CN117711207A - Method and system for early warning collision of meeting vehicles at curve intersection - Google Patents

Abstract

The invention relates to the field of intelligent traffic systems, in particular to a method and a system for early warning collision of meeting vehicles at a curve intersection, comprising the following steps: collecting a plurality of vehicle operation data sequences and the environmental interference degree of a curve intersection; obtaining a vehicle state disturbance factor according to a vehicle operation data sequence; obtaining collision risk factors according to the vehicle state disturbance factors; obtaining an abnormal factor according to the environmental disturbance degree of the curve intersection; acquiring a plurality of comparison operation data; obtaining a relative noise factor according to the comparison operation data and the collision risk factor; obtaining a predicted collision coefficient according to the relative noise factor and the anomaly factor; obtaining a vehicle collision state estimated value according to the predicted collision coefficient; and early warning is carried out according to the estimated value of the collision state of the vehicle. The invention reduces the possibility of noise mistakenly considered by normal vehicle operation data and improves the accuracy of vehicle collision early warning.

Description

Method and system for early warning collision of meeting vehicles at curve intersection

Technical Field

The invention relates to the field of intelligent traffic systems, in particular to a method and a system for early warning collision at a curve intersection.

Background

The curve intersection is taken as an important node of the traffic network and is a high-rise place of traffic safety accidents; in order to reduce potential safety hazards when a vehicle runs at a curve intersection, the vehicle traffic at the curve intersection is more intelligent; the vehicle running data of the vehicle-mounted sensor in the vehicle is generally transmitted to a road side unit RSU of the curve intersection, the probability of collision of the vehicle is predicted through the road side unit RSU, and corresponding collision early warning is carried out.

The traditional method generally uses the extended Kalman filtering to perform collision early warning on the vehicle operation data, but the vehicle operation data can be changed rapidly due to the actual driving situation, more abrupt vehicle operation data exist, and the traditional extended Kalman filtering can mistakes part of abrupt normal vehicle operation data into noise data, so that the accuracy of the vehicle collision early warning is reduced.

Disclosure of Invention

The invention provides a method and a system for early warning collision of meeting vehicles at a curve intersection, which aim to solve the existing problems: the vehicle operation data can be changed rapidly according to the actual driving condition, more abrupt vehicle operation data exist, the traditional extended Kalman filtering can not effectively distinguish the abrupt vehicle operation data and noise, and the accuracy of vehicle collision early warning is reduced.

The invention relates to a method and a system for early warning collision of vehicles meeting at a curve intersection, which adopts the following technical scheme:

the embodiment of the invention provides a collision early warning method for meeting vehicles at a curve intersection, which comprises the following steps:

collecting a plurality of vehicle operation data sequences of a target vehicle, a plurality of vehicle operation data sequences of a plurality of historical vehicles and the environment interference degree of a curve intersection, wherein the vehicle operation data sequences comprise a plurality of vehicle operation data;

Obtaining a vehicle state disturbance factor of each vehicle operation data sequence according to the numerical value difference of the vehicle operation data between the vehicle operation data sequences of the target vehicle and the historical vehicle, wherein the vehicle state disturbance factor is used for describing the probability that the vehicle operation data in the vehicle operation data sequence is erroneously recognized as noise; obtaining collision risk factors of each vehicle operation data sequence according to the difference of the vehicle state disorder factors among the vehicle operation data sequences;

obtaining an abnormal factor of each vehicle operation data of the target vehicle according to the environmental disturbance degree of the curve intersection and the change difference of the vehicle operation data of the target vehicle between adjacent positions, wherein the abnormal factor is used for describing the probability that the vehicle operation data is influenced by the operation of a driver; for any one vehicle operation data in any one vehicle operation data sequence, screening a plurality of comparison operation data of each vehicle operation data from the vehicle operation data sequence; according to the distribution discrete condition of the comparison operation data and the collision risk factor in the vehicle operation data sequence of the target vehicle, obtaining the relative noise factor of each vehicle operation data of the target vehicle, wherein the relative noise factor is used for describing the initial difference between the vehicle operation data and the normal vehicle operation data;

Obtaining a predicted collision coefficient of each vehicle operation data of the target vehicle according to the relative noise factor and the anomaly factor, wherein the predicted collision coefficient is used for describing the final difference between the vehicle operation data and the normal vehicle operation data; obtaining a vehicle collision state estimated value of the target vehicle according to the predicted collision coefficient, wherein the vehicle collision state estimated value is used for describing an error of a collision result of the target vehicle; and pre-warning the target vehicle according to the estimated value of the collision state of the vehicle.

Preferably, the method for obtaining the vehicle state disturbance factor of each vehicle operation data sequence according to the numerical value difference of the vehicle operation data between the vehicle operation data sequences of the target vehicle and the historical vehicle comprises the following specific steps:

marking the target vehicle and the historical vehicle as one reference vehicle, and fitting any vehicle operation data sequence of any reference vehicle by using a least square method to obtain a fitting curve;

for any one vehicle operation data in the vehicle operation data sequence, marking the numerical value of the vehicle operation data on the fitting curve as original smooth operation data of the vehicle operation data, marking the absolute value of the difference value between the original smooth operation data and the vehicle operation data as an original operation data difference value of the vehicle operation data, acquiring the original operation data difference values of all the vehicle operation data in the vehicle operation data sequence, carrying out linear normalization on all the original operation data difference values, and marking each normalized original operation data difference value as a smooth operation difference value;

Wherein, beta represents an initial vehicle state disturbance factor of any one vehicle operation data sequence of any one reference vehicle; n represents the number of all vehicle operation data in the vehicle operation data sequence; a, a _n Representing nth vehicle operation data in the sequence of vehicle operation data;representing a mean value of all vehicle operation data in the sequence of vehicle operation data; Δa1 _n A smooth running difference value representing nth vehicle running data in the vehicle running data sequence; the absolute value is taken; obtaining initial vehicle state disturbance factors of all vehicle operation data sequences of the reference vehicles, obtaining initial vehicle state disturbance factors of all vehicle operation data sequences of all reference vehicles, carrying out linear normalization on all initial vehicle state disturbance factors, and recording the normalized initial vehicle state disturbance factors as vehicle state disturbance factors.

