CN111785010A

CN111785010A - Method and device for detecting traffic efficiency information

Info

Publication number: CN111785010A
Application number: CN201910266365.1A
Authority: CN
Inventors: 邹莉; 徐琪琪; 孙伟力
Original assignee: Beijing Didi Infinity Technology and Development Co Ltd
Current assignee: Beijing Didi Infinity Technology and Development Co Ltd
Priority date: 2019-04-03
Filing date: 2019-04-03
Publication date: 2020-10-16

Abstract

The application provides a method and a device for detecting traffic efficiency information, wherein the method comprises the following steps: acquiring target vehicle track data of a vehicle driving through a target intersection in a time period to be detected and intersection information of the target intersection; generating a characteristic value of the target intersection under the influence characteristic of the traffic efficiency according to the target vehicle track data and the intersection information; and inputting the characteristic value into a pre-trained traffic efficiency information detection model, and acquiring traffic efficiency information corresponding to the target intersection. Compared with the mode of determining the traffic intersection traffic efficiency information through a manual field judgment method in the prior art, the method is shorter in time consumption, higher in efficiency and higher in accuracy.

Description

Method and device for detecting traffic efficiency information

Technical Field

The application relates to the technical field of big data, in particular to a method and a device for detecting traffic efficiency information.

Background

With the increasing number of urban vehicles, the problem of low efficiency of traffic becomes one of the hot spots concerned by people at present. The discovery of the cause of traffic inefficiency is a prerequisite to alleviating the problem of traffic inefficiency. The location where traffic congestion occurs is typically at a road intersection. When the green light of the road intersection is turned on, vehicles in the corresponding direction should smoothly drive away from the stop line and pass through the intersection. If the vehicle is behind the stop line but the inside of the intersection still runs slowly, the passing efficiency of the intersection is affected, and therefore the traffic inefficient state is caused.

At present, the traffic intersection traffic efficiency information is determined by adopting a manual field judgment method, the method is long in time consumption and low in efficiency, and due to the fact that judgment is carried out manually, judgment can be carried out only on the basis of the intersection traffic condition at the current moment, and the judgment accuracy is also low.

Disclosure of Invention

In view of the above, an object of the embodiments of the present application is to provide a method and an apparatus for detecting traffic efficiency information, which can detect traffic efficiency information of an intersection based on vehicle trajectory data and traffic conditions of a downstream intersection, and have short time consumption, high efficiency and high accuracy in judgment compared with manual judgment.

In a first aspect, an embodiment of the present application provides a method for detecting traffic efficiency information, including:

acquiring target vehicle track data of a vehicle driving through a target intersection in a time period to be detected and intersection information of the target intersection;

generating a characteristic value of the target intersection under the influence characteristic of the traffic efficiency according to the target vehicle track data and the intersection information;

and inputting the characteristic value into a pre-trained traffic efficiency information detection model, and acquiring traffic efficiency information corresponding to the target intersection.

In an alternative embodiment, the traffic efficiency information includes information indicating a traffic inefficiency and a reason for the traffic inefficiency; or, information indicating the passage efficiency and the passage efficiency reason.

In an alternative embodiment, the traffic efficiency affecting feature comprises: one or more of road condition characteristics, road characteristics, and time characteristics.

In an optional implementation manner, for a case that the traffic efficiency influence characteristics include road condition characteristics and road characteristics, the method further includes:

acquiring historical vehicle track data of the target intersection driven by a historical detection time period before the time period to be detected;

generating a characteristic value of the target intersection under the influence characteristic of traffic efficiency according to the target vehicle track data and the intersection information, wherein the characteristic value comprises the following steps:

generating a first characteristic value of the target intersection under the road condition characteristic according to the target vehicle track data, generating a second characteristic value of the target intersection under the road condition characteristic according to the historical vehicle track data, and generating a third characteristic value of the target intersection under the road characteristic according to the intersection information;

inputting the characteristic value into a pre-trained traffic efficiency information detection model, and acquiring traffic efficiency information corresponding to the target intersection, wherein the method comprises the following steps:

and inputting the first characteristic value, the second characteristic value and the third characteristic value into the pre-trained traffic efficiency information detection model to obtain traffic efficiency information corresponding to the target intersection.

In an alternative embodiment, for the case that the traffic efficiency influence characteristics comprise road condition characteristics, road characteristics and time characteristics,

the method further comprises the following steps: acquiring a fourth characteristic value of the target intersection under the time characteristic;

the inputting the first characteristic value, the second characteristic value and the third characteristic value into a pre-trained traffic efficiency information detection model to obtain traffic efficiency information corresponding to the target intersection includes:

and inputting the first characteristic value, the second characteristic value, the third characteristic value and the fourth characteristic value into a pre-trained traffic efficiency information detection model, and acquiring traffic efficiency information corresponding to the target intersection.

In an alternative embodiment, the road condition characteristics include: one or more of the speed of the vehicle flow in the intersection, the speed of the vehicle flow on the downstream road section of the intersection and the parking coverage rate on the downstream road section of the intersection.

In an alternative embodiment, the road feature comprises: the bus stop comprises one or more of turning radius, intersection slope, intersection sight distance, the number of intersection stop lines, intersection exit lane width, the distance between a road entrance nearest to an intersection exit and an intersection, the distance between a bus station nearest to the intersection exit and the intersection, and the number of buses stopped at the bus station nearest to the intersection exit.

In an alternative embodiment, the temporal characteristics include: the time period to be detected is one or more of a date corresponding to the time period to be detected, whether the time period to be detected is a traffic peak, the number of weeks corresponding to the time period to be detected, whether the time period to be detected is a holiday and the number of months corresponding to the time period to be detected.

In an alternative embodiment, for the case where the road condition characteristics include speed of vehicle flow within the intersection:

acquiring the speed of the vehicle flow in the intersection by adopting the following modes:

calculating the speed of each target vehicle passing through the target intersection, and the entrance time and the exit time of each target vehicle passing through the target intersection according to the target vehicle track data of each vehicle;

and fitting the speed, the driving-in time and the driving-out time of each vehicle passing through the target intersection to obtain the speed of the vehicle flow in the intersection.

In an alternative embodiment, the target vehicle trajectory data comprises: the position information of a plurality of positioning points and the positioning time corresponding to each positioning point;

calculating the speed of each vehicle passing through the target intersection according to the target vehicle track data of each vehicle, and the method comprises the following steps:

for each vehicle, performing:

calculating the distance between every two adjacent positioning points according to the position information of every two adjacent positioning points;

calculating the time difference of every two positioning points according to the positioning time of every two adjacent positioning points;

calculating the temporary speed corresponding to each two adjacent positioning points according to the distance and the time difference;

and calculating the speed of the vehicle passing through the target intersection based on each temporary speed.

calculating the entering time and the leaving time of each target vehicle passing through the target intersection according to the target vehicle track data of each vehicle, and the method comprises the following steps:

determining an entering positioning point and an exiting positioning point from each positioning point according to the position information of each positioning point and the area range of the target intersection;

and determining the entrance time according to the positioning time corresponding to the entrance positioning point, and determining the exit time according to the positioning time of the exit positioning point.

In an optional implementation manner, for a case that the road condition characteristic includes a traffic flow speed of a downstream road section of the intersection, the traffic flow speed of the downstream road section of the intersection in the time period to be detected is obtained in the following manner:

acquiring downstream vehicle track data of vehicles driving through a downstream road section of a target intersection in a time period to be detected;

and generating the traffic flow speed of the downstream road section of the intersection according to the downstream vehicle track data.

In an alternative embodiment, the traffic efficiency information detection model is obtained by training in the following way:

acquiring sample vehicle track data of a vehicle driving through at least one sample intersection in a sample time period, sample intersection information of each sample intersection and sample traffic efficiency information corresponding to each sample intersection;

generating a sample characteristic value of each sample intersection under the traffic efficiency influence characteristic according to the sample vehicle track data and the sample intersection information;

and constructing a decision tree model by taking the sample characteristic value of each sample intersection under the traffic efficiency influence characteristic as an input characteristic value of the decision tree model and taking the sample traffic efficiency information corresponding to each sample intersection as an output characteristic value of the decision tree model, and taking the constructed decision tree model as the traffic efficiency information detection model.

acquiring sample vehicle track data of a vehicle driving through at least one sample intersection in a sample time period, sample historical vehicle track data of the vehicle driving through the sample intersection in a sample historical detection time period corresponding to the sample time period, sample intersection information of each sample intersection and sample traffic efficiency information corresponding to each sample intersection;

generating a first sample characteristic value of the sample intersection under the road condition characteristic according to the sample vehicle track data, generating a second sample characteristic value of the sample intersection under the road condition characteristic according to the sample historical vehicle track data, and generating a third sample characteristic value of the sample intersection under the road condition characteristic according to the intersection information;

inputting the first sample characteristic value, the second sample characteristic value and the third sample characteristic value into a basic prediction model, and obtaining a sample passing low-efficiency detection result corresponding to each sample intersection;

and training the basic prediction model according to the sample traffic inefficiency detection result and the sample traffic efficiency information, and taking the trained basic prediction model as the traffic efficiency information detection model.

In an alternative embodiment, the base prediction model comprises: a first neural network, a second neural network, and a third neural network;

inputting the first sample characteristic value, the second sample characteristic value and the third sample characteristic value into a basic prediction model, and obtaining a sample passage inefficiency detection result corresponding to each sample intersection includes:

for each of the sample intersections, performing:

fusing a first feature vector formed by the first sample feature value and a second feature vector formed by the second sample feature value to generate a first intermediate sample feature vector;

inputting the first intermediate sample feature vector into the first neural network to obtain a second intermediate sample feature vector;

inputting a third sample feature vector formed by the third sample feature value into the second neural network to obtain a third intermediate sample feature vector;

splicing the second intermediate sample feature vector and the third intermediate sample feature vector to form a spliced vector;

and inputting the splicing vector into a third neural network to obtain the detection result of the passage inefficiency of the sample.

In an alternative embodiment, the method further comprises: acquiring a fourth sample characteristic value under the time characteristic;

inputting the first sample characteristic value, the second sample characteristic value and the third sample characteristic value into a basic prediction model, and obtaining a sample passage inefficiency detection result corresponding to each sample intersection, including:

inputting the first sample characteristic value, the second sample characteristic value, the third sample characteristic value and the fourth sample characteristic value into a basic prediction model, and obtaining a sample passing low-efficiency detection result corresponding to each sample intersection.

inputting the first sample characteristic value, the second sample characteristic value, the third sample characteristic value and the fourth sample characteristic value into a basic prediction model, and obtaining a sample passage inefficiency detection result corresponding to each sample intersection, including:

for each of the sample intersections, performing:

splicing a third sample characteristic vector formed by the third sample characteristic value and a fourth sample characteristic vector formed by the fourth sample characteristic value, and inputting the spliced third sample characteristic vector and the fourth sample characteristic vector into the second neural network to obtain a third intermediate sample characteristic vector;

In an optional implementation manner, before generating a feature value of the target intersection under a traffic efficiency influence characteristic according to the target vehicle trajectory data and the intersection information, the method further includes:

determining whether the target intersection is in a low-efficiency passing state or not according to the target vehicle track data and the reference speed of the target intersection;

and after the target intersection is detected to be in a low-efficiency passing state, generating a characteristic value of the target intersection under the passing efficiency influence characteristic according to the target vehicle track data and the intersection information.

