling into each grid.
In the theoretical study of traffic flow, the relationship of traffic flow parameters is generally described in the form of a scatter diagram, such as a traffic flow-occupancy relationship scatter diagram shown in fig. 4 a. In the embodiment of the present application, in order to extract the feature of the two-parameter joint distribution, for the normalized parameter X "and parameter Y", the Y "-X" relationship scattergram is gridded, the size of the grid is preset, for example, the size of the grid is 0.05 × 0.05, as shown in the gridding schematic diagram of the normalized flow-occupancy relationship scattergram shown in fig. 4 b. From this, the frequency at which the scatter points fall within each grid can be calculated.
Alternatively, the frequency distribution table may be used to characterize the two-parameter distribution, as shown in table 1.
Table 1: two parameter frequency distribution table
Alternatively, the 20 × 20 matrix in table 1 above may be converted into a one-dimensional vector:
P＝{p 1，1 ，p 1，2 ，...，p 1，20 ，p 2，1 ，p 2，2 ，...，p 20,20 }＝{p(1)，p(2)，...，p(400)}。
in the aspect of standardized expression of traffic flow parameter distribution characteristics, in the traditional traffic flow theory, single-parameter and multi-parameter distribution is mainly depicted by adopting a method for establishing a mathematical model, and the method is more dependent on a traffic flow dynamics theoretical model. Aiming at single traffic flow parameter distribution, the industry describes the distribution condition of traffic flow parameters such as flow, occupancy, speed and the like through mathematical models such as mixed Gaussian distribution, gamma distribution, lognormal distribution, negative exponential distribution and the like aiming at different traffic states such as free flow, blocking state and the like. However, the model selection has difficulty because the traffic state is in dynamic change. For the relationship between traffic flow Two parameters, the classical traffic models include GreenShield Model, greenberg Model, two-regiomemodel, three-regiome Model, etc. However, the model introduces a large number of parameters, requires a relatively complex calibration process, and has insufficient operability in engineering application. In addition, in the traditional traffic flow theory, a double-parameter theoretical model is mainly based on density, a flow-density or speed-density mathematical model is established, actually acquired traffic flow parameters are occupancy, and the two traffic flow parameters need to be converted through effective vehicle length, so that errors are further introduced in the process of calibrating the model, and the effectiveness judgment effect is poor. How to overcome the application bottlenecks of difficult selection of a traffic flow distribution model and complex parameter calibration, and realizing the description of the distribution form of the traffic data of a single parameter popularized to multiple parameters by adopting a unified standardized method without depending on a strong assumption condition of the traffic flow is a great technical difficulty to be solved at present.
In the embodiment of the application, the traffic data single parameter is popularized to the unified depiction of multi-parameter distribution characteristics by aiming at single-parameter distribution such as flow, speed and occupancy and a representation method of double-parameter distribution gridding. The idea of discrete gridding is adopted to realize the uniform standard expression and depiction of single-parameter distribution and multi-parameter distribution characteristics, and the influence of factors such as line shape, traffic, environment and the like is overcome.
S104: and obtaining the probability that the single-day traffic flow data belongs to each category according to the distribution characteristics of the single-day traffic flow data based on the trained neural network classifier.
In this step, classification can be achieved using a trained neural network classifier. The input of the neural network classifier is the distribution characteristic of the traffic flow data of a single day, and the output is the probability that the traffic flow data of the single day belongs to each category.
S105: and determining whether the distribution characteristics of the traffic flow data of the single day are effective or not according to the probability that the traffic flow data of the single day belongs to each category. Further, for the single-day traffic flow data that is determined to be valid, the category to which the single-day traffic flow data belongs may also be determined.
Optionally, if the probability that the single-day traffic flow data belongs to each category is lower than the set threshold, the distribution characteristic of the single-day traffic flow data is determined to be invalid, otherwise, the distribution characteristic is determined to be valid. Further, for the single-day traffic flow data judged to be valid, the highest probability can be selected according to the probability that the data belongs to each category, and the category corresponding to the highest probability is taken as the category to which the data belongs.