Preferably, the collision risk factor of each vehicle operation data sequence is obtained according to the difference of the vehicle state disorder factors between the vehicle operation data sequences, and the specific method comprises the following steps:

in the formula, beta 3 _z A collision risk factor representing a z-th vehicle operation data sequence of the target vehicle; beta 2 _z A vehicle state disturbance factor representing a z-th vehicle operation data sequence of the target vehicle;a vehicle condition disturbance factor representing a z-th vehicle operation data sequence of all historical vehicles.

Preferably, the method for obtaining the anomaly factor of each vehicle operation data of the target vehicle according to the environmental disturbance degree of the curve intersection and the variation difference of the vehicle operation data of the target vehicle between adjacent positions includes the following specific steps:

performing DTW matching on the jth vehicle operation data sequence of the target vehicle and the jth vehicle operation data sequence of the ith historical vehicle to obtain a DTW distance, and recording the DTW distance as the sequence data similarity of the jth vehicle operation data sequence of the target vehicle and the jth vehicle operation data sequence of the ith historical vehicle;

the method comprises the steps of recording w-th vehicle operation data in a j-th vehicle operation data sequence of a target vehicle as target data, recording a vehicle operation data sequence with the type of the vehicle operation data as position data as a vehicle distance operation data sequence, and recording w-th vehicle operation data in the vehicle distance operation data sequence of the target vehicle as target position data of the target data;

in the vehicle distance operation data sequence of the ith historical vehicle, the vehicle operation data with the smallest absolute value of the difference value with the target position data is recorded as the minimum vehicle distance operation data of the target position data in the vehicle distance operation data sequence of the ith historical vehicle; the absolute value of the difference value between the target position data and the minimum vehicle distance operation data is recorded as a vehicle distance difference value of the target data; recording the corresponding vehicle operation data of the minimum vehicle distance operation data in the j-th vehicle operation data sequence of the i-th historical vehicle as final reference operation data of the target data in the j-th vehicle operation data sequence of the i-th historical vehicle;

In the formula g _j,w An initial anomaly factor representing the w-th vehicle operation data in the j-th vehicle operation data sequence of the target vehicle; i represents all history vehiclesThe number of vehicles; a2 _j,w The jth vehicle operation data sequence representing the target vehicle; a is that _i,j,w Representing the jth vehicle operation data sequence of the target vehicle, and final reference operation data in the jth vehicle operation data sequence of the ith historical vehicle; b (B) _i,j,w A vehicle distance difference value representing the w-th vehicle operation data in the j-th vehicle operation data sequence of the target vehicle; d, d _i,j A sequence data similarity representing a jth vehicle operation data sequence of the target vehicle and a jth vehicle operation data sequence of the ith historical vehicle; τ represents the environmental disturbance degree of the curve intersection; the absolute value is taken; exp () represents an exponential function based on a natural constant; acquiring initial abnormal factors of all vehicle operation data in all vehicle operation data sequences of the target vehicle, carrying out linear normalization on all the initial abnormal factors, and recording each normalized initial abnormal factor as an abnormal factor.

Preferably, the screening a plurality of comparison operation data of each vehicle operation data from the vehicle operation data sequence includes the following specific methods:

Presetting a vehicle operation data quantity T1, and recording a j 1-th vehicle operation data sequence of a target vehicle as a first target sequence; the vehicle operation data sequences except the first target sequence are marked as first reference sequences in all vehicle operation data sequences of the target vehicle; in the first target sequence, the T1 th vehicle operation data before the w1 st vehicle operation data and the T1 st vehicle operation data after the w1 st vehicle operation data are recorded as comparison operation data of the w1 st vehicle operation data in the first target sequence.

Preferably, the method for obtaining the relative noise factor of each vehicle operation data of the target vehicle according to the distribution discrete condition of the comparison operation data and the collision risk factor in the vehicle operation data sequence of the target vehicle includes the following specific steps:

wherein L is _j1,w1 An initial relative noise factor representing w1 st vehicle operation data in the j1 st vehicle operation data sequence of the target vehicle; v (V) _j1 A number of all first reference sequences representing a j1 st vehicle operation data sequence of the target vehicle; epsilon _j1,w1 Standard deviation of all the collation operation data representing the w1 st vehicle operation data in the j1 st vehicle operation data sequence of the target vehicle; epsilon _j1,v,w1 Standard deviations of all the collation operation data of the w1 st vehicle operation data in the v first reference sequence representing the j1 st vehicle operation data sequence of the target vehicle; beta 3 _j1 A collision risk factor representing a j1 st vehicle operation data sequence of the target vehicle; beta 3 _j1,v A collision risk factor representing a v first reference sequence of a j1 st vehicle operation data sequence of the target vehicle; exp () represents an exponential function based on a natural constant; acquiring initial relative noise factors of all vehicle operation data in all vehicle operation data sequences of the target vehicle, carrying out linear normalization on all initial relative noise factors, and recording each normalized initial relative noise factor as a relative noise factor.

Preferably, the method for obtaining the predicted collision coefficient of each vehicle operation data of the target vehicle according to the relative noise factor and the anomaly factor includes the following specific steps:

f _j2,w2 ＝1-(g1 _j2,w2 ×L _j2,w2 )

wherein f _j2,w2 A predicted collision coefficient indicating w 2-th vehicle operation data in the j 2-th vehicle operation data sequence of the target vehicle; g1 _j2,w2 An abnormality factor representing w 2-th vehicle operation data in a j 2-th vehicle operation data sequence of the target vehicle; l (L) _j2,w2 Representing the relative noise factor of the w2 nd vehicle operation data in the j2 nd vehicle operation data sequence of the target vehicle.

Preferably, the method for obtaining the estimated value of the collision state of the target vehicle according to the predicted collision coefficient includes the following specific steps:

the method comprises the steps of recording a predicted value obtained by expanding Kalman filtering on w3 vehicle operation data in a j3 vehicle operation data sequence of a target vehicle as a traditional predicted value, and recording an absolute value of a difference value between the traditional predicted value and the w3 vehicle operation data as a comparison filtering value of the w3 vehicle operation data;

wherein P represents a vehicle collision state estimation value of the target vehicle; y represents the number of all vehicle operation data sequences of the target vehicle; r is R _j3 Representing the number of all vehicle operation data in the j3 rd vehicle operation data sequence of the target vehicle; omega _j3,w3 A comparison filter value representing w3 vehicle operation data in the j3 rd vehicle operation data sequence of the target vehicle; f (f) _j3,w3 A predicted collision coefficient representing the w3 rd vehicle operation data in the j3 rd vehicle operation data sequence of the target vehicle.