In an alternative embodiment, the determining whether the target intersection is in a traffic inefficiency state according to the target vehicle trajectory data and the reference speed of the target intersection includes:

acquiring the speed of the target intersection in the intersection of the time period to be detected according to the target vehicle track data;

detecting whether the speed difference between the reference speed and the speed in the intersection is greater than a preset first speed difference threshold value or not, and detecting whether the speed in the intersection is less than the preset first speed threshold value or not;

judging whether the reference speed is smaller than a preset second speed threshold value or not;

and if the speed difference between the reference speed and the speed in the intersection is detected to be larger than a preset first speed difference threshold value and the speed in the intersection is detected to be smaller than a preset second speed threshold value, or the reference speed is smaller than the preset first speed threshold value, determining that the target intersection is in a passing inefficient state.

determining whether the target intersection is in a high-efficiency passing state or not according to the target vehicle track data and the reference speed of the target intersection;

and after detecting that the target intersection is in a high-efficiency passing state, generating a characteristic value of the target intersection under the influence characteristic of the passing efficiency according to the target vehicle track data and the intersection information.

In an alternative embodiment, determining whether the target intersection is in a traffic-efficient state based on the target vehicle trajectory data and a reference speed of the target intersection comprises:

detecting whether the speed difference between the reference speed and the speed in the intersection is smaller than a preset second speed difference threshold value or not, and detecting whether the speed in the intersection is larger than a preset third speed threshold value or not;

and if the speed difference between the reference speed and the speed in the intersection is detected to be smaller than a preset second speed difference threshold value and the speed in the intersection is detected to be larger than a preset third speed threshold value, determining that the target intersection is in a high-efficiency passing state.

In an alternative embodiment, the method further comprises: determining the intersection internal area of the target intersection according to target vehicle track data of a plurality of vehicles driving through the target intersection;

the acquiring of the target vehicle track data of the vehicle which passes through the target intersection in the time period to be detected comprises the following steps:

and acquiring target vehicle track data of vehicles driving through the intersection inner area of the target intersection in the time period to be detected.

In an alternative embodiment, the reference speed is determined by:

acquiring reference vehicle track data of vehicles driving through the target intersection within a reference time period;

and determining the reference speed corresponding to the target intersection according to the reference vehicle track data.

In an optional implementation, the acquiring target vehicle trajectory data of a vehicle that passes through a target intersection in a time period to be detected includes:

acquiring target vehicle track data of vehicles driving in the road direction in a time period to be detected aiming at each road direction of the target intersection;

generating characteristic values of all road directions under the traffic efficiency influence characteristic according to target vehicle track data of vehicles driving through all road directions in a time period to be detected;

and inputting the characteristic value of each road direction under the traffic efficiency influence characteristic into a traffic efficiency information detection model trained in advance, and acquiring traffic efficiency information corresponding to each road direction.

In a second aspect, an embodiment of the present application further provides an apparatus for detecting traffic efficiency information, including:

the data acquisition module is used for acquiring target vehicle track data of vehicles which drive through a target intersection in a time period to be detected and intersection information of the target intersection;

the characteristic generating module is used for generating a characteristic value of the target intersection under the influence characteristic of the traffic efficiency according to the target vehicle track data and the intersection information;

and the result acquisition module is used for inputting the characteristic value into a pre-trained traffic efficiency information detection model and acquiring traffic efficiency information corresponding to the target intersection.

In an alternative embodiment, for the case that the traffic efficiency influence characteristics comprise road condition characteristics and road characteristics,

the data acquisition module is further configured to: acquiring historical vehicle track data of the target intersection driven by a historical detection time period before the time period to be detected;

the feature generation module is specifically configured to: generating a first characteristic value of the target intersection under the road condition characteristic according to the target vehicle track data, generating a second characteristic value of the target intersection under the road condition characteristic according to the historical vehicle track data, and generating a third characteristic value of the target intersection under the road characteristic according to the intersection information;

the result obtaining module is specifically configured to: and inputting the first characteristic value, the second characteristic value and the third characteristic value into the pre-trained traffic efficiency information detection model to obtain traffic efficiency information corresponding to the target intersection.

the feature generation module is further configured to: acquiring a fourth characteristic value of the target intersection under the time characteristic;

the result obtaining module is specifically configured to: and inputting the first characteristic value, the second characteristic value, the third characteristic value and the fourth characteristic value into a pre-trained traffic efficiency information detection model, and acquiring traffic efficiency information corresponding to the target intersection.

the characteristic generating module is used for acquiring the speed of the vehicle flow in the intersection by adopting the following modes:

the characteristic generating module is used for calculating the speed of each vehicle passing through the target intersection according to the target vehicle track data of each vehicle by adopting the following modes:

for each vehicle, performing:

the characteristic generating module is used for calculating the entering time and the leaving time of each target vehicle passing through the target intersection according to the target vehicle track data of each vehicle by adopting the following modes:

In an alternative embodiment, for the case that the road condition characteristics include the traffic flow speed of the road section downstream of the intersection,

the characteristic generating module is used for acquiring the traffic flow speed of the downstream road section of the intersection in the time period to be detected by adopting the following modes:

In an alternative embodiment, the method further comprises: the first model training module is used for training in the following way to obtain the traffic efficiency information detection model:

In an alternative embodiment, the method further comprises: the second model training module is used for training to obtain the traffic efficiency information detection model by adopting the following modes:

the second model training module is configured to input the first sample feature value, the second sample feature value, and the third sample feature value into a basic prediction model in the following manner, and obtain a sample passage inefficiency detection result corresponding to each sample intersection:

for each of the sample intersections, performing:

In an optional implementation, the second model training module is further configured to: acquiring a fourth sample characteristic value under the time characteristic;

the second model training module is configured to input the first sample feature value, the second sample feature value, the third sample feature value, and the fourth sample feature value into a basic prediction model in the following manner, and obtain a sample passage inefficiency detection result corresponding to each sample intersection:

for each of the sample intersections, performing:

In an alternative embodiment, the method further comprises: a traffic efficiency determination module, configured to, before the feature generation module generates a feature value of the target intersection under a traffic efficiency influence feature according to the target vehicle trajectory data and the intersection information:

the feature generation module is specifically configured to: and after the target intersection is detected to be in a low-efficiency passing state, generating a characteristic value of the target intersection under the passing efficiency influence characteristic according to the target vehicle track data and the intersection information.

In an optional embodiment, the traffic efficiency determination module is specifically configured to determine whether the target intersection is in a traffic inefficiency state according to the target vehicle trajectory data and the reference speed of the target intersection in the following manner:

In an alternative embodiment, the apparatus further comprises: an interior region determining module, configured to determine an intersection interior region of the target intersection according to target vehicle trajectory data of a plurality of vehicles driving through the target intersection;

the data acquisition module is specifically configured to: and acquiring target vehicle track data of vehicles driving through the intersection inner area of the target intersection in the time period to be detected.

In an alternative embodiment, the traffic efficiency determination module is specifically configured to determine the reference speed by:

In an alternative embodiment, the method further comprises: a second traffic efficiency determination module, configured to, before the feature generation module generates the feature value of the target intersection under the traffic efficiency influence feature according to the target vehicle trajectory data and the intersection information:

the feature generation module is specifically configured to: and after detecting that the target intersection is in a high-efficiency passing state, generating a characteristic value of the target intersection under the influence characteristic of the passing efficiency according to the target vehicle track data and the intersection information.

In an optional embodiment, the second traffic efficiency determination module is configured to determine whether the target intersection is in a traffic efficiency state according to the target vehicle trajectory data and the reference speed of the target intersection by:

In an optional implementation manner, the data obtaining module is specifically configured to: acquiring target vehicle track data of vehicles driving in the road direction in a time period to be detected aiming at each road direction of the target intersection;

the feature generation module is specifically configured to: generating characteristic values of all road directions under the traffic efficiency influence characteristic according to target vehicle track data of vehicles driving through all road directions in a time period to be detected;

the result obtaining module is specifically configured to: and inputting the characteristic value of each road direction under the traffic efficiency influence characteristic into a traffic efficiency information detection model trained in advance, and acquiring traffic efficiency information corresponding to each road direction.

In a third aspect, an embodiment of the present application further provides an electronic device, including: a processor, a memory and a bus, the memory storing machine-readable instructions executable by the processor, the processor and the memory communicating via the bus when the electronic device is running, the machine-readable instructions when executed by the processor performing the steps of the first aspect described above, or any possible implementation of the first aspect.

In a fourth aspect, this application further provides a computer-readable storage medium, on which a computer program is stored, where the computer program is executed by a processor to perform the steps in the first aspect or any one of the possible implementation manners of the first aspect.

The method comprises the steps of obtaining target vehicle track data of vehicles driving through a target intersection in a time period to be detected and intersection information of the target intersection, generating a characteristic value of the target intersection under the influence characteristic of traffic efficiency according to the target vehicle track data and the intersection information, inputting the characteristic value into a pre-trained traffic efficiency information detection model, obtaining traffic efficiency information corresponding to the target intersection, and compared with a mode that the traffic efficiency information of the traffic intersection is determined through a manual field judgment method in the prior art, the method is shorter in time consumption, higher in efficiency and higher in accuracy.

In order to make the aforementioned objects, features and advantages of the present application more comprehensible, preferred embodiments accompanied with figures are described in detail below.

Drawings

In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings that are required to be used in the embodiments will be briefly described below, it should be understood that the following drawings only illustrate some embodiments of the present application and therefore should not be considered as limiting the scope, and for those skilled in the art, other related drawings can be obtained from the drawings without inventive effort.

Fig. 1 is a flowchart illustrating a method for detecting traffic efficiency information according to an embodiment of the present application;

fig. 2 is a flowchart illustrating a specific manner of obtaining a speed of a vehicle flow in an intersection in a method for detecting traffic efficiency information according to a first embodiment of the present application;

fig. 3 is a flowchart illustrating a specific manner of acquiring a traffic flow speed of a road section downstream of an intersection in the method for detecting traffic efficiency information according to the first embodiment of the present application;

fig. 4 is a flowchart illustrating a specific manner of obtaining a parking coverage rate of a road section downstream of an intersection in the method for detecting traffic efficiency information according to the first embodiment of the present application;

fig. 5 is a flowchart illustrating a specific manner of training a traffic efficiency information detection model in the method for detecting traffic efficiency information according to an embodiment of the present application;

fig. 6 is a flowchart illustrating another method for detecting traffic efficiency information according to the second embodiment of the present application;

fig. 7 is a flowchart illustrating a specific method for determining a reference speed of a target intersection in the method for detecting traffic efficiency information according to the second embodiment of the present application;

fig. 8 is a flowchart illustrating a specific method for determining whether a target intersection is in a traffic inefficiency state in the method for detecting traffic efficiency information provided in the second embodiment of the present application;

fig. 9 is a flowchart illustrating another method for detecting traffic efficiency information according to a third embodiment of the present application;

fig. 10 is a flowchart illustrating a specific method for determining whether a target intersection is in a traffic-efficient state in the method for detecting traffic efficiency information provided in the third embodiment of the present application;

fig. 11 is a flowchart illustrating another method for detecting traffic efficiency information according to the fourth embodiment of the present application;

fig. 12 is a flowchart illustrating a method for detecting traffic efficiency information according to a fifth embodiment of the present application;

fig. 13 is a flowchart illustrating a specific manner of training a traffic efficiency information detection model in the method for detecting traffic efficiency information according to the fifth embodiment of the present application;

fig. 14 is a flowchart illustrating a specific manner of obtaining a detection result of a sample traffic inefficiency corresponding to each sample intersection in the method for detecting traffic efficiency information according to the fifth embodiment of the present application;

fig. 15 is a flowchart illustrating another specific manner of obtaining the detection result of the sample traffic inefficiency corresponding to each sample intersection in the method for detecting traffic efficiency information provided in the fifth embodiment of the present application;

fig. 16 is a flowchart illustrating another method for detecting traffic efficiency information according to a sixth embodiment of the present application;

fig. 17 is a schematic diagram illustrating an apparatus for detecting traffic efficiency information according to a seventh implementation example of the present application;

fig. 18 shows a schematic diagram of an electronic device 180 according to an eighth embodiment of the present application.