Optionally, in the embodiment of the application, whether the distribution characteristics of the traffic flow data of a single day are effective or not may be determined based on a traditional traffic data distribution effectiveness determination method, such as a threshold determination method, a three-parameter zero-value consistency determination method, or an effective vehicle length determination method. The threshold value discrimination method judges that the value of the parameter exceeding the threshold value is invalid by setting a reasonable value range of each parameter; the three-parameter zero-value consistency judgment needs to meet the condition that the values of the three parameters are all not 0 when a vehicle passes through or are all 0 in the state of no vehicle passing through; and the effective vehicle length judgment estimates the effective vehicle length of the detected road section based on the flow, the speed and the occupancy rate, and eliminates abnormal data according to the threshold value of the effective vehicle length.
The classification is carried out based on the neural network classifier, on one hand, the effectiveness of the traffic flow data of a single day can be judged according to the probability that the traffic flow data of the single day belongs to each category, on the other hand, the category to which the effective data belongs can be identified, and therefore the neural network classifier is helpful for providing reference for subsequent data repair and other operations or specific management decisions based on similar traffic flow detection equipment.
In the embodiment of the application, after the distribution characteristics of the traffic flow data of a single day are extracted, the probability that the traffic flow data of the single day belongs to each category is obtained according to the distribution characteristics of the traffic flow data of the single day based on the trained neural network classifier, whether the distribution characteristics of the traffic flow data of the single day are effective or not is determined according to the probability that the traffic flow data of the single day belongs to each category, the category to which the traffic flow data of the single day is determined to be effective is determined, and therefore the judgment on the effectiveness of the traffic data distribution is achieved. By adopting the embodiment of the application, the internal basic logic of the traffic data is mined, the machine learning method is adopted for modeling, the abnormal distribution data is identified, the systematic distribution abnormality of the traffic flow detection equipment can be identified efficiently, the abnormal traffic flow detection equipment can be identified more accurately, the support is provided for improving the management efficiency and operation maintenance, and the existing traffic flow data effectiveness judging method is effectively supplemented.
In the traditional traffic flow data quality management, whether the distribution characteristics of traffic flow data are effective or not is judged by a threshold value judgment method, a three-parameter zero-value consistency judgment method or an effective vehicle length judgment method, and the method is widely applied, but the methods mainly aim at the condition that the value of the traffic flow data collected at a certain moment is abnormal. Although the manufacturers of traffic flow detection equipment adopted in different cities are different, parameter distribution abnormity caused by unstable system operation or equipment failure occurs at the same time, and the specific expression is as follows: the value range of a single parameter is normal, but the integral acquisition value of the existing parameter is low, or the distribution of the parameter value combination such as speed-occupancy or flow-occupancy does not accord with the internal logic rule among three parameters of the traffic flow, and the macroscopic basic diagram shows the abnormal distribution form. Due to the introduction of the step of judging the distribution effectiveness, the embodiment of the application can firstly position the abnormal parameters, thereby avoiding the misjudgment of the effectiveness of other parameters in the subsequent combined judgment caused by the abnormal parameters; on the basis of confirming that the parameter distribution is effective, the method can further utilize a threshold value discrimination method, a zero value consistency method or an effective vehicle length discrimination method to carry out the validity discrimination of the traffic flow data, thereby more comprehensively identifying the abnormal traffic flow data.
Referring to fig. 5, a training process of a neural network classifier provided in the embodiment of the present application may include the following steps as shown in the figure:
s501: and acquiring the traffic flow data collected by the traffic flow detection equipment with the identified data distribution effectiveness as a sample for training the neural network classifier.
The specific implementation of this step is basically similar to S101 in fig. 1.
S502: the sample is pre-treated.
The specific implementation manner of this step is basically similar to S102 in fig. 1.
S503: and performing parameter distribution characteristic extraction on the preprocessed sample to obtain the distribution characteristic of the preprocessed sample.
The specific implementation manner of this step is basically similar to S103 in fig. 1.