Preferably, the method for early warning the target vehicle according to the estimated value of the collision state of the vehicle includes the following specific steps:

replacing the state estimation value of the extended Kalman filtering algorithm with the vehicle collision state estimation value of the target vehicle, and marking the vehicle collision state estimation value as a prediction error covariance matrix according to all vehicle operation data of the target vehicle and the vehicle collision state estimation value; and carrying out iterative optimization of a prediction model by taking the vehicle collision prediction matrix as a prediction error covariance matrix, carrying out vehicle track prediction by taking model parameters corresponding to the minimum error matrix as an optimal Kalman filtering model, and carrying out vehicle collision early warning.

The invention also provides a collision early warning system for the vehicles meeting at the curve intersection, which comprises a memory and a processor, wherein the processor executes a computer program stored in the memory so as to realize the steps of the collision early warning method for the vehicles meeting at the curve intersection.

The technical scheme of the invention has the beneficial effects that: obtaining a vehicle state disturbance factor according to the numerical value difference of vehicle operation data among the vehicle operation data sequences, obtaining a collision risk factor according to the vehicle state disturbance factor, obtaining an abnormal factor according to the change difference of the vehicle operation data among adjacent positions, obtaining a relative noise factor according to the collision risk factor, obtaining a predicted collision coefficient according to the relative noise factor and the abnormal factor, obtaining a vehicle collision state estimated value according to the predicted collision coefficient, and carrying out early warning according to the vehicle collision state estimated value; the vehicle state disturbance factor reflects the probability that the vehicle operation data in the vehicle operation data sequence are erroneously identified as noise, the abnormality factor reflects the probability that the vehicle operation data are influenced by the operation of a driver, the relative noise factor reflects the initial difference between the vehicle operation data and the normal vehicle operation data, the predicted collision coefficient reflects the final difference between the vehicle operation data and the normal vehicle operation data, and the vehicle collision state estimated value reflects the error of the collision result of the target vehicle; the possibility that normal vehicle operation data is mistaken for noise is reduced, the denoising effect is improved, and the accuracy of vehicle collision early warning is improved.

Drawings

In order to more clearly illustrate the embodiments of the invention or the technical solutions in the prior art, the drawings that are required in the embodiments or the description of the prior art will be briefly described, it being obvious that the drawings in the following description are only some embodiments of the invention, and that other drawings may be obtained according to these drawings without inventive effort for a person skilled in the art.

FIG. 1 is a flow chart showing the steps of a method for early warning a collision at a curve intersection.

Detailed Description

In order to further describe the technical means and effects adopted by the invention to achieve the preset aim, the following detailed description is given below of a method and a system for early warning of collision at a curve intersection according to the invention, which are specific embodiments, structures, features and effects thereof, with reference to the accompanying drawings and preferred embodiments. In the following description, different "one embodiment" or "another embodiment" means that the embodiments are not necessarily the same. Furthermore, the particular features, structures, or characteristics of one or more embodiments may be combined in any suitable manner.

Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.

The invention provides a method and a system for early warning of collision at a curve crossing.

Referring to fig. 1, a step flow chart of a method for early warning of collision at a curve intersection according to an embodiment of the invention is shown, and the method includes the following steps:

step S001: and collecting a plurality of vehicle operation data sequences of the target vehicle, a plurality of vehicle operation data sequences of a plurality of historical vehicles and the environment interference degree of the curve intersection.

It should be noted that, in the conventional method, collision early warning is usually performed on vehicle operation data by using an extended kalman filter, but the vehicle operation data can be changed rapidly due to actual driving conditions, and more abrupt vehicle operation data exist, and the conventional extended kalman filter can mistakes part of abrupt normal vehicle operation data into noise data, so that the accuracy of vehicle collision early warning is reduced. Therefore, the embodiment provides a collision early warning method for meeting at a curve intersection.

Specifically, in order to implement the method for early warning a collision between vehicles at a curve intersection provided in this embodiment, firstly, a plurality of vehicle operation data sequences of a target vehicle, a plurality of vehicle operation data sequences of a plurality of historical vehicles, and an environmental interference degree at the curve intersection need to be collected, and the specific process is as follows: taking a vehicle provided with an on-vehicle sensor as a target vehicle, parking the target vehicle at a place with the road length of 300 meters from a curve intersection provided with a road side unit RSU, starting the target vehicle, simultaneously establishing communication between the target vehicle and the road side unit RSU, and enabling the target vehicle to freely drive towards the curve intersection; the road side unit records four vehicle operation data types, namely speed data, pressure data, acceleration data and position data, which are respectively communicated and transmitted with the target vehicle once every 2 seconds until the vehicle leaves the maximum communication distance Q of the road side unit, and stops collecting the vehicle operation data types; recording the speed data, the pressure data, the acceleration data and the position data recorded each time as vehicle running data; taking any kind of vehicle operation data as an example, arranging all vehicle operation data of the vehicle operation data according to the recording time from small to large, and recording the arranged sequence as a vehicle operation data sequence; a plurality of vehicle operation data sequences of the vehicle are acquired. The maximum communication distance Q of the road side unit RSU adopted in this embodiment is 300 meters, and each vehicle operation data sequence corresponds to one vehicle operation data type.

Further, acquiring a plurality of vehicle operation data of a plurality of vehicles in the near week from a database of the road side unit RSU, and recording the vehicles as historical vehicles; the method for acquiring the plurality of vehicle operation data sequences of the reference target vehicle obtains a plurality of vehicle operation data sequences of a plurality of historical vehicles.

Further, acquiring a plurality of meteorological data of four meteorological data types, namely temperature, humidity, wind speed and rainfall in a week in a database of a road side unit RSU; taking any one kind of meteorological data as an example, carrying out linear normalization on all meteorological data of the meteorological data kind, recording the normalized meteorological data as initial meteorological data, and recording the average value of all meteorological data of the meteorological data kind as initial characterization environmental data of the meteorological data kind; and acquiring initial representation environment data of all meteorological data types, and recording the average value of the initial representation environment data of all meteorological data types as the environmental interference degree of the curve intersection.