Detailed Description

In order to make the objects, technical solutions and advantages of the embodiments of the present application clearer, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application, and it is obvious that the described embodiments are only a part of the embodiments of the present application, and not all the embodiments. The components of the embodiments of the present application, generally described and illustrated in the figures herein, can be arranged and designed in a wide variety of different configurations. Thus, the following detailed description of the embodiments of the present application, presented in the accompanying drawings, is not intended to limit the scope of the claimed application, but is merely representative of selected embodiments of the application. All other embodiments, which can be derived by a person skilled in the art from the embodiments of the present application without making any creative effort, shall fall within the protection scope of the present application.

It should be noted that: like reference numbers and letters refer to like items in the following figures, and thus, once an item is defined in one figure, it need not be further defined and explained in subsequent figures.

For the convenience of understanding of the present embodiment, first, a detailed description is given to a method for detecting traffic efficiency information disclosed in the embodiments of the present application, and an execution subject of the method for detecting traffic efficiency information provided in the embodiments of the present application is generally an electronic device with computing capability.

Example one

Referring to fig. 1, a flowchart of a method for detecting traffic efficiency information according to an embodiment of the present application is shown, where the method includes steps S101 to S103, where:

s101: the method comprises the steps of obtaining target vehicle track data of vehicles driving through a target intersection in a time period to be detected and intersection information of the target intersection.

In a specific implementation, there may be one or more time periods to be detected. The time period to be detected in the embodiment of the application can be specifically set according to actual needs; for example, 7: 00-10: 00, 10: 00-14: 00, 14: 00-17: 00, 17: 00-20: 00, 20: 00-22: 00 of each day are respectively set as the time periods to be detected. Aiming at the same target intersection, the traffic efficiency information corresponding to the time periods to be detected can be respectively acquired aiming at each time period to be detected.

Specifically, the traffic efficiency information in the embodiment of the present application includes: information indicating and cause of traffic inefficiency; or, information indicating the passage efficiency and the passage efficiency reason.

The target vehicle track data refers to track data of vehicles which drive through a target intersection in a time period to be detected.

In some embodiments, the target intersection is generally made up of two parts: an intersection interior region, and other regions except the intersection interior region.

Wherein, crossing interior region includes: the range between the intersection entrance lane stop line and the exit lane start line.

Generally, the target vehicle trajectory data corresponding to the target intersection may be vehicle trajectory data when the vehicle is traveling through an intersection interior region, or may be trajectory data when the vehicle is traveling through an intersection interior region and traveling through a road connected to the target intersection.

When the target vehicle trajectory data corresponding to the target intersection only includes vehicle trajectory data for a vehicle to drive through an internal area of the intersection, in another embodiment of the present application, before executing S101, the method further includes: and determining the intersection internal area of the target intersection according to the vehicle track data of a plurality of vehicles driving through the target intersection.

At this time, the vehicle trajectory data of the plurality of vehicles that have traveled the target intersection includes the trajectory data of the vehicle that has traveled the inside area of the intersection and the road connected to the target intersection, and then the target vehicle trajectory data of the vehicle that has traveled the target intersection for the time period to be detected is determined based on the vehicle trajectory data.

It should be noted here that the plurality of vehicles corresponding to the vehicle trajectory data of the internal area of the confirmation target intersection may be the same as or different from the vehicle for which the target vehicle trajectory data is determined.

And when the passing efficiency information of the target intersection is obtained for multiple times, the intersection internal area of the target intersection is determined once.

When the intersection internal area of the target intersection is determined, the intersection internal area can be marked, the starting point of the acquired target vehicle track data falls on the stop line of the intersection entrance lane, and the end point of the target vehicle track data falls on the starting point line of the intersection exit lane.

In addition, the related information of the intersection internal area of the target intersection can be directly obtained from a traffic management department or a related department of city planning and construction.

The traffic efficiency is affected by the traffic itself and also by the facilities associated with the intersection. Therefore, in the implementation of the present application, intersection information of the target intersection is also obtained.

The intersection information of the target intersection is related facility information of the target intersection, such as a road direction of the target intersection, a width of an exit corresponding to each road direction, a gradient of the intersection, a sight distance, the number of stop lines at the intersection, a timing of a signal light, a distance between another intersection closest to the downstream of the target route, a specific setting mode of a bus stop near the target intersection, and the like.

Receiving the above S101, after acquiring the target vehicle trajectory data and the intersection information, the method further includes:

s102: and generating a characteristic value of the target intersection under the traffic efficiency influence characteristic according to the target vehicle track data and the intersection information.

In particular implementations, the traffic efficiency impact characteristics include: one or more of road condition characteristics, road characteristics, and time characteristics.

A, to the condition that current efficiency influences the characteristic and includes road conditions characteristic, road conditions characteristic includes: one or more of the speed of the vehicle flow in the intersection, the speed of the vehicle flow on the downstream road section of the intersection and the parking coverage rate on the downstream road section of the intersection.

a 1: as for the condition that the road condition characteristics include the speed of the vehicle flow in the intersection, referring to fig. 2, an embodiment of the present application provides a specific way of obtaining the speed of the vehicle flow in the intersection, including:

s201: and calculating the speed of each target vehicle passing through the target intersection, and the entrance time and the exit time of each target vehicle passing through the target intersection according to the target vehicle track data of each vehicle.

Here, the target vehicle trajectory data corresponding to each vehicle includes: the position information of a plurality of positioning points and the positioning time corresponding to each positioning point.

If the acquired target road track data only comprises track data of vehicles passing through the internal area of the target intersection, according to the sequence of positioning time, an initial positioning point of the target vehicle track data is an initial position of the vehicles entering the target intersection, and the positioning time corresponding to the initial positioning point is the driving-in time of the vehicles driving through the target intersection; the end positioning point of the target vehicle track data is the position of the vehicle driving out of the target intersection, and the positioning time corresponding to the end positioning point is the driving-out time of the vehicle driving through the target intersection.

If the acquired target road track data comprises track data of the vehicle passing through the inner area of the target intersection and track data of the outer area of the target intersection, an entering positioning point of the vehicle entering the target intersection and an exiting positioning point of the vehicle exiting the target intersection can be determined from the target vehicle track data according to the area range of the target intersection and the position information of the positioning point. The positioning time corresponding to the entry positioning point is the entry time; and the positioning time corresponding to the exit positioning point is the exit time.

Calculating the speed of each vehicle passing through the target intersection by adopting the following modes:

for each vehicle, performing:

Here, the average of the respective temporary speeds may be determined as the speed at which the vehicle passes through the target intersection. Alternatively, all temporary speeds are directly determined as the speed at which the vehicle is driving through the target intersection.

S202: and fitting the speed of each vehicle passing through the target intersection, the entrance time and the exit time to obtain the speed of the vehicle flow in the intersection.

Here, since the speed at which the vehicle passes through the target intersection is the speed in a period of time from the time at which the vehicle enters the target intersection to the time at which the vehicle exits the target intersection, it is expressed as a curve from a certain time point to another time point in mathematical expression, for example, in a coordinate system established with the time as a horizontal axis and the speed as a vertical axis. And fitting the speed of each vehicle passing through the target intersection, the driving-in time and the driving-out time, namely fitting a curve of the first time speed corresponding to each vehicle, so as to obtain a second time speed curve capable of representing that all vehicles drive through the target intersection within the time period to be detected, wherein the second time speed curve can be converted into a function related to time and speed.

For example, the speed of the vehicle flow in the intersection may be an average speed of all vehicles driving through the target intersection in the time period to be detected. Specifically, according to the obtained time-speed curve of all vehicles driving through the target intersection in the time period to be detected, taking a preset time period as a time interval, a plurality of points are taken on the curve, and the average value of the speeds corresponding to the points is the average speed of all vehicles driving through the target intersection in the time period to be detected.

For example, the traffic flow speed in the intersection may also be the traffic flow speed corresponding to each time segment obtained by dividing the time period to be detected into a plurality of time segments. Specifically, a plurality of points may be taken on the curve with a preset time period as a time interval according to the obtained second time-speed curve that all vehicles drive through the target intersection within the time period to be detected, where the duration length of each time period is equal to the duration length of the time segment. The speed corresponding to each point obtained is the traffic speed corresponding to each time segment.

a 2: as for the condition that the road condition characteristics include the traffic flow speed of the road section downstream of the intersection, referring to fig. 3, an embodiment of the present application provides a specific manner for obtaining the traffic flow speed of the road section downstream of the intersection, including:

s301: and acquiring downstream vehicle track data of vehicles driving through a downstream road section of the target intersection in a time period to be detected.

Here, the intersection downstream road section is a downstream road section of the target intersection, and means a road section located downstream of the target intersection and having a distance to the target intersection smaller than a preset distance threshold.

For example, a road segment downstream of the target intersection and less than 50 meters away from the target intersection is determined as a downstream road segment of the target intersection.

Illustratively, each target intersection will typically include a plurality of road directions, including, for example: one or more of east-west straight going, east-north turning, east-south turning, west-east straight going, west-north turning, west-south turning, north-south straight going, north-west turning, north-east turning, south-north straight going, south-west turning, and south-east turning.

And each road direction corresponds to a downstream road section.

After downstream vehicle track data of vehicles driving through a downstream road section of the target intersection in a time period to be detected is acquired, the method further comprises the following steps:

s302: and generating the traffic flow speed of the downstream road section of the intersection according to the downstream vehicle track data.

Here, the downstream vehicle trajectory data also includes position information of a plurality of positioning points and a positioning time of each positioning point. The generation mode of the traffic flow speed of the downstream road section of the intersection is similar to the generation mode of the traffic flow speed in the intersection, and details are not repeated here.

a 3: the road condition characteristics comprise the condition of parking coverage rate of the road sections at the downstream of the intersection, and the parking coverage rate refers to the proportion of the area covered by the vehicles to the total area of the road sections at the downstream of the intersection when the road sections at the downstream of the intersection stop. Referring to fig. 4, an embodiment of the present application provides a specific manner for obtaining a parking coverage rate of a downstream road segment of an intersection, including:

s401: determining parking information of each vehicle corresponding to the downstream vehicle track data on the downstream road section of the intersection according to the downstream vehicle track data; the parking information includes: a dwell position and dwell start and end times;

s402: the vehicles are divided into a plurality of groups according to the stop start time and the stop end time.

S403: and for each group, determining the queuing length according to the stopping position of the vehicle in the group, and calculating the parking coverage rate according to the queuing length and the length of the road section downstream of the intersection.

Here, the vehicle is divided into a plurality of groups according to the stop start time and the stop termination time, and when a stop occurs at a certain time on a section downstream of the intersection, the vehicle staying on the section downstream of the intersection is distinguished from other vehicles. Wherein one group corresponds to a parking event occurring on a downstream road segment of the intersection.