S504: and determining the distance between the distribution characteristics of each sample in the preprocessed samples based on the relative entropy function, and clustering each sample according to the distance between the distribution characteristics of each sample in the preprocessed samples by adopting a density-based clustering algorithm to obtain the category of each sample in the preprocessed samples.
In this step, for the known normal running traffic flow detection device, after the distribution features (such as single-parameter distribution or double-parameter joint distribution features) of the samples (i.e., single-day traffic flow data) are extracted, a density-based clustering algorithm (such as a DBSCAN algorithm) is used to perform cluster analysis on the samples, so as to obtain class labels, and the class labels are used as training data of the classifier.
Alternatively, the distance between samples may be calculated first. Alternatively, taking sample A and sample B as an example, the distribution P of sample A may be calculated based on a relative entropy (relative entropy) function A Distribution P with sample B B The distance D (A, B) between them, calculatingThe formula is as follows:
wherein KL () represents a relative entropy function operation; p is a radical of A (i) Representing the probability, p, of the ith parameter in sample A B (i) Representing the probability of the ith parameter in sample B. When single parameter distribution distance calculation is performed, N =20; when a two-parameter joint distribution distance calculation is performed, N =400.
After the distances between the samples are calculated, the samples can be clustered by using a clustering algorithm (such as a DBSCAN algorithm) according to the distances between the samples. The DBSCAN is a clustering algorithm based on density, and the cluster to which the sample point belongs is determined by setting a given radius Eps and a threshold Minpts of the number of points in the neighborhood of the radius. The clustering algorithm has the advantages that similar traffic flow detection equipment can be clustered only according to data, and detailed information of the arrangement position of the traffic flow detection equipment and the surrounding traffic environment of the traffic flow detection equipment is not relied on, so that the operability of the method can be improved, and certain universality is achieved.
S505: the neural network classifier is trained using class-labeled samples.
Through the clustering, normal traffic flow detection devices can be grouped, samples in each group are marked (namely, labels of the classes to which the samples belong are marked on each sample, and different classes can mark different traffic flow detection devices), and a neural network classifier can be trained by using the marked samples. The input clustered data training neural network classifier with the class label distribution characteristics can determine the class of the output sample data according to the probability that the sample data belong to each class, and the class is compared with the class marked in the sample to adjust the weight or threshold of the classifier so as to learn the internal logic relationship among the characteristics, thereby achieving the purpose of training the neural network classifier.
In the embodiment of the application, the clustering result is used as the class label instead of the normal/abnormal two-class classification of the equipment, so that on one hand, the validity of the data can be judged according to the probability that the sample data belongs to various types, and on the other hand, the class to which the valid data belongs can be identified, and the method is favorable for providing reference for subsequent data restoration and other operations or specific management decisions based on similar traffic flow detection equipment.
In the aspect of traffic abnormal data discrimination algorithm design, normal traffic data distribution characteristics have similarity but difference due to factors such as road conditions, environmental facilities and the like. How to be compatible with the difference between normal data distributions, designing a proper similarity measurement method, and effectively identifying abnormal data distributions based on extracted distribution characteristics is a problem to be considered. In the training process of the neural network classifier provided by the embodiment of the application, the weights or the threshold values are adjusted in the clustering and classifying algorithm to learn the internal logic among the features, the distribution of similar data is identified, and reference can be provided for subsequent data restoration and other operations or specific management decisions.
Based on the same technical concept, the embodiment of the application also provides a device for judging the distribution effectiveness of the traffic data. The device can realize the traffic data distribution effectiveness judging method provided by the embodiment.
Referring to fig. 6, a schematic structural diagram of a traffic data distribution validity determination apparatus provided in the embodiment of the present application is shown, and as shown in the drawing, the apparatus may include: an acquisition module 601, a preprocessing module 602, a feature extraction module 603, a classification module 604, and an effectiveness judgment module 605.