So far, a plurality of vehicle operation data sequences of the target vehicle, a plurality of vehicle operation data sequences of a plurality of historical vehicles and the environment interference degree of the curve crossing are obtained through the method.

Step S002: obtaining a vehicle state disturbance factor of each vehicle operation data sequence according to the numerical value difference of the vehicle operation data between the vehicle operation data sequences of the target vehicle and the historical vehicle; and obtaining collision risk factors of each vehicle operation data sequence according to the difference of the vehicle state disorder factors among the vehicle operation data sequences.

In the process of driving the vehicle to the curve intersection, the curve intersection can shield part of road surface conditions, and under normal conditions, in order to ensure the safety of the vehicle when passing through the curve intersection, the vehicle speed can change in a gentle trend; however, in the actual driving situation, there may be some disturbance factors such as uneven or smoother road surface, and these disturbance factors may damage the gentle trend of the original speed change to some extent; according to the embodiment, the vehicle state disturbance factor is obtained through the variation difference between the vehicle operation data in the vehicle operation data sequence, and the collision risk factor is obtained according to the local vehicle state disturbance factor and the overall vehicle state disturbance factor variation condition, so that subsequent analysis and processing can be facilitated.

Specifically, the target vehicle and the history vehicle are marked as a reference vehicle, any one vehicle operation data sequence of any one reference vehicle is taken as an example, and the least square method is utilized to fit the vehicle operation data sequence to obtain a fitting curve; taking any one vehicle operation data in the vehicle operation data sequence as an example, marking the numerical value of the vehicle operation data on the fitting curve as original smooth operation data of the vehicle operation data, marking the absolute value of the difference value between the original smooth operation data and the vehicle operation data as an original operation data difference value of the vehicle operation data, acquiring the original operation data difference values of all the vehicle operation data in the vehicle operation data sequence, carrying out linear normalization on all the original operation data difference values, and marking each normalized original operation data difference value as a smooth operation difference value. The least square method is a known technique, and the description of this embodiment is omitted.

Further, according to the smooth running difference values of all the vehicle running data in the vehicle running data sequence, an initial vehicle state disturbance factor of the vehicle running data sequence is obtained. The calculation method of the initial vehicle state disturbance factor of the vehicle operation data sequence comprises the following steps:

wherein, beta represents an initial vehicle state disturbance factor of the vehicle operation data sequence; n represents the number of all vehicle operation data in the vehicle operation data sequence; a, a _n Representing an nth vehicle operation data in the sequence of vehicle operation data;representing a mean value of all vehicle operation data in the sequence of vehicle operation data; Δa1 _n A smooth running difference value representing nth vehicle running data in the vehicle running data sequence; the absolute value is taken. The larger the disturbance factor of the initial vehicle state of the vehicle operation data sequence, the more irregular the change condition of the vehicle operation data type represented by the vehicle operation data sequence is, and the greater the possibility that the vehicle operation data in the vehicle operation data sequence is erroneously identified as noise is reflected. Obtaining initial vehicle state disturbance factors of all vehicle operation data sequences of the reference vehicles, obtaining initial vehicle state disturbance factors of all vehicle operation data sequences of all reference vehicles, carrying out linear normalization on all initial vehicle state disturbance factors, and recording the normalized initial vehicle state disturbance factors as vehicle state disturbance factors.

Further, according to the vehicle state disturbance factors of the z-th vehicle operation data sequence of all the historical vehicles, collision risk factors of the z-th vehicle operation data sequence of the target vehicle are obtained. The method for calculating the collision risk factor of the z-th vehicle operation data sequence of the target vehicle comprises the following steps:

in the formula, beta 3 _z A collision risk factor representing a z-th vehicle operation data sequence of the target vehicle; beta 2 _z A vehicle state disturbance factor representing a z-th vehicle operation data sequence of the target vehicle;a vehicle condition disturbance factor representing a z-th vehicle operation data sequence of all historical vehicles. If the collision risk factor of the z-th vehicle operation data sequence of the target vehicle is larger, it is indicated that the change rule of the vehicle operation data in the z-th vehicle operation data sequence of the target vehicle does not conform to the normal driving rule, and from the perspective of the z-th vehicle operation data sequence of the target vehicle, the more likely the target vehicle collides, and the greater the degree of noise interference is. And acquiring collision risk factors of all vehicle operation data sequences of the target vehicle.

So far, the collision risk factors of all vehicle operation data sequences of the target vehicle are obtained through the method.

Step S003: obtaining an abnormal factor of each vehicle running data of the target vehicle according to the environmental disturbance degree of the curve intersection and the variation difference of the vehicle running data of the target vehicle between adjacent positions; screening a plurality of comparison operation data of each vehicle operation data from the vehicle operation data sequence; and obtaining the relative noise factor of each vehicle operation data of the target vehicle according to the distribution discrete condition of the comparison operation data and the collision risk factor in the vehicle operation data sequence of the target vehicle.

It should be noted that, the collision risk factor is obtained based on the overall vehicle operation data sequence, and may represent the risk degree of the overall vehicle operation data sequence of the target vehicle; however, noise is typically distributed over local vehicle operation data in the sequence of vehicle operation data, where the collision risk factor is insufficient to characterize the risk level of the local vehicle operation data; under normal conditions, the roughness of the pavement surfaces of different areas is not the same, larger or smaller differences exist, the corresponding pavement conditions have hidden trouble conditions of concave-convex or smooth, and when a vehicle runs on the pavement with hidden trouble conditions, the running state of the vehicle is greatly changed to ensure the safety of the vehicle; therefore, in normal circumstances, for any one position data, if all the vehicle operation data corresponding to the position data are changed greatly, the vehicle operation data which are changed greatly are most likely to be caused by the position data; if the vehicle operation data corresponding to the position data is changed greatly, it is indicated that the possibility that the vehicle operation data changed greatly is caused by the position data is low, and meanwhile, the local vehicle operation data is changed greatly due to noise, so that the vehicle operation data changed greatly is likely to be caused by noise interference. Therefore, the present embodiment obtains the anomaly factor by analyzing the difference relation between the vehicle operation data corresponding to the position data, obtains the relative noise factor according to the collision risk factor, and obtains the predicted collision coefficient according to the anomaly factor and the relative noise factor, so as to facilitate the subsequent analysis and processing.