Then, according to the stopping position of each vehicle stopping on the downstream road section of the intersection in the parking event, the queuing length can be calculated, and the queuing length corresponds to the length of the road section covered by the parking on the downstream road section of the intersection. And then the parking coverage rate can be calculated according to the queuing length and the length of the road section at the downstream of the intersection.

B: the road characteristics include: the bus stop comprises one or more of turning radius, intersection slope, intersection sight distance, the number of intersection stop lines, intersection exit lane width, the distance between a road entrance nearest to an intersection exit and an intersection, the distance between a bus station nearest to the intersection exit and the intersection, and the number of buses stopped at the bus station nearest to the intersection exit.

For the case that the road characteristics include the turning radius, the turning radius can be directly calculated according to the road direction of the target intersection and the road width of each road direction of the target intersection.

In addition, in order to accurately obtain the accurate turning radius of the moment to be detected, the turning radius can be calculated according to target track data when the vehicle passes through the target intersection.

Specifically, the turning radius of each vehicle can be determined based on target vehicle trajectory data of the vehicle, each of which turns left or right in the road direction; then, the average value of the turning radius of all vehicles is obtained and is used as the turning radius of the target intersection.

C: the time characteristics include: the time period to be detected is one or more of a date corresponding to the time period to be detected, whether the time period to be detected is a traffic peak, the number of weeks corresponding to the time period to be detected, whether the time period to be detected is a holiday and the number of months corresponding to the time period to be detected.

After receiving the above S102, after generating the characteristic value of the target intersection under the traffic efficiency influence characteristic in the target vehicle trajectory data and the intersection information, the method further includes:

s103: and inputting the characteristic value into a pre-trained traffic efficiency information detection model, and acquiring traffic efficiency information corresponding to the target intersection.

In specific implementation, the traffic efficiency information detection model is obtained by training the characteristic values of the sample intersections under the traffic efficiency influence characteristics and the traffic efficiency information corresponding to each sample road.

Specifically, referring to fig. 5, in the embodiment of the present application, a traffic efficiency information detection model is obtained through training in the following manner:

s501: the method comprises the steps of obtaining sample vehicle track data of vehicles driving through at least one sample intersection in a sample time period, sample intersection information of each sample intersection and sample traffic efficiency information corresponding to each sample intersection.

Here, the sample time periods corresponding to different sample intersections may be the same or may not be the same.

Specifically, the sample vehicle track data of the vehicle at the sample intersection and the acquisition manner of the sample intersection information are similar to the acquisition manner of the intersection information of the target vehicle track data of the vehicle at the target intersection, and are not described herein again.

S502: and generating a sample characteristic value of each sample intersection under the traffic efficiency influence characteristic according to the sample vehicle track data and the sample intersection information.

Here, the traffic efficiency affecting characteristics include: one or more of road condition characteristics, road characteristics, and time characteristics.

The road condition characteristics and the road characteristics are similar to those in the above S102, and are not described herein again.

In addition, it should be noted that the temporal characteristics include: the date corresponding to the sample time period, whether the sample time period is a traffic peak, the week number corresponding to the sample time period, whether the sample time period is a holiday, and the month number corresponding to the sample time period.

The obtaining mode of the sample characteristics of each sample intersection under the traffic efficiency influence characteristic is similar to the obtaining mode of the characteristic value of the target intersection under the traffic efficiency influence characteristic, and is not repeated here.

S503: and constructing a decision tree model by taking the sample characteristic value of each sample intersection under the traffic efficiency influence characteristic as an input characteristic value of the decision tree model and taking the sample traffic efficiency information corresponding to each sample intersection as an output characteristic value of the decision tree model, and taking the constructed decision tree model as the traffic efficiency information detection model.

Specifically, the constructed decision tree model may include one decision tree or may include a plurality of decision trees.

Firstly, the method comprises the following steps: when the constructed decision tree model comprises a decision tree, the following method can be adopted:

and calculating the information gain of each traffic efficiency influence characteristic according to the sample characteristic value of each sample intersection under the traffic efficiency influence characteristic.

And taking the traffic efficiency influence characteristic with the maximum corresponding information gain as a parent node of the decision tree.

And dividing the sample intersection into each characteristic value interval respectively according to the plurality of characteristic value intervals corresponding to the parent node.

And aiming at each characteristic value range, calculating the information gain corresponding to each traffic efficiency influence characteristic except the node existing in the path where the father node is located according to the sample characteristic value under the traffic efficiency influence characteristic of the sample intersection in the characteristic value range, and taking the traffic efficiency influence characteristic with the maximum corresponding information gain as a child node of the father node.

And taking the child node as a new father node, forming a new sample intersection set by the sample intersections of the characteristic value range corresponding to the new father node, returning to the step of calculating the information gain of each traffic efficiency influence characteristic according to the sample characteristic value of each sample intersection under the traffic efficiency influence characteristic, and knowing that each road contains all traffic efficiency influence characteristics.

Then, according to a plurality of characteristic value intervals corresponding to the last node of each road in the decision tree, the sample intersections corresponding to the node are divided into each characteristic value interval, and the sample traffic efficiency information of the sample intersections in each characteristic value interval is determined as the traffic efficiency information corresponding to each characteristic value interval corresponding to the node.

In another embodiment of the present application, after the construction of the decision tree model is completed, the accuracy of the traffic efficiency information detection model can be verified by using the test feature values of the test intersections under the traffic efficiency influence features and the test traffic efficiency information corresponding to each test intersection.

If the precision of the decision tree model is smaller than a certain threshold value, the traffic efficiency information detection model can be reconstructed by using the relevant data of the test intersections and the sample intersections, and the precision of the traffic efficiency information detection model is verified by using the test characteristic values of the new test intersections under the traffic efficiency influence characteristic and the test traffic efficiency information corresponding to each test intersection until the precision of the constructed traffic efficiency information detection model meets the requirement.

Specifically, the process of verifying the constructed traffic efficiency information detection model includes:

acquiring test vehicle track data of a vehicle driving through at least one test intersection in a test time period, test intersection information of each test intersection and test traffic efficiency information corresponding to each test intersection;

generating a test characteristic value of each test intersection under the traffic efficiency influence characteristic according to the test vehicle track data and the test intersection information;

inputting the test characteristic value of each test intersection under the traffic efficiency influence characteristic into the traffic efficiency information detection model, and acquiring the detection result of the low-efficiency reason corresponding to each test intersection;

and verifying the low-efficiency reason detection model according to the low-efficiency reason detection result and the test traffic efficiency information corresponding to each test intersection.

II, secondly: when a plurality of decision trees are included in the constructed decision tree model, the following method can be adopted:

and selecting a plurality of target influence characteristics from the traffic efficiency influence characteristics, and constructing a decision tree based on the target influence characteristics.

And executing the construction process of the decision tree for multiple times, wherein the target influence characteristics used by the constructed decision tree each time have distinguishing characteristics.

Here, the process of constructing a decision tree based on the target influence characteristics is similar to the above process of constructing a decision tree using all traffic efficiency influence characteristics, and is not described herein again.

In addition, when the constructed decision tree model comprises a plurality of decision trees, the precision of the decision tree model is ensured by adopting a mode of verifying the constructed decision trees while constructing.

Specifically, after each decision tree is constructed, all the currently constructed decision trees form a current decision tree set. Aiming at each test intersection, inputting a test characteristic value of the test intersection under the target influence characteristic corresponding to each decision tree into the corresponding decision tree, and acquiring an inefficient reason detection result corresponding to each decision tree;

and then weighting and summing the detection results of the ineffective reasons corresponding to the carved decision tree to obtain the detection result of the ineffective reasons corresponding to the test intersection.

And verifying the accuracy of the current decision tree set according to the detection result of the low-efficiency reasons corresponding to each test intersection and the test traffic efficiency information. If the verification is passed, stopping building a new decision tree; and if the verification fails, continuing to construct a new decision tree.

Example two

Referring to fig. 6, a flowchart of a method for detecting traffic efficiency information according to a second embodiment of the present application is shown, where the method includes steps S601 to S604, where:

s601: the method comprises the steps of obtaining target vehicle track data of vehicles driving through a target intersection in a time period to be detected and intersection information of the target intersection.

Here, the implementation of S601 is similar to S101 described above, and is not described here again.

S602: and determining whether the target intersection is in a low-efficiency passing state or not according to the target vehicle track data and the reference speed of the target intersection. If yes, jumping to S603; if not, then jump to S605.

In a specific implementation, the reference speed of the target intersection refers to the speed of the vehicle passing through the target intersection when the traffic flow of the target intersection is small, for example, 0:00 to 5:00 in the morning. That is, the speed at which the vehicle smoothly passes through the target intersection.

Specifically, referring to fig. 7, the embodiment of the present application provides a specific way to determine the reference speed of the target intersection, including:

s701: reference vehicle trajectory data of a vehicle driving through the target intersection within a reference time period is acquired.

In specific implementation, the reference time period can be specifically set according to actual needs. And the reference time periods of different intersections can be the same or different.

For example, if the traffic flow of a certain intersection is less at 0: 00-5: 00, 0: 00-5: 00 is determined as the reference time period of the intersection. And if the traffic flow of a certain intersection is less at 1: 00-3: 00, determining 1: 00-3: 00 as the reference time period of the intersection.

The acquired reference vehicle trajectory data may be reference vehicle trajectory data corresponding to a reference time period of each day within a preset time period, for example, within one week, within one month, within three months, within five months, within one year.

The method for acquiring the reference vehicle track data of the target intersection in the reference time period is similar to the method for acquiring the target vehicle track data of the target intersection in the time period to be detected, and is not repeated herein.

S702: and determining the reference speed corresponding to the target intersection according to the reference vehicle track data.

In a specific implementation, the speed of each vehicle driving through the target intersection in a reference time period can be determined according to the reference vehicle track data; and determining the average value of the speeds of all vehicles driving through the target intersection in the reference time period as the reference speed corresponding to the target intersection.

Specifically, since the vehicle trajectory data is a set of data composed of different positioning points, each positioning point includes the geographic coordinates of the positioning point, and the positioning time of the positioning point. It is therefore possible to calculate the time difference between two adjacent positioning points, that is, the time required for the vehicle to move from the upstream positioning point to the downstream positioning point of the two adjacent positioning points, based on the positioning time of the two adjacent positioning points. Then, calculating the distance between the two adjacent positioning points according to the geographic coordinates of the two adjacent positioning points; and then calculating the moving speed of the vehicle between the two adjacent positioning points according to the distance between the two positioning points and the time difference. And then determining the speed of the vehicle passing through the target intersection in the reference time period according to the speed of the vehicle moving between each two adjacent positioning points.

Referring to fig. 8, an embodiment of the present application further provides a specific method for determining whether the target intersection is in a traffic inefficiency state according to the target vehicle trajectory data and the reference speed of the target intersection, including:

s801: and acquiring the speed of the target intersection in the intersection of the time period to be detected according to the target vehicle track data.

Here, the method for determining the speed in the intersection is similar to the method for determining the reference speed, and is not described herein again.

S802: detecting whether the speed difference between the reference speed and the speed in the intersection is greater than a preset first speed difference threshold value, and detecting whether the speed in the intersection is less than the preset first speed threshold value. If not, jumping to S803; if so, it jumps to S805.

Here, the first speed difference threshold value and the first speed threshold value are specifically set according to actual needs. When setting, the corresponding first speed difference threshold values of different target intersections can be the same or different; similarly, the first speed thresholds corresponding to different target intersections may be the same or different.