An obtaining module 601, configured to obtain traffic flow data for a single day collected by a traffic flow detection device for validity of distribution of data to be identified, where the traffic flow data for a single day includes at least one traffic flow parameter collected for at least one lane;
a preprocessing module 602 configured to preprocess the single-day traffic flow data;
the feature extraction module 603 is configured to perform parameter distribution feature extraction on the preprocessed single-day traffic flow data to obtain distribution features of the single-day traffic flow data;
a classification module 604 configured to obtain, based on the trained neural network classifier, a probability that the traffic flow data of the single day belongs to each category according to a distribution characteristic of the traffic flow data of the single day;
a validity judging module 605 configured to determine whether the distribution characteristics of the traffic flow data on a single day are valid according to the probability that the traffic flow data on a single day belongs to each class, and determine the class to which the traffic flow data on a single day belongs for which the traffic flow data on a single day is determined to be valid.
Optionally, the preprocessing module 602 may perform the following operations: and carrying out normalization processing on the single-day traffic flow data, wherein any traffic flow parameter after normalization processing meets the formula (4).
Optionally, the preprocessing module 602 may further perform the following operations before performing the normalization processing on the traffic flow data of a single day: and according to a set aggregation period, aggregating the traffic flow data of the single day into the traffic flow data of the single day of the aggregation period. Optionally, if the single-day traffic flow data includes at least one of a flow rate, a speed, and an occupancy, the aggregated flow rate satisfies the above formula (1), the aggregated speed satisfies the above formula (2), and the aggregated occupancy satisfies the above formula (3).
Optionally, the preprocessing module 602 may further perform the following operations before performing the normalization processing on the traffic flow data of a single day: and eliminating abnormal values in the traffic flow data of the single day by adopting a box type graph method.
Optionally, the single-day traffic flow data includes a single traffic flow parameter. The feature extraction module 603 may perform the following operations: uniformly dividing the value intervals of the single-day traffic flow parameters into sub-intervals with set number; and counting the frequency of the single-day traffic flow parameters falling into each subinterval.
Optionally, the single-day traffic flow data includes a first traffic flow parameter and a second traffic flow parameter, and the first traffic flow parameter and the second traffic flow parameter are different types of traffic flow parameters. The feature extraction module 603 may perform the following operations: dividing the grids according to the set grid size, the value interval of the first traffic flow parameter and the value interval of the second traffic flow parameter, wherein the horizontal dimension of each divided grid is the first traffic flow parameter, and the vertical dimension is the second traffic flow parameter; and counting the frequency of the single-day traffic flow parameters falling into each grid.
It should be noted that the apparatus for determining a saturation of a business turn provided in the embodiment of the present application can implement all the method steps implemented by the method embodiment and achieve the same technical effects, and detailed descriptions of the same parts and beneficial effects as those of the method embodiment in this embodiment are omitted here.
Based on the same technical concept, the embodiment of the present application further provides a communication device, and the communication device can implement the method flows provided by the above embodiments of the present application.
Fig. 7 illustrates a schematic structural diagram of a communication device provided in an embodiment of the present application. As shown, the apparatus may comprise: a processor 701, a memory 702, and a bus interface 703.
The processor 701 is responsible for managing the bus architecture and general processing, and the memory 702 may store data used by the processor 701 in performing operations.
The bus architecture may include any number of interconnected buses and bridges, with one or more processors, represented by processor 701, and various circuits, represented by memory 702, being linked together. The bus architecture may also link together various other circuits such as peripherals, voltage regulators, power management circuits, and the like, which are well known in the art, and therefore, will not be described any further herein. The bus interface provides an interface. The processor 701 is responsible for managing the bus architecture and general processing, and the memory 702 may store data used by the processor 701 in performing operations.
The processes disclosed in the embodiments of the present application may be applied to the processor 701, or implemented by the processor 701. In implementation, the steps of the signal processing flow may be implemented by integrated logic circuits of hardware or instructions in the form of software in the processor 701. The processor 701 may be a general purpose processor, a digital signal processor, an application specific integrated circuit, a field programmable gate array or other programmable logic device, discrete gate or transistor logic, discrete hardware components, or any combination thereof that may implement or perform the methods, steps, and logic blocks of the present application in embodiments thereof. A general purpose processor may be a microprocessor or any conventional processor or the like. The steps of a method applied in connection with the embodiments of the present application may be directly implemented by a hardware processor, or may be implemented by a combination of hardware and software modules in the processor. The software module may be located in ram, flash memory, rom, prom, or eprom, registers, etc. storage media as is well known in the art. The storage medium is located in the memory 702, and the processor 701 reads the information in the memory 702 and completes the steps of the information processing flow in combination with the hardware thereof.