Specifically, performing DTW matching on a jth vehicle operation data sequence of the target vehicle and a jth vehicle operation data sequence of an ith historical vehicle to obtain a DTW distance, and recording the DTW distance as the sequence data similarity of the jth vehicle operation data sequence of the target vehicle and the jth vehicle operation data sequence of the ith historical vehicle; the DTW matching algorithm is a known technique, and this embodiment is not described in detail.

Further, the w-th vehicle operation data in the j-th vehicle operation data sequence of the target vehicle is recorded as target data, the vehicle operation data sequence with the vehicle operation data type being the position data is recorded as a vehicle distance operation data sequence, and the w-th vehicle operation data in the vehicle distance operation data sequence of the target vehicle is recorded as target position data of the target data; in the vehicle distance operation data sequence of the ith historical vehicle, the vehicle operation data with the smallest absolute value of the difference value with the target position data is recorded as the minimum vehicle distance operation data of the target position data in the vehicle distance operation data sequence of the ith historical vehicle; recording the absolute value of the difference value between the target position data and the minimum vehicle distance operation data as the vehicle distance difference value of the target data; and recording the corresponding vehicle operation data of the minimum vehicle distance operation data in the j-th vehicle operation data sequence of the i-th historical vehicle as final reference operation data of the target data in the j-th vehicle operation data sequence of the i-th historical vehicle. Each vehicle corresponds to a vehicle distance operation data sequence, and for any one vehicle operation data in any one vehicle distance operation data sequence, the vehicle operation data corresponds to one vehicle operation data in each vehicle operation data sequence of the vehicle. In addition, if there are a plurality of minimum distance operation data in the i-th historical vehicle distance operation data sequence, the sequence number of the target position data in the i-th historical vehicle distance operation data sequence is marked as a first sequence number, the sequence number of each minimum distance operation data in the i-th historical vehicle distance operation data sequence is marked as a second sequence number, the absolute value of the difference between the second sequence number and the first sequence number is marked as a first absolute value, and the minimum distance operation data with the minimum first absolute value is used as the final minimum distance operation data in the i-th historical vehicle distance operation data sequence of the target position data.

Further, according to the sequence data similarity of the jth vehicle operation data sequence of the target vehicle and the jth vehicle operation data sequence of the ith historical vehicle, the environment interference degree of the curve road junction and the minimum vehicle distance operation data of the w vehicle operation data in the vehicle distance operation data sequence of the target vehicle in the vehicle distance operation data sequence of the ith historical vehicle, initial abnormal factors of the w vehicle operation data in the jth vehicle operation data sequence of the target vehicle are obtained. The method for calculating the initial abnormality factor of the w-th vehicle operation data in the j-th vehicle operation data sequence of the target vehicle comprises the following steps:

in the formula g _j,w An initial anomaly factor representing the w-th vehicle operation data in the j-th vehicle operation data sequence of the target vehicle; i represents the number of all historical vehicles; a2 _j,w The jth vehicle operation data sequence representing the target vehicle; a is that _i,j,w The jth vehicle operation data sequence representing the jth vehicle operation data of the target vehicle, and the jth vehicle operation data sequence representing the ith history vehicleFinally referencing the operational data; b (B) _i,j,w A vehicle distance difference value representing the w-th vehicle operation data in the j-th vehicle operation data sequence of the target vehicle; d, d _i,j A sequence data similarity representing a jth vehicle operation data sequence of the target vehicle and a jth vehicle operation data sequence of the ith historical vehicle; τ represents the environmental disturbance degree of the curve intersection; the absolute value is taken; exp () represents an exponential function based on a natural constant; the embodiment adopts exp (-) function to present inverse proportion relation and normalization processing, and an implementer can select the inverse proportion function and the normalization function according to actual conditions. If the initial abnormality factor of the w-th vehicle operation data in the j-th vehicle operation data sequence of the target vehicle is larger, the degree that the w-th vehicle operation data is affected by the environment of the curve intersection is lower, the relevance with the driving operation of the driver is lower, and the w-th vehicle operation data is more likely to be interfered by noise. Acquiring initial abnormal factors of all vehicle operation data in all vehicle operation data sequences of the target vehicle, carrying out linear normalization on all the initial abnormal factors, and recording each normalized initial abnormal factor as an abnormal factor.

Further, a vehicle running data amount T1 is preset, where the embodiment is described by taking t1=5 as an example, and the embodiment is not specifically limited, where T1 may be determined according to specific implementation conditions; the j 1-th vehicle operation data sequence of the target vehicle is recorded as a first target sequence; the vehicle operation data sequences except the first target sequence are marked as first reference sequences in all vehicle operation data sequences of the target vehicle; in the first target sequence, recording the front T1 vehicle operation data and the rear T1 vehicle operation data of the w1 vehicle operation data as comparison operation data of the w1 vehicle operation data in the first target sequence; and obtaining the comparison operation data of the w1 st vehicle operation data in each first reference sequence of the first target sequence. In the process of acquiring the comparison operation data of the w1 th vehicle operation data, if the quantity of the vehicle operation data actually existing before and after the w1 st vehicle operation data does not meet the preset T1, the comparison operation data is acquired until the quantity of the vehicle operation data actually existing before and after the w1 st vehicle operation data.

Further, according to the collision risk factor of the vehicle operation data sequence of the target vehicle and the comparison operation data of the w1 th vehicle operation data in the corresponding vehicle operation data sequence, the initial relative noise factor of the w1 st vehicle operation data in the j1 st vehicle operation data sequence of the target vehicle is obtained. The method for calculating the initial relative noise factor of the w1 th vehicle operation data in the j1 st vehicle operation data sequence of the target vehicle comprises the following steps:

wherein L is _j1,w1 An initial relative noise factor representing w1 st vehicle operation data in the j1 st vehicle operation data sequence of the target vehicle; v (V) _j1 A number of all first reference sequences representing a j1 st vehicle operation data sequence of the target vehicle; epsilon _j1,w1 Standard deviation of all the collation operation data representing the w1 st vehicle operation data in the j1 st vehicle operation data sequence of the target vehicle; epsilon _j1,v,w1 Standard deviations of all the collation operation data of the w1 st vehicle operation data in the v first reference sequence representing the j1 st vehicle operation data sequence of the target vehicle; beta 3 _j1 A collision risk factor representing a j1 st vehicle operation data sequence of the target vehicle; beta 3 _j1,v A collision risk factor representing a v first reference sequence of a j1 st vehicle operation data sequence of the target vehicle; exp () represents an exponential function based on a natural constant; the embodiment adopts exp (-) function to present inverse proportion relation and normalization processing, and an implementer can select the inverse proportion function and the normalization function according to actual conditions. The greater the initial relative noise factor of the w1 th vehicle operation data in the j1 st vehicle operation data sequence of the target vehicle is, the more obvious the numerical value abnormal change characteristic of the w1 st vehicle operation data is, and the greater the degree of noise interference of the w1 st vehicle operation data is reflected. Acquiring initial relative noise factors of all vehicle operation data in all vehicle operation data sequences of target vehicles, for And (3) carrying out linear normalization on all the initial relative noise factors, and recording each normalized initial relative noise factor as a relative noise factor.