Illustratively, with the sophistication of urban traffic networks, the number of intersections to be detected is very large, and it is obviously not practical to determine a first speed threshold for each intersection corresponding to that intersection. The same first speed threshold can be used for each intersection of the same area.

And if the speed difference between the reference speed and the speed in the intersection is greater than a preset first speed difference threshold value, the speed in the intersection of the time period to be detected is considered to be too large different from the speed of the vehicle when the vehicle smoothly passes through the target intersection.

In addition, if the speed difference between the reference speed and the speed in the intersection is greater than a preset first speed difference threshold value, the reference speed is very high due to the setting reason of the target intersection; at the moment, the speed difference between the reference speed and the speed in the intersection is greater than a preset first speed difference threshold value, but the speed in the intersection of the target intersection is still in a higher state, and the target intersection is not actually in a low-efficiency passing state; therefore, to further constrain the intersection with the second speed threshold, the target intersection is determined to be in a traffic inefficiency state only after the speed in the intersection is less than the preset second speed threshold and the speed difference between the reference speed and the speed in the intersection is greater than the preset first speed difference threshold.

And only when the two conditions are simultaneously met, determining that the target intersection is in the low-efficiency passing state.

There may be a traffic inefficiency caused by traffic causes.

S803: and judging whether the reference speed is smaller than a preset second speed threshold value or not. If yes, jumping to S805; if not, jumping to S804.

Here, if the reference speed is less than the preset second speed threshold, even if there is no inefficiency due to traffic causes, there may be a traffic inefficiency state due to a road setting problem of the target intersection itself.

S804: confirming that the target intersection is not in the traffic inefficiency state.

S805: confirming that the target intersection is in the traffic inefficiency state.

It should be noted that in the above S802 and S803, S803 may be executed first, and when the determination result in S803 is no, S802 may be executed again. Namely: after the reference speed is judged to be not less than a preset first speed threshold, whether the speed difference between the reference speed and the speed in the intersection is greater than a preset first speed difference threshold or not and whether the speed in the intersection is less than a preset second speed threshold or not are detected.

As long as any one of the conditions of S802 and S803 is met, it can be determined that the target intersection is in the traffic inefficiency state.

Receiving the step S602, the method for detecting traffic efficiency information provided in the second embodiment of the present application further includes:

s603: and generating a characteristic value of the target intersection under the traffic efficiency influence characteristic according to the target vehicle track data and the intersection information.

Here, the implementation of S603 is similar to that of S102 described above, and is not described here again.

S604: and inputting the characteristic value into a pre-trained traffic efficiency information detection model, and acquiring traffic efficiency information corresponding to the target intersection.

Here, the implementation of S604 is similar to the implementation of S103 described above, and is not described here again.

S605: and (6) ending.

After acquiring target vehicle track data of vehicles at a target intersection and intersection information of the target intersection, determining whether the target intersection is in a low-efficiency passing state or not according to the target vehicle track data and the reference speed of the target intersection; after the target intersection is determined to be in the low-efficiency passing state, according to the target vehicle track data and the intersection information, a characteristic value of the target intersection under the influence characteristic of the passing efficiency is generated, the characteristic value is input into a pre-trained passing efficiency information detection model, and the passing efficiency information corresponding to the target intersection is obtained, so that the condition that the low-efficiency passing state exists can be filtered out firstly, the corresponding passing efficiency information is determined only for the condition that the low-efficiency passing state exists, the data processing amount of the model is reduced, and the calculation resources are saved.

EXAMPLE III

Referring to fig. 9, a flowchart of a method for detecting traffic efficiency information according to a second embodiment of the present application is shown, where the method includes steps S901 to S904, where:

s901: the method comprises the steps of obtaining target vehicle track data of vehicles driving through a target intersection in a time period to be detected and intersection information of the target intersection.

Here, the implementation of S901 is similar to S101 described above, and is not described here again.

S902: and determining whether the target intersection is in a high-efficiency passing state or not according to the target vehicle track data and the reference speed of the target intersection. If yes, jumping to S903; if not, then jump to S905.

Here, the determination method of the reference speed of the target intersection is similar to the determination method of the reference speed in the second embodiment, and is not described herein again.

Referring to fig. 10, an embodiment of the present application further provides a specific method for determining whether the target intersection is in a traffic-efficient state according to the target vehicle trajectory data and the reference speed of the target intersection, including:

s1001: and acquiring the speed of the target intersection in the intersection of the time period to be detected according to the target vehicle track data.

S1002: and detecting whether the speed difference between the reference speed and the speed in the intersection is smaller than a preset second speed difference threshold value or not, and detecting whether the speed in the intersection is larger than a preset third speed threshold value or not. If not, jumping to S1003; if so, it jumps to S1004.

Here, the second speed difference threshold and the third speed threshold are specifically set according to actual needs. In setting, corresponding second speed difference thresholds of different target intersections can be the same or different; the third speed thresholds corresponding to different target intersections may be the same or different.

Illustratively, with the complexity of the urban traffic network, the number of intersections to be detected is very large, and it is obviously not practical to determine a third speed threshold corresponding to each intersection. The same third speed threshold can be used for each intersection of the same area.

And only when the two conditions are simultaneously met, determining that the target intersection is in the high-efficiency passing state.

S1003: and confirming that the target intersection is not in the high-efficiency passing state.

S1004: and confirming that the target intersection is in the high-efficiency passing state.

It should be noted that, in the above S1002 and S1003, S1003 may be executed first, and when the determination result of S1003 is no, S1002 may be executed again.

With reference to the foregoing S902, the method for detecting traffic efficiency information provided in the second embodiment of the present application further includes:

s903: and generating a characteristic value of the target intersection under the traffic efficiency influence characteristic according to the target vehicle track data and the intersection information.

Here, the implementation of S903 is similar to that of S102 described above, and is not described here again.

S904: and inputting the characteristic value into a pre-trained traffic efficiency information detection model, and acquiring traffic efficiency information corresponding to the target intersection.

Here, the implementation of S904 is similar to that of S103 described above, and is not described here again.

S905: and (6) ending.

After acquiring target vehicle track data of vehicles at a target intersection and intersection information of the target intersection, determining whether the target intersection is in a high-efficiency passing state or not according to the target vehicle track data and the reference speed of the target intersection; after the target intersection is determined to be in the high-efficiency passing state, according to the target vehicle track data and the intersection information, a characteristic value of the target intersection under the characteristic of the influence of passing efficiency is generated, the characteristic value is input into a pre-trained passing efficiency information detection model, and the passing efficiency information corresponding to the target intersection is obtained, so that the situation that the high-efficiency passing state does not exist is filtered out at first, the corresponding passing efficiency information is determined only according to the situation that the high-efficiency passing state exists, the data processing amount of the model is reduced, and the calculation resources are saved.

In addition, after the target intersection is determined to be in a high-efficiency passing state or a low-efficiency passing state, a characteristic value of the target intersection under the passing efficiency influence characteristic is generated according to the target vehicle track data and the intersection information, and the passing efficiency information corresponding to the target intersection is obtained based on the characteristic value under the passing efficiency influence characteristic and a pre-trained passing path information detection model, so that the situations of no high-efficiency passing state and low-efficiency passing state can be filtered out, the corresponding passing efficiency information is determined only according to the situations of the high-efficiency passing state and the low-efficiency passing state, the data processing amount of the model is reduced, and the calculation resources are saved.

Example four

Referring to fig. 11, a fourth embodiment of the present application further provides another method for detecting traffic efficiency information, including:

s1101: acquiring target vehicle track data of vehicles driving in the road direction in a time period to be detected aiming at each road direction of the target intersection;

s1102: generating characteristic values of all road directions under the traffic efficiency influence characteristic according to target vehicle track data of vehicles driving through all road directions in a time period to be detected;

s1103: and inputting the characteristic value of each road direction under the traffic efficiency influence characteristic into a traffic efficiency information detection model trained in advance, and acquiring traffic efficiency information corresponding to each road direction.

Here, the specific implementation manners of S1101 to S1103 are similar to the implementation manners of S101 to S103, and are not described again here.

The embodiment of the application can detect the traffic efficiency information aiming at each road direction of the target intersection, and has pertinence.

EXAMPLE five

The road condition of the target intersection in the time period to be detected is not only influenced by the road condition in the time period to be detected, but also continuously influenced by the road condition of the target intersection in the time period to be detected before the time period to be detected, so in another embodiment of the present application, not only the target vehicle track data of the vehicle driving through the target intersection in the time period to be detected, but also the historical vehicle track data of the vehicle driving through the target road in the historical detection time period before the time period to be detected are acquired.

Referring to fig. 12, another method for detecting traffic efficiency information according to a fifth embodiment of the present application includes:

s1201: the method comprises the steps of obtaining target vehicle track data of vehicles driving through a target intersection in a time period to be detected and intersection information of the target intersection, and obtaining historical vehicle track data of vehicles driving through the target intersection in a historical detection time period before the time period to be detected.

Here, the specific manner of acquiring the target vehicle trajectory data and the intersection information of the target intersection is referred to the first embodiment, and is not described herein again.

The historical detection time period before the time period to be detected is generally a time period which has a continuous relation with the time period to be detected, for example, the time period to be detected is 7: 00-10: 00 in 2018, 1 month, 15 days, and 6: 00-7: 00 in 2018, 1 month, 15 days.

In addition, if the time length of the time period to be detected is long, the congestion degree and the congestion type may be different at different moments of the same intersection in the time period to be detected, for example, 7: 00-9:00 are set as the time period to be detected, but a road in one road direction of a road connected with a target intersection is changed into a bus-only line in 8:00-9:00, so that if the target intersection has a traffic inefficiency, 7: 00-8: 00 may be inefficient due to traffic conditions, and 8:00-9:00 may be inefficient due to road settings besides the traffic conditions. Therefore, the obtained traffic efficiency information of the time period to be detected can be multiple.

In addition, the time length of the time period to be detected can be shortened, so that the obtained traffic efficiency information of the time period to be detected only aims at the time period to be detected. For example, the length of the time period to be detected is set to be 15 minutes, the time period is from 7:00 to 9:00, the detection is carried out once every 15 minutes, and a plurality of pieces of traffic efficiency information from 7:00 to 9:00 can be obtained.

S1202: generating a first characteristic value of the target intersection under the road condition characteristic according to the target vehicle track data, generating a second characteristic value of the target intersection under the road condition characteristic according to the historical vehicle track data, and generating a third characteristic value of the target intersection under the road characteristic according to the intersection information.

S1203: and inputting the first characteristic value, the second characteristic value and the third characteristic value into a pre-trained traffic efficiency information detection model, and acquiring traffic efficiency information corresponding to the target intersection.

In another embodiment, when a fourth feature value of the target intersection under the time feature is obtained, the first feature value, the second feature value, the third feature value and the fourth feature value are input into a traffic efficiency information detection model trained in advance, and traffic efficiency information corresponding to the target intersection is obtained.

I: the training process of the traffic efficiency information detection model to obtain the first feature value, the second feature value and the third feature value is described below.

Specifically, referring to fig. 13, in the embodiment of the present application, a traffic efficiency information detection model is obtained through training in the following manner:

s1301: the method comprises the steps of obtaining sample vehicle track data of vehicles driving through at least one sample intersection in a sample time period, sample historical vehicle track data of vehicles driving through the sample intersections in a sample historical detection time period corresponding to the sample time period, sample intersection information of each sample intersection and sample traffic efficiency information corresponding to each sample intersection.