Specifically, the processor 701 is configured to read a computer instruction in the memory 702 and execute the method for determining the validity of the traffic flow data distribution in the embodiment of the present application.
It should be noted that, the communication apparatus provided in the embodiment of the present application can implement all the method steps implemented by the method embodiment and achieve the same technical effect, and detailed descriptions of the same parts and beneficial effects as the method embodiment in this embodiment are omitted here.
Based on the same technical concept, the embodiment of the application also provides a computer readable storage medium. The computer-readable storage medium stores computer-executable instructions for causing a computer to perform the process illustrated in any one of fig. 1 and 5.
Based on the same technical concept, the embodiment of the present application further provides a computer program product, which, when called by a computer, causes the computer to execute the flow shown in any one of fig. 1 and 5.
The present application is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the application. It will be understood that each flow and/or block of the flow diagrams and/or block diagrams, and combinations of flows and/or blocks in the flow diagrams and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
While the preferred embodiments of the present application have been described, additional variations and modifications in those embodiments may occur to those skilled in the art once they learn of the basic inventive concepts. Therefore, it is intended that the appended claims be interpreted as including preferred embodiments and all alterations and modifications as fall within the scope of the application.
It will be apparent to those skilled in the art that various changes and modifications may be made in the present application without departing from the spirit and scope of the application. Thus, if such modifications and variations of the present application fall within the scope of the claims of the present application and their equivalents, the present application is intended to include such modifications and variations as well.
Claims (9)
1. A traffic data distribution validity judging method is characterized by comprising the following steps:
acquiring single-day traffic flow data collected by traffic flow detection equipment for distribution effectiveness of data to be identified, wherein the single-day traffic flow data comprises at least one traffic flow parameter collected aiming at least one lane;
preprocessing the single-day traffic flow data;
performing parameter distribution characteristic extraction on the preprocessed single-day traffic flow data to obtain the distribution characteristics of the single-day traffic flow data;
based on the trained neural network classifier, obtaining the probability that the traffic flow data of the single day belongs to each category according to the distribution characteristics of the traffic flow data of the single day;
determining whether the distribution characteristics of the single-day traffic flow data are effective or not according to the probability of the single-day traffic flow data belonging to each category, and determining the category of the single-day traffic flow data which is determined to be effective;
wherein the training process of the neural network classifier comprises the following steps:
acquiring single-day traffic flow data acquired by traffic flow detection equipment with data distribution effectiveness identified as a sample for training, wherein the single-day traffic flow data comprises at least one traffic flow parameter acquired aiming at least one lane;
pre-treating the sample;
performing parameter distribution characteristic extraction on the preprocessed sample to obtain the distribution characteristic of the preprocessed sample;
determining the distance between the distribution characteristics of each sample in the preprocessed sample based on a relative entropy function, and clustering the preprocessed sample according to the distance between the distribution characteristics of each sample in the preprocessed sample by adopting a density-based clustering algorithm to obtain the category of each sample in the preprocessed sample;
training the neural network classifier using class-labeled samples.
2. The method of claim 1, wherein the pre-processing the single-day traffic flow data comprises:
carrying out normalization processing on the single-day traffic flow data, wherein any traffic flow parameter after normalization processing meets the following formula:
wherein X "represents a traffic flow parameter after normalization, X 'represents a traffic flow parameter before normalization, and X' max Represents the maximum value X 'of the traffic flow parameter in the single-day traffic flow data' min And the minimum value of the traffic flow parameter in the single-day traffic flow data is represented.
3. The method of claim 2, wherein prior to normalizing the single-day traffic flow data, the method further comprises:
and according to a set aggregation period, aggregating the traffic flow data of the single day into the traffic flow data of the single day of the aggregation period.