So far, the abnormal factors and the relative noise factors of all the vehicle operation data in each vehicle operation data sequence of the target vehicle are obtained through the method.

Step S004: obtaining a predicted collision coefficient of each vehicle operation data of the target vehicle according to the relative noise factor and the anomaly factor; obtaining a vehicle collision state estimated value of the target vehicle according to the predicted collision coefficient; and pre-warning the target vehicle according to the estimated value of the collision state of the vehicle.

Specifically, according to the relative noise factor and the abnormal factor of the w2 th vehicle operation data in the j2 nd vehicle operation data sequence of the target vehicle, the predicted collision coefficient of the w2 nd vehicle operation data in the j2 nd vehicle operation data sequence of the target vehicle is obtained. The calculation method of the predicted collision coefficient of the w2 th vehicle operation data in the j2 nd vehicle operation data sequence of the target vehicle comprises the following steps:

f _j2,w2 ＝1-(g1 _j2,w2 ×L _j2,w2 )

wherein f _j2,w2 A predicted collision coefficient indicating w 2-th vehicle operation data in the j 2-th vehicle operation data sequence of the target vehicle; g1 _j2,w2 An abnormality factor representing w 2-th vehicle operation data in a j 2-th vehicle operation data sequence of the target vehicle; l (L) _j2,w2 Representing the relative noise factor of the w2 nd vehicle operation data in the j2 nd vehicle operation data sequence of the target vehicle. The greater the predicted collision coefficient of the w2 th vehicle operation data in the j2 nd vehicle operation data sequence of the target vehicle, the higher the predicted value of the w2 nd vehicle operation data is, and the lower the degree of the final noise interference of the w2 nd vehicle operation data is reflected. A predicted collision coefficient for each vehicle operation data in each vehicle operation data sequence of the target vehicle is obtained.

Further, a predicted value obtained by expanding Kalman filtering on the w3 th vehicle operation data in the j3 rd vehicle operation data sequence of the target vehicle is recorded as a traditional predicted value, and an absolute value of a difference value between the traditional predicted value and the w3 rd vehicle operation data is recorded as a comparison filtering value of the w3 rd vehicle operation data. The process of obtaining the predicted value of the data is a well-known content of the extended kalman filtering algorithm, and this embodiment will not be described in detail.

Further, according to the predicted collision coefficients of all the vehicle operation data in each vehicle operation data sequence of the target vehicle and the comparison filtering values, the vehicle collision state estimated value of the target vehicle is obtained. The method for calculating the vehicle collision state estimated value of the target vehicle comprises the following steps:

Wherein P represents a vehicle collision state estimation value of the target vehicle; y represents the number of all vehicle operation data sequences of the target vehicle; r is R _j3 Representing the number of all vehicle operation data in the j3 rd vehicle operation data sequence of the target vehicle; omega _j3,w3 A comparison filter value representing w3 vehicle operation data in the j3 rd vehicle operation data sequence of the target vehicle; f (f) _j3,w3 A predicted collision coefficient representing the w3 rd vehicle operation data in the j3 rd vehicle operation data sequence of the target vehicle.

Further, replacing the state estimation value of the traditional extended Kalman filtering algorithm with the vehicle collision state estimation value of the target vehicle, and marking the vehicle collision state estimation value as a vehicle collision prediction matrix according to all vehicle operation data of the target vehicle and the vehicle collision state estimation value; and carrying out iterative optimization of a prediction model by taking the vehicle collision prediction matrix as a prediction error covariance matrix, carrying out vehicle track prediction by taking model parameters corresponding to the minimum error matrix as an optimal Kalman filtering model, and carrying out vehicle collision early warning. The process of predicting the vehicle track and performing collision pre-warning according to the optimal kalman filter model is known in the patent CN116526454a, and the description of this embodiment is omitted. In addition, the process of performing collision early warning according to the prediction error covariance matrix is a well-known content of "vehicle cooperative collision early warning simulation study based on track prediction", and this embodiment will not be described in detail.

Through the steps, the method for early warning the collision of vehicles meeting at the curve intersection is completed.

Another embodiment of the present invention provides a collision warning system for a vehicle meeting at a curve intersection, the system including a memory and a processor, the processor executing a computer program stored in the memory, the processor performing the following operations:

collecting a plurality of vehicle operation data sequences of a target vehicle, a plurality of vehicle operation data sequences of a plurality of historical vehicles and the environment interference degree of a curve intersection, wherein the vehicle operation data sequences comprise a plurality of vehicle operation data;

obtaining a vehicle state disturbance factor of each vehicle operation data sequence according to the numerical value difference of the vehicle operation data between the vehicle operation data sequences of the target vehicle and the historical vehicle, wherein the vehicle state disturbance factor is used for describing the probability that the vehicle operation data in the vehicle operation data sequence is erroneously recognized as noise; obtaining collision risk factors of each vehicle operation data sequence according to the difference of the vehicle state disorder factors among the vehicle operation data sequences;

obtaining an abnormal factor of each vehicle operation data of the target vehicle according to the environmental disturbance degree of the curve intersection and the change difference of the vehicle operation data of the target vehicle between adjacent positions, wherein the abnormal factor is used for describing the probability that the vehicle operation data is influenced by the operation of a driver; for any one vehicle operation data in any one vehicle operation data sequence, screening a plurality of comparison operation data of each vehicle operation data from the vehicle operation data sequence; according to the distribution discrete condition of the comparison operation data and the collision risk factor in the vehicle operation data sequence of the target vehicle, obtaining the relative noise factor of each vehicle operation data of the target vehicle, wherein the relative noise factor is used for describing the initial difference between the vehicle operation data and the normal vehicle operation data;

Obtaining a predicted collision coefficient of each vehicle operation data of the target vehicle according to the relative noise factor and the anomaly factor, wherein the predicted collision coefficient is used for describing the final difference between the vehicle operation data and the normal vehicle operation data; obtaining a vehicle collision state estimated value of the target vehicle according to the predicted collision coefficient, wherein the vehicle collision state estimated value is used for describing an error of a collision result of the target vehicle; and pre-warning the target vehicle according to the estimated value of the collision state of the vehicle.