Specifically, the acquisition modes of the sample vehicle track data, the sample historical vehicle track data and the sample intersection information of the vehicles at the sample intersection are similar to the acquisition modes of the target vehicle track data, the historical vehicle track data and the intersection information at the target intersection, and are not described herein again.

S1302: generating a first sample characteristic value of the sample intersection under the road condition characteristic according to the sample vehicle track data, generating a second sample characteristic value of the sample intersection under the road condition characteristic according to the sample historical vehicle track data, and generating a third sample characteristic value of the sample intersection under the road condition characteristic according to the intersection information.

The obtaining manner of the first sample characteristic value, the second sample characteristic value and the third sample characteristic value at each sample intersection is similar to the obtaining method of the first characteristic value, the second characteristic value and the third characteristic value at the target intersection, and is not described herein again.

S1303: inputting the first sample characteristic value, the second sample characteristic value and the third sample characteristic value into a basic prediction model, and obtaining a sample passing low-efficiency detection result corresponding to each sample intersection.

S1304: and training the basic prediction model according to the sample traffic inefficiency detection result and the sample traffic efficiency information, and taking the trained basic prediction model as the traffic efficiency information detection model.

In a specific implementation, the base prediction model includes: a first neural network, a second neural network, and a third neural network.

Referring to fig. 14, an embodiment of the present application provides a specific manner of inputting the first sample feature value, the second sample feature value, and the third sample feature value into a basic prediction model to obtain a sample traffic inefficiency detection result corresponding to each of the sample intersections, including:

for each sample intersection, performing:

s1401: and fusing a first feature vector consisting of the first sample feature values and a second feature vector consisting of the second sample feature values to generate a first intermediate sample feature vector.

Here, the first feature vector formed of the first sample feature values and the second feature vector formed of the second sample feature values may be fused, and the first feature vector and the second feature vector may be subjected to weighted summation. Generally, since the first feature vector is obtained based on the target vehicle track of the period to be detected, the second feature vector is obtained based on the historical vehicle track of the historical detection period, and the first feature vector has a greater influence on the inefficient state of the current target intersection, the weight of the first feature vector is greater than that of the second feature vector when the first feature vector and the second feature vector are subjected to weighted summation. The weights may be adjusted during training of the underlying predictive model, or may be used as hyper-parameters.

In addition, a first feature vector composed of the first sample feature value and a second feature vector composed of the second sample feature value may be fused, or the first feature vector and the second feature vector may be spliced.

S1402: and inputting the first intermediate sample feature vector into the first neural network to obtain a second intermediate sample feature vector.

Here, the first neural network includes at least one fully-connected layer. After the first intermediate sample feature vector passes through the first neural network, each element in the obtained second intermediate sample feature vector is influenced by all features in the first intermediate sample feature vector, so that the second intermediate sample feature vector establishes a relation between each element in the first intermediate sample feature vector, namely the features of the first feature vector and the second feature vector are extracted through the first neural network.

S1403: inputting a third sample feature vector formed by the third sample feature value into the second neural network, and obtaining a third intermediate sample feature vector.

Here, the second neural network also includes at least one fully-connected layer. After the third sample feature vector passes through the second neural network, each element in the obtained third intermediate sample feature vector is influenced by all features in the third sample feature vector, so that each element in the third intermediate sample feature vector can characterize each element in the third sample feature vector to a certain extent.

It should be noted that S1402 and S1403 are not executed in a sequential order. Only when both are executed, the following S1404 is executed.

S1404: and splicing the second intermediate sample feature vector and the third intermediate sample feature vector to form a spliced vector.

S1405: and inputting the splicing vector into a third neural network to obtain the detection result of the passage inefficiency of the sample.

Here, the third neural network also includes at least one fully-connected layer and includes a classifier. And after passing through the third neural network, the spliced vector enters a classifier to obtain a final classification result, namely corresponding to the sample intersection.

And obtaining a sample passing low-efficiency detection result.

And after a sample passing low-efficiency detection result is obtained, training the basic prediction model according to the sample passing low-efficiency detection result and the sample passing efficiency information.

Specifically, the basic prediction model may be trained in the following manner:

and calculating the model loss according to the sample passing low-efficiency detection result corresponding to each sample intersection and the sample passing efficiency information.

Comparing the model loss with a preset loss threshold;

and if the model loss is larger than a preset loss threshold value, adjusting parameters of the basic prediction model, re-obtaining the sample passing low-efficiency detection result corresponding to each sample intersection based on the basic prediction model after the parameters are adjusted, and calculating the model loss based on the newly obtained sample passing low-efficiency detection result and the corresponding sample passing efficiency information.

And if the model loss is not greater than a preset loss threshold value, determining the current basic prediction model as a traffic efficiency information detection model.

Here, the model loss can be calculated in the following manner:

and comparing the sample passing low-efficiency detection result corresponding to each sample intersection with the sample passing efficiency information.

And taking the number of sample intersections with different sample traffic efficiency information and the sample traffic inefficiency detection result as the error number.

The ratio between the number of errors and the number of sample intersections is determined as the model loss.

II: the following describes a training process of obtaining a first feature value, a second feature value, a third feature value, and a fourth feature value for a traffic efficiency information detection model, where the training process includes:

the method comprises the steps of obtaining sample vehicle track data of vehicles driving through at least one sample intersection in a sample time period, sample historical vehicle track data of vehicles driving through the sample intersections in a sample historical detection time period corresponding to the sample time period, sample intersection information of each sample intersection and sample traffic efficiency information corresponding to each sample intersection.

Generating a first sample characteristic value of the sample intersection under the road condition characteristic according to the sample vehicle track data, generating a second sample characteristic value of the sample intersection under the road condition characteristic according to the sample historical vehicle track data, generating a third sample characteristic value of the sample intersection under the road condition characteristic according to the intersection information, and acquiring a fourth sample characteristic value of the sample time under the time characteristic.

Specific implementation procedures can be found in the above description of I.

Referring to fig. 15, an embodiment of the present application further provides a specific manner of inputting the first sample feature value, the second sample feature value, the third sample feature value, and the fourth sample feature value into a basic prediction model to obtain a sample passage inefficiency detection result corresponding to each of the sample intersections, including:

s1501: and fusing a first feature vector consisting of the first sample feature values and a second feature vector consisting of the second sample feature values to generate a first intermediate sample feature vector.

Here, the implementation of S1501 is similar to S1401 described above, and is not described herein again.

S1502: and inputting the first intermediate sample feature vector into the first neural network to obtain a second intermediate sample feature vector.

Here, the implementation of S1502 is similar to S1402 described above, and is not described herein again.

S1503: and splicing a third sample feature vector formed by the third sample feature value and a fourth sample feature vector formed by the fourth sample feature value, and inputting the spliced third sample feature vector and the fourth sample feature vector into the second neural network to obtain a third intermediate sample feature vector.

S1504: and splicing the second intermediate sample feature vector and the third intermediate sample feature vector to form a spliced vector.

Here, the implementation of S1504 is similar to S1404 described above, and is not described herein again.

S1505: and inputting the splicing vector into a third neural network to obtain the detection result of the passage inefficiency of the sample.

Here, the implementation of S1505 is similar to S1405 described above, and is not described herein again.

After the traffic efficiency information detection model is obtained through training, aiming at the condition that the traffic efficiency information detection model is obtained through training based on the first sample characteristic value, the second sample characteristic value and the third sample characteristic value, the first characteristic value, the second characteristic value and the third characteristic value corresponding to the target intersection can be input into the traffic efficiency information detection model, and the traffic efficiency information of the target intersection is obtained.

And inputting a first characteristic value, a second characteristic value, a third characteristic value and a fourth characteristic value corresponding to the target intersection into the traffic efficiency information detection model to obtain the traffic efficiency information of the target intersection aiming at the condition obtained by training the traffic efficiency information detection model based on the first sample characteristic value, the second sample characteristic value, the third sample characteristic value and the fourth sample characteristic value.

The embodiment of the application can automatically detect the traffic efficiency information of the target intersection, and compared with a method for judging through a manual site in the prior art, the method is short in time consumption, high in efficiency and higher in accuracy.

EXAMPLE six

Referring to fig. 16, a flowchart of a method for detecting traffic efficiency information according to a sixth embodiment of the present application is shown, where the method includes steps S1601 to S1604, where:

s1601: acquiring target vehicle track data of a vehicle driving through a target intersection in a time period to be detected, historical vehicle track data of the vehicle driving through the target intersection in a historical detection time period before the time period to be detected, and intersection information of the target intersection.

Here, the implementation of S1601 is similar to S1201 described above, and is not described here again.

S1602: and determining whether the target intersection is in a low-efficiency passing state or not according to the target vehicle track data and the reference speed of the target intersection. If so, jump to S1603; if not, jumping to S1605.

The specific way of determining the reference speed of the target intersection is similar to the way of determining the reference speed of the target intersection in fig. 7, and is not described herein again.

The specific manner of determining whether the target intersection is in the low-efficiency passing state according to the target vehicle trajectory data and the reference speed of the target intersection is similar to the embodiment corresponding to fig. 8, and is not described herein again.

S1603: generating a first characteristic value of the target intersection under the road condition characteristic according to the target vehicle track data, generating a second characteristic value of the target intersection under the road condition characteristic according to the historical vehicle track data, and generating a third characteristic value of the target intersection under the road characteristic according to the intersection information.

Here, the implementation of S1603 is similar to the implementation of S1202 described above and is not described here again.

S1604: and inputting the first characteristic value, the second characteristic value and the third characteristic value into a pre-trained traffic efficiency information detection model, and acquiring traffic efficiency information corresponding to the target intersection.

Here, the implementation of S1604 is similar to that of S1203 described above, and is not described herein again.

S1605: and (6) ending.

After acquiring target vehicle track data, historical vehicle track data and intersection information of a target intersection, determining whether the target intersection is in a low-efficiency passing state or not according to the target vehicle track data and the reference speed of the target intersection; after the target intersection is determined to be in the low-efficiency passing state, a first characteristic value, a second characteristic value and a third characteristic value are generated according to target vehicle track data, historical vehicle track data and intersection information, the first characteristic value, the second characteristic value and the third characteristic value are input into a pre-trained passing efficiency information detection model, and passing efficiency information corresponding to the target intersection is obtained, so that the situation that the low-efficiency passing state exists can be filtered out firstly, the corresponding passing efficiency information is determined only for the situation that the low-efficiency passing state exists, the data processing amount of the model is reduced, and computing resources are saved.

Based on the same inventive concept, the embodiment of the present application further provides a device for detecting traffic efficiency information corresponding to the method for detecting traffic efficiency information, and as the principle of solving the problem of the device in the embodiment of the present application is similar to the method for detecting traffic efficiency information in the embodiment of the present application, the implementation of the device can refer to the implementation of the method, and repeated details are not repeated.

EXAMPLE seven

Referring to fig. 17, a schematic diagram of an apparatus for detecting traffic efficiency information according to a seventh embodiment of the present application is shown, where the apparatus includes: a data acquisition module 171, a feature generation module 172, a result acquisition module 173; wherein:

the data acquisition module 171 is configured to acquire target vehicle trajectory data of a vehicle that has driven through a target intersection in a time period to be detected, and intersection information of the target intersection;

the characteristic generating module 172 is configured to generate a characteristic value of the target intersection under the traffic efficiency influence characteristic according to the target vehicle trajectory data and the intersection information;

the result obtaining module 173 is configured to input the feature value into a traffic efficiency information detection model trained in advance, and obtain traffic efficiency information corresponding to the target intersection.