4. The method of claim 3, wherein if the traffic flow data for a single day includes at least one of traffic, speed, occupancy, then:
the flow after polymerization satisfies the following formula:
the velocity after polymerization satisfies the following formula:
the occupancy after polymerization satisfies the following formula:
wherein, q' k Denotes the flow rate of the k-th aggregation period, q i Representing the flow of the ith acquisition cycle; v' k Denotes the velocity, v, of the k-th polymerization cycle i Representing the speed of the ith acquisition cycle; o 'to' k Denotes the occupancy of the k-th polymerization cycle, o i Representing the occupancy of the ith acquisition cycle; t is a unit of 0 Denotes the aggregation period and T denotes the acquisition period.
5. The method of claim 2, wherein prior to normalizing the single-day traffic flow data, the method further comprises:
and eliminating abnormal values in the traffic flow data of the single day by adopting a box type graph method.
6. The method of claim 1, wherein the single day traffic flow data comprises a single traffic flow parameter;
the parameter distribution characteristic extraction of the preprocessed traffic flow data on a single day comprises the following steps:
uniformly dividing the value intervals of the single-day traffic flow parameters into sub-intervals with set number;
and counting the frequency of the single-day traffic flow parameters falling into each subinterval.
7. The method of claim 1, wherein the single-day traffic flow data includes a first traffic flow parameter and a second traffic flow parameter, the first traffic flow parameter and the second traffic flow parameter being different types of traffic flow parameters;
the parameter distribution characteristic extraction of the preprocessed traffic flow data on a single day comprises the following steps:
dividing the grids according to the set grid size, the value interval of the first traffic flow parameter and the value interval of the second traffic flow parameter, wherein the horizontal dimension of each divided grid is the first traffic flow parameter, and the vertical dimension is the second traffic flow parameter;
and counting the frequency of the single-day traffic flow parameters falling into each grid.
8. A traffic data distribution validity determination device is characterized by comprising:
the acquisition module is configured to acquire single-day traffic flow data acquired by traffic flow detection equipment for validity of distribution of data to be identified, wherein the single-day traffic flow data comprises at least one traffic flow parameter acquired aiming at least one lane;
a preprocessing module configured to preprocess the single-day traffic flow data;
the characteristic extraction module is configured to perform parameter distribution characteristic extraction on the preprocessed single-day traffic flow data to obtain the distribution characteristics of the single-day traffic flow data;
the classification module is configured to obtain the probability that the single-day traffic flow data belongs to each category according to the distribution characteristics of the single-day traffic flow data based on the trained neural network classifier;
the validity judging module is configured to determine whether the distribution characteristics of the single-day traffic flow data are valid according to the probability that the single-day traffic flow data belong to each class, and determine the class to which the single-day traffic flow data are determined to be valid;
wherein the training process of the neural network classifier comprises the following steps:
acquiring single-day traffic flow data acquired by traffic flow detection equipment with data distribution effectiveness identified as a sample for training, wherein the single-day traffic flow data comprises at least one traffic flow parameter acquired aiming at least one lane;
pre-treating the sample;
performing parameter distribution characteristic extraction on the preprocessed sample to obtain the distribution characteristic of the preprocessed sample;
determining the distance between the distribution characteristics of each sample in the preprocessed samples based on a relative entropy function, and clustering the preprocessed samples according to the distance between the distribution characteristics of each sample in the preprocessed samples by adopting a density-based clustering algorithm to obtain the category of each sample in the preprocessed samples;
training the neural network classifier using class-labeled samples.