The above description is only of the preferred embodiments of the present invention and is not intended to limit the invention, but any modifications, equivalent substitutions, improvements, etc. within the principles of the present invention should be included in the scope of the present invention.

Claims (10)

1. The method for early warning the collision of the vehicles meeting at the curve intersection is characterized by comprising the following steps of:

collecting a plurality of vehicle operation data sequences of a target vehicle, a plurality of vehicle operation data sequences of a plurality of historical vehicles and the environment interference degree of a curve intersection, wherein the vehicle operation data sequences comprise a plurality of vehicle operation data;

obtaining a vehicle state disturbance factor of each vehicle operation data sequence according to the numerical value difference of the vehicle operation data between the vehicle operation data sequences of the target vehicle and the historical vehicle, wherein the vehicle state disturbance factor is used for describing the probability that the vehicle operation data in the vehicle operation data sequence is erroneously recognized as noise; obtaining collision risk factors of each vehicle operation data sequence according to the difference of the vehicle state disorder factors among the vehicle operation data sequences;

Obtaining an abnormal factor of each vehicle operation data of the target vehicle according to the environmental disturbance degree of the curve intersection and the change difference of the vehicle operation data of the target vehicle between adjacent positions, wherein the abnormal factor is used for describing the probability that the vehicle operation data is influenced by the operation of a driver; for any one vehicle operation data in any one vehicle operation data sequence, screening a plurality of comparison operation data of each vehicle operation data from the vehicle operation data sequence; according to the distribution discrete condition of the comparison operation data and the collision risk factor in the vehicle operation data sequence of the target vehicle, obtaining the relative noise factor of each vehicle operation data of the target vehicle, wherein the relative noise factor is used for describing the initial difference between the vehicle operation data and the normal vehicle operation data;

obtaining a predicted collision coefficient of each vehicle operation data of the target vehicle according to the relative noise factor and the anomaly factor, wherein the predicted collision coefficient is used for describing the final difference between the vehicle operation data and the normal vehicle operation data; obtaining a vehicle collision state estimated value of the target vehicle according to the predicted collision coefficient, wherein the vehicle collision state estimated value is used for describing an error of a collision result of the target vehicle; and pre-warning the target vehicle according to the estimated value of the collision state of the vehicle.

2. The method for pre-warning the collision of vehicles at a curve intersection according to claim 1, wherein the method for obtaining the vehicle state disturbance factor of each vehicle operation data sequence according to the numerical difference of the vehicle operation data between the vehicle operation data sequences of the target vehicle and the historical vehicle comprises the following specific steps:

marking the target vehicle and the historical vehicle as one reference vehicle, and fitting any vehicle operation data sequence of any reference vehicle by using a least square method to obtain a fitting curve;

for any one vehicle operation data in the vehicle operation data sequence, marking the numerical value of the vehicle operation data on the fitting curve as original smooth operation data of the vehicle operation data, marking the absolute value of the difference value between the original smooth operation data and the vehicle operation data as an original operation data difference value of the vehicle operation data, acquiring the original operation data difference values of all the vehicle operation data in the vehicle operation data sequence, carrying out linear normalization on all the original operation data difference values, and marking each normalized original operation data difference value as a smooth operation difference value;

Wherein beta represents any one of the vehicle operation data sequences of any one of the reference vehiclesIs a factor of the initial vehicle state disturbance; n represents the number of all vehicle operation data in the vehicle operation data sequence; a, a _n Representing nth vehicle operation data in the sequence of vehicle operation data;representing a mean value of all vehicle operation data in the sequence of vehicle operation data; Δa1 _n A smooth running difference value representing nth vehicle running data in the vehicle running data sequence; the absolute value is taken; obtaining initial vehicle state disturbance factors of all vehicle operation data sequences of the reference vehicles, obtaining initial vehicle state disturbance factors of all vehicle operation data sequences of all reference vehicles, carrying out linear normalization on all initial vehicle state disturbance factors, and recording the normalized initial vehicle state disturbance factors as vehicle state disturbance factors.

3. The method for pre-warning the collision of vehicles meeting at the curve intersection according to claim 1, wherein the method for obtaining the collision risk factor of each vehicle operation data sequence according to the difference of the vehicle state disorder factors among the vehicle operation data sequences comprises the following specific steps:

in the formula, beta 3 _z A collision risk factor representing a z-th vehicle operation data sequence of the target vehicle; beta 2 _z A vehicle state disturbance factor representing a z-th vehicle operation data sequence of the target vehicle;a vehicle condition disturbance factor representing a z-th vehicle operation data sequence of all historical vehicles.

4. The method for pre-warning the collision of vehicles at a curve intersection according to claim 1, wherein the method for obtaining the abnormal factor of each vehicle operation data of the target vehicle according to the environmental disturbance degree of the curve intersection and the variation difference of the vehicle operation data of the target vehicle between adjacent positions comprises the following specific steps:

performing DTW matching on the jth vehicle operation data sequence of the target vehicle and the jth vehicle operation data sequence of the ith historical vehicle to obtain a DTW distance, and recording the DTW distance as the sequence data similarity of the jth vehicle operation data sequence of the target vehicle and the jth vehicle operation data sequence of the ith historical vehicle;

the method comprises the steps of recording w-th vehicle operation data in a j-th vehicle operation data sequence of a target vehicle as target data, recording a vehicle operation data sequence with the type of the vehicle operation data as position data as a vehicle distance operation data sequence, and recording w-th vehicle operation data in the vehicle distance operation data sequence of the target vehicle as target position data of the target data;