In one possible embodiment, the traffic efficiency information includes information indicating traffic inefficiency and traffic efficiency information; or, information indicating the passage efficiency and the passage efficiency reason.

In one possible embodiment, the traffic efficiency affecting feature comprises: one or more of road condition characteristics, road characteristics, and time characteristics.

In a possible implementation manner, for a case that the traffic efficiency influence characteristics include road condition characteristics and road characteristics:

the data acquisition module 171 is further configured to: acquiring historical vehicle track data of the target intersection driven by a historical detection time period before the time period to be detected;

the feature generation module 172 is specifically configured to: generating a first characteristic value of the target intersection under the road condition characteristic according to the target vehicle track data, generating a second characteristic value of the target intersection under the road condition characteristic according to the historical vehicle track data, and generating a third characteristic value of the target intersection under the road characteristic according to the intersection information;

the result obtaining module 173 is specifically configured to: and inputting the first characteristic value, the second characteristic value and the third characteristic value into the pre-trained traffic efficiency information detection model to obtain traffic efficiency information corresponding to the target intersection.

In one possible embodiment, for a case where the traffic efficiency influence characteristics include road condition characteristics, road characteristics, and time characteristics:

the feature generation module 172 is further configured to: acquiring a fourth characteristic value of the target intersection under the time characteristic;

the result obtaining module 173 is specifically configured to: and inputting the first characteristic value, the second characteristic value, the third characteristic value and the fourth characteristic value into a pre-trained traffic efficiency information detection model, and acquiring traffic efficiency information corresponding to the target intersection.

In one possible embodiment, the road condition characteristics include: one or more of the speed of the vehicle flow in the intersection, the speed of the vehicle flow on the downstream road section of the intersection and the parking coverage rate on the downstream road section of the intersection.

In one possible embodiment, the road feature comprises: the bus stop comprises one or more of turning radius, intersection slope, intersection sight distance, the number of intersection stop lines, intersection exit lane width, the distance between a road entrance nearest to an intersection exit and an intersection, the distance between a bus station nearest to the intersection exit and the intersection, and the number of buses stopped at the bus station nearest to the intersection exit.

In one possible embodiment, the temporal characteristics include: the time period to be detected is one or more of a date corresponding to the time period to be detected, whether the time period to be detected is a traffic peak, the number of weeks corresponding to the time period to be detected, whether the time period to be detected is a holiday and the number of months corresponding to the time period to be detected.

In one possible embodiment, for a case where the road condition characteristics include vehicle flow speed in the intersection:

the feature generation module 172 is configured to acquire the speed of the vehicle flow in the intersection in the following manner:

In one possible embodiment, the target vehicle trajectory data includes: the position information of a plurality of positioning points and the positioning time corresponding to each positioning point;

the feature generation module 172 is configured to calculate a speed of each vehicle passing through the target intersection according to the target vehicle trajectory data of each vehicle in the following manner:

for each vehicle, performing:

the feature generation module 172 is configured to calculate an entry time and an exit time of each target vehicle passing through the target intersection according to target vehicle trajectory data of each vehicle in the following manner:

In one possible implementation, for the case that the road condition characteristics include the traffic flow speed of the road section downstream of the intersection,

the characteristic generating module 172 is configured to acquire the traffic flow speed of the downstream road segment of the intersection in the time period to be detected by using the following method:

In a possible embodiment, the method further comprises: a first model training module 174, configured to train to obtain the traffic efficiency information detection model in the following manner:

In a possible embodiment, the method further comprises: a second model training module 175, configured to train to obtain the traffic efficiency information detection model in the following manner:

In one possible embodiment, the base prediction model includes: a first neural network, a second neural network, and a third neural network;

the second model training module 175 is configured to input the first sample feature value, the second sample feature value, and the third sample feature value into a basic prediction model in the following manner, and obtain a sample passage inefficiency detection result corresponding to each of the sample intersections:

for each of the sample intersections, performing:

In a possible implementation, the second model training module 175 is further configured to: acquiring a fourth sample characteristic value under the time characteristic;

the second model training module 175 is configured to input the first sample feature value, the second sample feature value, the third sample feature value, and the fourth sample feature value into a basic prediction model in the following manner, and obtain a sample passage inefficiency detection result corresponding to each sample intersection:

for each of the sample intersections, performing:

In a possible embodiment, the method further comprises: the first traffic efficiency determining module 176 is configured to, before the feature generating module 172 generates the feature value of the target intersection under the traffic efficiency influence feature according to the target vehicle trajectory data and the intersection information:

the feature generation module 172 is specifically configured to: and after the target intersection is detected to be in a low-efficiency passing state, generating a characteristic value of the target intersection under the passing efficiency influence characteristic according to the target vehicle track data and the intersection information.

In one possible embodiment, the first traffic efficiency determining module 176 is specifically configured to determine whether the target intersection is in a traffic inefficiency state according to the target vehicle trajectory data and the reference speed of the target intersection in the following manner:

and if the speed difference between the reference speed and the speed in the intersection is detected to be larger than a preset first speed difference threshold value and the speed in the intersection is detected to be smaller than a preset first speed threshold value, or the reference speed is detected to be smaller than a preset second speed threshold value, determining that the target intersection is in a passing low-efficiency state.

In a possible embodiment, the device further comprises: an interior region determining module 177 for determining the intersection interior region of the target intersection according to target vehicle trajectory data of a plurality of vehicles driving through the target intersection;

the data obtaining module 171 is specifically configured to: and acquiring target vehicle track data of vehicles driving through the intersection inner area of the target intersection in the time period to be detected.

In a possible implementation, the first traffic efficiency determining module 176 is specifically configured to determine the reference speed by:

In a possible embodiment, the method further comprises: a second traffic efficiency determining module 178, configured to, before the feature generating module 172 generates the feature value of the target intersection under the traffic efficiency influence feature according to the target vehicle trajectory data and the intersection information:

the feature generation module 172 is specifically configured to: and after detecting that the target intersection is in a high-efficiency passing state, generating a characteristic value of the target intersection under the influence characteristic of the passing efficiency according to the target vehicle track data and the intersection information.

In one possible implementation, the second traffic efficiency determination module 178 is configured to determine whether the target intersection is in a traffic efficiency state according to the target vehicle trajectory data and the reference speed of the target intersection in the following manner:

judging whether the reference speed is greater than a preset fourth speed threshold value or not;

and if the speed difference between the reference speed and the speed in the intersection is detected to be smaller than a preset second speed difference threshold value and the speed in the intersection is detected to be larger than a preset third speed threshold value, or the reference speed is detected to be larger than a preset fourth speed threshold value, determining that the target intersection is in a high-efficiency passing state.

In a possible implementation manner, the data obtaining module 171 is specifically configured to: acquiring target vehicle track data of vehicles driving in the road direction in a time period to be detected aiming at each road direction of the target intersection;

the feature generation module 172 is specifically configured to: generating characteristic values of all road directions under the traffic efficiency influence characteristic according to target vehicle track data of vehicles driving through all road directions in a time period to be detected;

the result obtaining module 173 is specifically configured to: and inputting the characteristic value of each road direction under the traffic efficiency influence characteristic into a traffic efficiency information detection model trained in advance, and acquiring traffic efficiency information corresponding to each road direction.

The description of the processing flow of each module in the device and the interaction flow between the modules may refer to the related description in the above method embodiments, and will not be described in detail here.

Example eight

An eighth embodiment of the present application further provides a computer device 180, as shown in fig. 18, which is a schematic structural diagram of the computer device 180 provided in the embodiment of the present application, and includes:

processor 181, memory 182, and bus 183; the memory 182 is used for storing instructions for execution and includes a memory 1821 and an external memory 1822; the memory 1821 is also referred to as an internal memory, and is configured to temporarily store operation data in the processor 181 and data exchanged with an external memory 1822 such as a hard disk, the processor 181 exchanges data with the external memory 1822 through the memory 1821, and when the user device 180 is operated, the processor 181 communicates with the memory 182 through the bus 183, so that the processor 181 executes the following instructions in a user mode:

In one possible implementation, the processor 181 executes instructions in which the traffic efficiency information includes information indicating that traffic is inefficient and traffic efficiency information; or, information indicating the passage efficiency and the passage efficiency reason.

In a possible implementation, the traffic efficiency affecting feature in the instructions executed by the processor 181 includes: one or more of road condition characteristics, road characteristics, and time characteristics.

In a possible implementation manner, the processor 181 executes instructions, and for a case that the traffic efficiency influence characteristic includes a road condition characteristic and a road characteristic, the method further includes:

In a possible embodiment, the processor 181 executes instructions to determine, for the case where the traffic efficiency influence characteristics include road condition characteristics, road characteristics and time characteristics,

In one possible embodiment, the processor 181 executes instructions to perform the following road condition features: one or more of the speed of the vehicle flow in the intersection, the speed of the vehicle flow on the downstream road section of the intersection and the parking coverage rate on the downstream road section of the intersection.

In a possible implementation, the processor 181 executes instructions that include, for the road feature: the bus stop comprises one or more of turning radius, intersection slope, intersection sight distance, the number of intersection stop lines, intersection exit lane width, the distance between a road entrance nearest to an intersection exit and an intersection, the distance between a bus station nearest to the intersection exit and the intersection, and the number of buses stopped at the bus station nearest to the intersection exit.

In a possible implementation, in the instructions executed by the processor 181, the time characteristic includes: the time period to be detected is one or more of a date corresponding to the time period to be detected, whether the time period to be detected is a traffic peak, the number of weeks corresponding to the time period to be detected, whether the time period to be detected is a holiday and the number of months corresponding to the time period to be detected.

In one possible embodiment, the processor 181 executes instructions to, for a case that the road condition characteristic includes a speed of a vehicle flow in the intersection:

In one possible implementation, the processor 181 executes instructions that include the target vehicle trajectory data: the position information of a plurality of positioning points and the positioning time corresponding to each positioning point;

for each vehicle, performing:

In one possible embodiment, the processor 181 executes instructions to determine, for the case that the road condition characteristic includes the traffic speed of the road segment downstream of the intersection,

the traffic flow speed of the downstream road section of the intersection in the time period to be detected is obtained by adopting the following method:

In one possible embodiment, the instructions executed by the processor 181 train the traffic efficiency information detection model to be obtained by:

In a possible implementation, in the instructions executed by the processor 181, the basic prediction model includes: a first neural network, a second neural network, and a third neural network;

for each of the sample intersections, performing:

In a possible implementation manner, the instructions executed by the processor 181 further include: acquiring a fourth sample characteristic value under the time characteristic;

for each of the sample intersections, performing:

In one possible embodiment, before the step of generating the characteristic value of the target intersection under the traffic efficiency influence characteristic according to the target vehicle trajectory data and the intersection information, the step executed by the processor 181 further includes:

In one possible embodiment, the processor 181 executes instructions for determining whether the target intersection is in a traffic inefficiency state according to the target vehicle trajectory data and the reference speed of the target intersection, including:

In a possible implementation manner, the instructions executed by the processor 181 further include: determining the intersection internal area of the target intersection according to target vehicle track data of a plurality of vehicles driving through the target intersection;

In one possible embodiment, the processor 181 executes instructions to determine the reference speed by:

In one possible embodiment, the processor 181 executes instructions for determining whether the target intersection is in a traffic-efficient state according to the target vehicle trajectory data and the reference speed of the target intersection, including:

In a possible implementation, the instructions executed by the processor 181 for obtaining target vehicle trajectory data of a vehicle that has traveled through a target intersection in a time period to be detected includes:

In addition, an embodiment of the present application further provides a computer-readable storage medium, where a computer program is stored on the computer-readable storage medium, and when the computer program is executed by a processor, the computer program performs the steps of the method for detecting traffic efficiency information described in the above method embodiment.

The method for detecting traffic efficiency information and the computer program product of the apparatus provided in the embodiment of the present application include a computer-readable storage medium storing a program code, where instructions included in the program code may be used to execute the method described in the foregoing method embodiment, and specific implementation may refer to the method embodiment, and will not be described herein again.

It is clear to those skilled in the art that, for convenience and brevity of description, the specific working processes of the system and the apparatus described above may refer to the corresponding processes in the foregoing method embodiments, and are not described herein again. In the several embodiments provided in the present application, it should be understood that the disclosed system, apparatus and method may be implemented in other ways. The above-described embodiments of the apparatus are merely illustrative, and for example, the division of the units is only one logical division, and there may be other divisions when actually implemented, and for example, a plurality of units or components may be combined or integrated into another system, or some features may be omitted, or not executed. In addition, the shown or discussed mutual coupling or direct coupling or communication connection may be an indirect coupling or communication connection of devices or units through some communication interfaces, and may be in an electrical, mechanical or other form.

The units described as separate parts may or may not be physically separate, and parts displayed as units may or may not be physical units, may be located in one place, or may be distributed on a plurality of network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of the embodiment.

In addition, functional units in the embodiments of the present application may be integrated into one processing unit, or each unit may exist alone physically, or two or more units are integrated into one unit.

The functions, if implemented in the form of software functional units and sold or used as a stand-alone product, may be stored in a non-volatile computer-readable storage medium executable by a processor. Based on such understanding, the technical solution of the present application or portions thereof that substantially contribute to the prior art may be embodied in the form of a software product stored in a storage medium and including instructions for causing a computer device (which may be a personal computer, a server, or a network device) to execute all or part of the steps of the method according to the embodiments of the present application. And the aforementioned storage medium includes: various media capable of storing program codes, such as a usb disk, a removable hard disk, a Read-only Memory (ROM), a Random Access Memory (RAM), a magnetic disk, or an optical disk.

Finally, it should be noted that: the above-mentioned embodiments are only specific embodiments of the present application, and are used for illustrating the technical solutions of the present application, but not limiting the same, and the scope of the present application is not limited thereto, and although the present application is described in detail with reference to the foregoing embodiments, those skilled in the art should understand that: any person skilled in the art can modify or easily conceive the technical solutions described in the foregoing embodiments or equivalent substitutes for some technical features within the technical scope disclosed in the present application; such modifications, changes or substitutions do not depart from the spirit and scope of the exemplary embodiments of the present application, and are intended to be covered by the scope of the present application. Therefore, the protection scope of the present application shall be subject to the protection scope of the claims.

Claims

1. A method for traffic efficiency information detection, comprising:

2. The method of claim 1, wherein the traffic efficiency information comprises information indicating a traffic inefficiency and a reason for the traffic inefficiency; or, information indicating the passage efficiency and the passage efficiency reason.

3. The method of claim 1, wherein the traffic efficiency impact characteristic comprises: one or more of road condition characteristics, road characteristics, and time characteristics.

4. The method of claim 3, wherein for the case that the traffic efficiency impact characteristics include road condition characteristics, road characteristics, the method further comprises:

5. The method of claim 4, wherein for the case where the traffic efficiency impact characteristics include road condition characteristics, road characteristics, and time characteristics,

6. The method of claim 3, wherein the road condition characteristics comprise: one or more of the speed of the vehicle flow in the intersection, the speed of the vehicle flow on the downstream road section of the intersection and the parking coverage rate on the downstream road section of the intersection.

7. The method of claim 3, wherein the road feature comprises: the bus stop comprises one or more of turning radius, intersection slope, intersection sight distance, the number of intersection stop lines, intersection exit lane width, the distance between a road entrance nearest to an intersection exit and an intersection, the distance between a bus station nearest to the intersection exit and the intersection, and the number of buses stopped at the bus station nearest to the intersection exit.

8. The method of claim 3, wherein the temporal characteristics comprise: the time period to be detected is one or more of a date corresponding to the time period to be detected, whether the time period to be detected is a traffic peak, the number of weeks corresponding to the time period to be detected, whether the time period to be detected is a holiday and the number of months corresponding to the time period to be detected.

9. The method of claim 6, wherein for the case that the road condition characteristic comprises a speed of vehicle flow within the intersection:

10. The method of claim 9, wherein the target vehicle trajectory data comprises: the position information of a plurality of positioning points and the positioning time corresponding to each positioning point;

for each vehicle, performing:

11. The method of claim 9, wherein the target vehicle trajectory data comprises: the position information of a plurality of positioning points and the positioning time corresponding to each positioning point;

12. The method according to claim 6, wherein the characteristic of the road condition comprises: the condition of the traffic flow speed of the road section downstream of the intersection,

13. The method of claim 1, wherein the traffic efficiency information detection model is trained by:

14. The method of claim 4, wherein the traffic efficiency information detection model is trained by:

15. The method of claim 14, wherein the base predictive model comprises: a first neural network, a second neural network, and a third neural network;

for each of the sample intersections, performing:

16. The method of claim 14, further comprising: acquiring a fourth sample characteristic value under the time characteristic;

inputting the first sample characteristic value, the second sample characteristic value, the third sample characteristic value and the fourth sample characteristic value into the basic prediction model, and obtaining a sample passing low-efficiency detection result corresponding to each sample intersection.

17. The method of claim 16, wherein the base predictive model comprises: a first neural network, a second neural network, and a third neural network;

for each of the sample intersections, performing:

18. The method according to claim 1, wherein before generating the characteristic value of the target intersection under the traffic efficiency influence characteristic according to the target vehicle trajectory data and the intersection information, further comprising:

19. The method of claim 18, wherein said determining whether said target intersection is in a traffic inefficiency state based on said target vehicle trajectory data and a reference speed of said target intersection comprises:

20. The method of claim 1 or 19, further comprising: determining an intersection internal area of the target intersection according to target vehicle track data of a plurality of vehicles driving through the target intersection;

21. The method of claim 18, wherein the reference speed is determined by:

22. The method according to claim 1, wherein before generating the characteristic value of the target intersection under the traffic efficiency influence characteristic according to the target vehicle trajectory data and the intersection information, further comprising:

23. The method of claim 22, wherein determining whether the target intersection is in a traffic-efficient state based on the target vehicle trajectory data and a reference speed of the target intersection comprises:

24. The method of claim 1, wherein the obtaining target vehicle trajectory data for vehicles driving through a target intersection for a time period to be detected comprises:

25. An apparatus for traffic efficiency information detection, comprising:

26. The apparatus of claim 25, wherein the traffic efficiency information comprises information indicating a traffic inefficiency and a reason for the traffic inefficiency; or, information indicating the passage efficiency and the passage efficiency reason.

27. The apparatus of claim 25, wherein the traffic efficiency impact characteristic comprises: one or more of road condition characteristics, road characteristics, and time characteristics.

28. The apparatus of claim 27, wherein for the traffic efficiency impact characteristics comprising road condition characteristics, road characteristics,

29. The apparatus of claim 28, wherein for the case that the traffic efficiency impact characteristics include road condition characteristics, road characteristics and time characteristics,

30. The apparatus of claim 27, wherein the road condition characteristics comprise: one or more of the speed of the vehicle flow in the intersection, the speed of the vehicle flow on the downstream road section of the intersection and the parking coverage rate on the downstream road section of the intersection.

31. The apparatus of claim 27, wherein the road feature comprises: the bus stop comprises one or more of turning radius, intersection slope, intersection sight distance, the number of intersection stop lines, intersection exit lane width, the distance between a road entrance nearest to an intersection exit and an intersection, the distance between a bus station nearest to the intersection exit and the intersection, and the number of buses stopped at the bus station nearest to the intersection exit.

32. The apparatus of claim 27, wherein the temporal characteristics comprise: the time period to be detected is one or more of a date corresponding to the time period to be detected, whether the time period to be detected is a traffic peak, the number of weeks corresponding to the time period to be detected, whether the time period to be detected is a holiday and the number of months corresponding to the time period to be detected.

33. The apparatus of claim 30, wherein for the condition that the road condition characteristic comprises a speed of vehicle flow within the intersection:

34. The apparatus of claim 33, wherein the target vehicle trajectory data comprises: the position information of a plurality of positioning points and the positioning time corresponding to each positioning point;

for each vehicle, performing:

35. The apparatus of claim 33, wherein the target vehicle trajectory data comprises: the position information of a plurality of positioning points and the positioning time corresponding to each positioning point;

36. The apparatus of claim 30, wherein for the condition that the road condition characteristic comprises traffic flow speed of a road segment downstream of the intersection,

37. The apparatus of claim 25, further comprising: the first model training module is used for training in the following way to obtain the traffic efficiency information detection model:

38. The apparatus of claim 28, further comprising: the second model training module is used for training to obtain the traffic efficiency information detection model by adopting the following modes:

39. The apparatus of claim 38, wherein the base predictive model comprises: a first neural network, a second neural network, and a third neural network;

for each of the sample intersections, performing:

40. The apparatus of claim 38, wherein the second model training module is further configured to: acquiring a fourth sample characteristic value under the time characteristic;

41. The apparatus of claim 40, wherein the base predictive model comprises: a first neural network, a second neural network, and a third neural network;

for each of the sample intersections, performing:

42. The apparatus of claim 25, further comprising: a first traffic efficiency determination module, configured to, before the feature generation module generates the feature value of the target intersection under the traffic efficiency influence feature according to the target vehicle trajectory data and the intersection information:

43. The apparatus of claim 42, wherein the first traffic efficiency determination module is specifically configured to determine whether the target intersection is in a traffic inefficiency state based on the target vehicle trajectory data and the reference speed of the target intersection by:

44. The apparatus of claim 25 or 43, further comprising: the internal area determining module is used for determining the internal area of the target intersection according to the target vehicle track data of a plurality of vehicles driving through the target intersection;

45. The apparatus of claim 42, wherein the first traffic efficiency determining module is specifically configured to determine the reference speed by:

46. The apparatus of claim 25, further comprising: a second traffic efficiency determination module, configured to, before the feature generation module generates the feature value of the target intersection under the traffic efficiency influence feature according to the target vehicle trajectory data and the intersection information:

47. The apparatus of claim 46, wherein the second traffic efficiency determination module is configured to determine whether the target intersection is in a traffic-efficient state based on the target vehicle trajectory data and the reference speed of the target intersection by:

48. The apparatus of claim 25, wherein the data acquisition module is specifically configured to: acquiring target vehicle track data of vehicles driving in the road direction in a time period to be detected aiming at each road direction of the target intersection;

49. An electronic device, comprising: a processor, a memory and a bus, the memory storing machine-readable instructions executable by the processor, the processor and the memory communicating via the bus when the electronic device is operating, the machine-readable instructions when executed by the processor performing the steps of the method of traffic efficiency information detection as claimed in any one of claims 1 to 24.

50. A computer-readable storage medium, having stored thereon a computer program which, when being executed by a processor, carries out the steps of the method for detecting traffic efficiency information according to any one of claims 1 to 24.