9. A communication device comprising a memory and a processor; the memory storing computer instructions; the processor, reading the computer instructions, performing the method of any one of claims 1-7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210063786.6A CN114495498B (en) | 2022-01-20 | 2022-01-20 | Traffic data distribution effectiveness judging method and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210063786.6A CN114495498B (en) | 2022-01-20 | 2022-01-20 | Traffic data distribution effectiveness judging method and device |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN114495498A CN114495498A (en) | 2022-05-13 |
| CN114495498B true CN114495498B (en) | 2023-01-10 |
Family
ID=81472195
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210063786.6A Active CN114495498B (en) | 2022-01-20 | 2022-01-20 | Traffic data distribution effectiveness judging method and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114495498B (en) |
Families Citing this family (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115171372B (en) * | 2022-06-20 | 2023-10-24 | 青岛海信网络科技股份有限公司 | Traffic abnormality detection method, equipment and device |
| CN115188186B (en) * | 2022-06-28 | 2024-02-20 | 公安部交通管理科学研究所 | Method for monitoring traffic flow in area |
Citations (13)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2018122585A1 (en) * | 2016-12-30 | 2018-07-05 | 同济大学 | Method for urban road traffic incident detecting based on floating-car data |
| CN108847022A (en) * | 2018-06-08 | 2018-11-20 | 浙江银江智慧交通集团有限公司 | A kind of rejecting outliers method of microwave traffic data collection equipment |
| CN109255956A (en) * | 2018-11-12 | 2019-01-22 | 长安大学 | A kind of charge station's magnitude of traffic flow method for detecting abnormality |
| CN109300315A (en) * | 2018-10-12 | 2019-02-01 | 山东交通学院 | Geomagnetic data anomalous discrimination method and system based on vehicle detection probability |
| CN109360415A (en) * | 2018-09-30 | 2019-02-19 | 北京工业大学 | A kind of road traffic flow disorder data recognition method |
| CN110751853A (en) * | 2019-10-25 | 2020-02-04 | 百度在线网络技术（北京）有限公司 | Parking space data validity identification method and device |
| CN110852516A (en) * | 2019-11-15 | 2020-02-28 | 上海电科智能系统股份有限公司 | Data quality judging method based on big data information entropy traffic flow detection equipment |
| CN111310786A (en) * | 2020-01-15 | 2020-06-19 | 青岛海信网络科技股份有限公司 | Traffic detector abnormity diagnosis method and device based on random forest classifier |
| JP2020154958A (en) * | 2019-03-22 | 2020-09-24 | 沖電気工業株式会社 | Abnormal traffic flow detection device, abnormal traffic flow detection method, and abnormal traffic flow detection program |
| CN112434075A (en) * | 2020-10-23 | 2021-03-02 | 北京千方科技股份有限公司 | ETC portal frame based traffic anomaly detection method and device, storage medium and terminal |
| KR20210061630A (en) * | 2019-11-20 | 2021-05-28 | 한국전자통신연구원 | Centralized scheduling apparatus and method considering non-uniform traffic |
| CN113242207A (en) * | 2021-04-02 | 2021-08-10 | 河海大学 | Iterative clustering network flow abnormity detection method |
| CN113362597A (en) * | 2021-06-03 | 2021-09-07 | 济南大学 | Traffic sequence data anomaly detection method and system based on non-parametric modeling |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20130066875A1 (en) * | 2011-09-12 | 2013-03-14 | Jacques Combet | Method for Segmenting Users of Mobile Internet |
-
2022
- 2022-01-20 CN CN202210063786.6A patent/CN114495498B/en active Active
Patent Citations (14)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2018122585A1 (en) * | 2016-12-30 | 2018-07-05 | 同济大学 | Method for urban road traffic incident detecting based on floating-car data |
| CN109923595A (en) * | 2016-12-30 | 2019-06-21 | 同济大学 | A kind of urban highway traffic method for detecting abnormality based on floating car data |
| CN108847022A (en) * | 2018-06-08 | 2018-11-20 | 浙江银江智慧交通集团有限公司 | A kind of rejecting outliers method of microwave traffic data collection equipment |
| CN109360415A (en) * | 2018-09-30 | 2019-02-19 | 北京工业大学 | A kind of road traffic flow disorder data recognition method |
| CN109300315A (en) * | 2018-10-12 | 2019-02-01 | 山东交通学院 | Geomagnetic data anomalous discrimination method and system based on vehicle detection probability |