In the vehicle distance operation data sequence of the ith historical vehicle, the vehicle operation data with the smallest absolute value of the difference value with the target position data is recorded as the minimum vehicle distance operation data of the target position data in the vehicle distance operation data sequence of the ith historical vehicle; the absolute value of the difference value between the target position data and the minimum vehicle distance operation data is recorded as a vehicle distance difference value of the target data; recording the corresponding vehicle operation data of the minimum vehicle distance operation data in the j-th vehicle operation data sequence of the i-th historical vehicle as final reference operation data of the target data in the j-th vehicle operation data sequence of the i-th historical vehicle;

in the formula g _j，w An initial anomaly factor representing the w-th vehicle operation data in the j-th vehicle operation data sequence of the target vehicle; i represents the number of all historical vehicles; a2 _j，w The jth vehicle operation data sequence representing the target vehicle; a is that _i，j，w The jth vehicle operation data sequence representing the target vehicle indicates the jth vehicle operation data sequence of the target vehicleFinal reference operational data in a jth vehicle operational data sequence of an ith historical vehicle; b (B) _i，j，w A vehicle distance difference value representing the w-th vehicle operation data in the j-th vehicle operation data sequence of the target vehicle; d, d _i，j A sequence data similarity representing a jth vehicle operation data sequence of the target vehicle and a jth vehicle operation data sequence of the ith historical vehicle; τ represents the environmental disturbance degree of the curve intersection; the absolute value is taken; exp () represents an exponential function based on a natural constant; acquiring initial abnormal factors of all vehicle operation data in all vehicle operation data sequences of the target vehicle, carrying out linear normalization on all the initial abnormal factors, and recording each normalized initial abnormal factor as an abnormal factor.

5. The method for pre-warning a collision between vehicles at a curve intersection according to claim 1, wherein the screening of a plurality of comparison operation data of each vehicle operation data from the vehicle operation data sequence comprises the following specific steps:

presetting a vehicle operation data quantity T1, and recording a j 1-th vehicle operation data sequence of a target vehicle as a first target sequence; the vehicle operation data sequences except the first target sequence are marked as first reference sequences in all vehicle operation data sequences of the target vehicle; in the first target sequence, the T1 th vehicle operation data before the w1 st vehicle operation data and the T1 st vehicle operation data after the w1 st vehicle operation data are recorded as comparison operation data of the w1 st vehicle operation data in the first target sequence.

6. The method for pre-warning the collision of vehicles at a curve intersection according to claim 5, wherein the obtaining the relative noise factor of each vehicle operation data of the target vehicle according to the distributed discrete condition of the comparison operation data and the collision risk factor in the vehicle operation data sequence of the target vehicle comprises the following specific steps:

wherein L is _j1，w1 An initial relative noise factor representing w1 st vehicle operation data in the j1 st vehicle operation data sequence of the target vehicle; v (V) _j1 A number of all first reference sequences representing a j1 st vehicle operation data sequence of the target vehicle; epsilon _j1，w1 Standard deviation of all the collation operation data representing the w1 st vehicle operation data in the j1 st vehicle operation data sequence of the target vehicle; epsilon _j1，v，w1 Standard deviations of all the collation operation data of the w1 st vehicle operation data in the v first reference sequence representing the j1 st vehicle operation data sequence of the target vehicle; beta 3 _j1 A collision risk factor representing a j1 st vehicle operation data sequence of the target vehicle; beta 3 _j1，v A collision risk factor representing a v first reference sequence of a j1 st vehicle operation data sequence of the target vehicle; exp () represents an exponential function based on a natural constant; acquiring initial relative noise factors of all vehicle operation data in all vehicle operation data sequences of the target vehicle, carrying out linear normalization on all initial relative noise factors, and recording each normalized initial relative noise factor as a relative noise factor.

7. The method for pre-warning the collision of vehicles at a curve intersection according to claim 1, wherein the method for obtaining the predicted collision coefficient of each vehicle operation data of the target vehicle according to the relative noise factor and the anomaly factor comprises the following specific steps:

f _j2,w2 ＝1-(g1 _j2,w2 ×L _j2,w2 )

wherein f _j2,w2 A predicted collision coefficient indicating w 2-th vehicle operation data in the j 2-th vehicle operation data sequence of the target vehicle; g1 _j2,w2 An abnormality factor representing w 2-th vehicle operation data in a j 2-th vehicle operation data sequence of the target vehicle; l (L) _j2,w2 Representing the relative noise factor of the w2 nd vehicle operation data in the j2 nd vehicle operation data sequence of the target vehicle.

8. The method for pre-warning the collision of vehicles at a curve intersection according to claim 1, wherein the method for obtaining the estimated value of the collision state of the target vehicle according to the predicted collision coefficient comprises the following specific steps:

the method comprises the steps of recording a predicted value obtained by expanding Kalman filtering on w3 vehicle operation data in a j3 vehicle operation data sequence of a target vehicle as a traditional predicted value, and recording an absolute value of a difference value between the traditional predicted value and the w3 vehicle operation data as a comparison filtering value of the w3 vehicle operation data;

Wherein P represents a vehicle collision state estimation value of the target vehicle; y represents the number of all vehicle operation data sequences of the target vehicle; r is R _j3 Representing the number of all vehicle operation data in the j3 rd vehicle operation data sequence of the target vehicle; omega _j3,w3 A comparison filter value representing w3 vehicle operation data in the j3 rd vehicle operation data sequence of the target vehicle; f (f) _j3,w3 A predicted collision coefficient representing the w3 rd vehicle operation data in the j3 rd vehicle operation data sequence of the target vehicle.

9. The method for pre-warning the collision of vehicles at the intersection of a curve according to claim 1, wherein the pre-warning the target vehicles according to the estimated value of the collision state of the vehicles comprises the following specific steps:

replacing the state estimation value of the extended Kalman filtering algorithm with the vehicle collision state estimation value of the target vehicle, and marking the vehicle collision state estimation value as a prediction error covariance matrix according to all vehicle operation data of the target vehicle and the vehicle collision state estimation value; and carrying out iterative optimization of a prediction model by taking the vehicle collision prediction matrix as a prediction error covariance matrix, carrying out vehicle track prediction by taking model parameters corresponding to the minimum error matrix as an optimal Kalman filtering model, and carrying out vehicle collision early warning.

10. A curve crossing collision warning system comprising a memory, a processor and a computer program stored on the memory and operable on the processor, wherein the computer program when executed by the processor implements the steps of a curve crossing collision warning method as claimed in any one of claims 1 to 9.