| CN109255956A (en) * | 2018-11-12 | 2019-01-22 | 长安大学 | A kind of charge station's magnitude of traffic flow method for detecting abnormality |
| JP2020154958A (en) * | 2019-03-22 | 2020-09-24 | 沖電気工業株式会社 | Abnormal traffic flow detection device, abnormal traffic flow detection method, and abnormal traffic flow detection program |
| CN110751853A (en) * | 2019-10-25 | 2020-02-04 | 百度在线网络技术（北京）有限公司 | Parking space data validity identification method and device |
| CN110852516A (en) * | 2019-11-15 | 2020-02-28 | 上海电科智能系统股份有限公司 | Data quality judging method based on big data information entropy traffic flow detection equipment |
| KR20210061630A (en) * | 2019-11-20 | 2021-05-28 | 한국전자통신연구원 | Centralized scheduling apparatus and method considering non-uniform traffic |
| CN111310786A (en) * | 2020-01-15 | 2020-06-19 | 青岛海信网络科技股份有限公司 | Traffic detector abnormity diagnosis method and device based on random forest classifier |
| CN112434075A (en) * | 2020-10-23 | 2021-03-02 | 北京千方科技股份有限公司 | ETC portal frame based traffic anomaly detection method and device, storage medium and terminal |
| CN113242207A (en) * | 2021-04-02 | 2021-08-10 | 河海大学 | Iterative clustering network flow abnormity detection method |
| CN113362597A (en) * | 2021-06-03 | 2021-09-07 | 济南大学 | Traffic sequence data anomaly detection method and system based on non-parametric modeling |
Non-Patent Citations (2)
| Title |
|---|
| 典型道路交通数据有效性评价方;赵盼明;《中国优秀硕士学位论文全文数据库 工程科技Ⅱ辑》;20200615(第6期);全文 * |
| 干线公路交通流数据有效性检验规则;朱雷雷;《东南大学学报( 自然科学版)》;20110131;第41卷(第1期);全文 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN114495498A (en) | 2022-05-13 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN114495498B (en) | Traffic data distribution effectiveness judging method and device | |
| CN108986465B (en) | Method, system and terminal equipment for detecting traffic flow | |
| WO2021043140A1 (en) | Method, apparatus and system for determining label | |
| CN104050361A (en) | Intelligent analysis early warning method for dangerousness tendency of prison persons serving sentences | |
| CN114519524A (en) | Enterprise risk early warning method and device based on knowledge graph and storage medium | |
| CN113269042B (en) | Intelligent traffic management method and system based on driving vehicle violation identification | |
| CN109995611B (en) | Traffic classification model establishing and traffic classification method, device, equipment and server | |
| CN111796957A (en) | Transaction abnormal root cause analysis method and system based on application log | |
| CN110704616B (en) | Equipment alarm work order identification method and device | |
| CN113592019A (en) | Fault detection method, device, equipment and medium based on multi-model fusion | |
| CN109726737B (en) | Track-based abnormal behavior detection method and device | |
| CN111949429A (en) | Server fault monitoring method and system based on density clustering algorithm | |
| CN111176953A (en) | Anomaly detection and model training method thereof, computer equipment and storage medium | |
| CN109598931A (en) | Group based on traffic safety risk divides and difference analysis method and system | |
| CN113516105B (en) | Lane detection method and device and computer readable storage medium | |
| CN115996249A (en) | Data transmission method and device based on grading | |
| CN111222968A (en) | Enterprise tax risk management and control method and system | |
| CN110474862B (en) | Network traffic anomaly detection method and device | |
| CN105447863A (en) | Residue detection method based on improved VIBE | |
| KR102433598B1 (en) | A System and Method for Deriving Data Boundary | |
| CN104537392A (en) | Object detection method based on distinguishing semantic component learning | |
| CN110807174B (en) | Effluent analysis and abnormity identification method for sewage plant group based on statistical distribution | |
| CN113824580A (en) | Network index early warning method and system | |
| CN105227410A (en) | Based on the method and system that the server load of adaptive neural network detects | |
| CN115187259A (en) | Block chain abnormal transaction identification method and system based on unsupervised machine learning |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |