CN116485012A - Prediction calculation method for space-time distribution envelope of traffic participant running track - Google Patents

Abstract

The prediction calculation method for the space-time distribution envelope of the running track of the traffic participant provided by the disclosure comprises the following steps: s1: screening a feasible running path; s2: calculating longitudinal and transverse uncertainty parameters; s3: determining a predicted endpoint state; s4: fitting the running track of the traffic participant to be predicted; s5: a spatiotemporal distribution envelope of the predicted trajectory is calculated. The method and the device fully utilize map topology information and historical track information, adopt a rule method, provide a calculation method framework for traffic participant running track prediction and space-time distribution envelope, have the characteristics of strong generalization capability, quick calculation and convenient deployment, and realize quick calculation of traffic participant space-time occupation prediction.

Description

Prediction calculation method for space-time distribution envelope of traffic participant running track

Technical Field

The disclosure belongs to the field of vehicle engineering, and in particular relates to a prediction calculation method of a traffic participant running track space-time distribution envelope.

Background

The future track prediction of the traffic participants is to predict and calculate the future track of the traffic participants according to the current and historical running tracks and environmental information of the traffic participants such as pedestrians, non-motor vehicles and motor vehicles, and the future track prediction is used for supporting an automatic driving decision and planning system to design a safe running track so as to avoid collision with the traffic participants.

At present, a track prediction method of a traffic participant mainly adopts a learning method, a coding-decoding structure of a depth neural network is established, input information such as a historical track of the traffic participant, road topology and the like is coded into a hidden state by using an encoder network, and then the hidden state is decoded into multi-mode prediction track coordinate information containing probability by using a decoder network. The method generally takes the error between the track predicted value and the true value as an index, and obtains a predicted neural network model in a gradient descent mode. However, this type of approach has three problems: firstly, multimode track prediction cannot be directly integrated into an automatic driving decision-making and planning module at the downstream, and the sampling track is generally subjected to post-screening by taking the constraint, so that the optimality of the planning track is affected; secondly, the track prediction cannot reach complete accuracy, the prediction result in the form of track coordinates loses the uncertainty of space-time distribution of vehicle operation, and the safety of planning tracks is difficult to strictly guarantee; thirdly, because of the heterogeneous characteristics of the road structure and traffic participants of the data set, the track prediction model crossing the data set needs to be retrained, and the application cannot be directly deployed.

In practice, the fundamental meaning of developing the track prediction work of the traffic participants is to analyze the future space-time occupation of the traffic participants, and ensure the running safety by restraining the track planned by the decision of the automatic driving vehicle from overlapping with the corresponding space-time position. From this point of view, the method has more significance to the space-time distribution envelope prediction of the running track of the traffic participants than to the track coordinate form; the space-time distribution envelope can reduce the dimension of the multi-mode track to a single mode, and can be directly used in decision and planning algorithms of automatic driving vehicles. In addition, the prediction of the space-time distribution envelope of the track does not require high-precision deterministic track coordinates, and the road map has a determined topological connection relationship, so that traffic participants generally tend to run close to the center line of the lane, and therefore, the characteristic can be utilized to develop a rule type prediction method, and the generalization capability, calculation speed and deployment convenience of the prediction model are improved.

Disclosure of Invention

The present disclosure aims to solve, at least to some extent, one of the technical problems in the related art.

Therefore, the prediction calculation method of the space-time distribution envelope of the running track of the traffic participant aims at the prediction calculation of the space-time distribution envelope of the running track of the traffic participant, and the historical track information and the road map topology information are utilized for estimation by adopting a rule type method, so that the development of driving auxiliary equipment facing the improvement of safety performance and the optimization design of an automatic driving decision and planning control strategy are supported.

In order to achieve the above purpose, the present disclosure adopts the following technical scheme:

the prediction calculation method for the space-time distribution envelope of the running track of the traffic participant provided by the disclosure comprises the following steps:

s1, screening feasible operation paths: comparing the historical track information of the traffic participant to be predicted with the central line information of the map road, screening and obtaining all feasible central lines of the road and the central lines of the subsequent lanes from the central line information of the map road, and splicing the central lines with topological relation to form a feasible running path set;

s2, calculating longitudinal and transverse uncertainty parameters: calculating longitudinal and transverse uncertainty parameters of each feasible running path by utilizing running state characteristics extracted from historical track information of a traffic participant to be predicted and the feasible running path set obtained by screening;

S3, determining a predicted endpoint state: calculating the running state of the predicted track end point of the traffic participant to be predicted, including coordinates, speed and orientation angle, by using the running state characteristics, longitudinal uncertainty parameters and each feasible running path extracted from the historical track information of the traffic participant to be predicted;

s4, fitting the running track of the traffic participant to be predicted: fitting and calculating the running track of the traffic participant to be predicted in the prediction time domain and outputting the running state of each moment in the prediction time domain;

s5, calculating a space-time distribution envelope of the predicted track: and drawing an uncertainty envelope of the running track of the traffic participant to be predicted in the prediction time domain by using the transverse and longitudinal uncertainty parameters of each feasible running path and adopting a geometric shape to obtain a space-time distribution envelope of the prediction track.

In some embodiments, S1 specifically includes:

s11, preprocessing historical track information of a traffic participant to be predicted, including denoising and extracting running state characteristics of the traffic participant to be predicted, wherein the running state characteristics include average speed, running direction and transverse and longitudinal acceleration of the historical track of the traffic participant to be predicted;

S12, screening a lane center line closest to the historical track of the traffic participant to be predicted at the current moment by using the preprocessed historical track information to serve as a candidate lane center line;

s13, traversing the center line of the candidate lane, and acquiring the center line of the subsequent lane by using map topology information until the farthest distance exceeds a set distance threshold L;

s14, splicing lane central line information according to the connection relation, removing duplication, and obtaining a feasible running path, so that a feasible running path set is formed.

In some embodiments, in S13, a depth-first or breadth-first search algorithm is employed to traverse the candidate lane centerline.

In some embodiments, the historical trajectory information of the traffic participant to be predicted includes longitudinal and lateral coordinates, speed, and heading angle of the traffic participant to be predicted in a historical time domain.

In some embodiments, the map roadway centerline information includes a number for each lane centerline, a number for a lane centerline subsequent to each lane centerline, longitudinal and lateral coordinate vectors for each lane centerline, a status of traffic lights within each lane, and a flag of whether a lane passes an intersection.

In some embodiments, in S2, the longitudinal uncertainty parameter for each feasible travel path includes an expected longitudinal acceleration and an ending longitudinal distance deviation calculated according to the following equation:

wherein,,for the expected longitudinal acceleration of the first possible path of travel, +.>Deviation of the longitudinal distance for the end of each possible path of travel,/->And->Longitudinal acceleration and average speed, respectively, derived from historical trajectory information of the traffic participant to be predicted,/->Signal lamp status for the first possible path, +.>For the first feasible path to pass through the sign of the intersection, alpha _px 、＝ _px 、γ _px The first, second and third influence coefficients are respectively.

In some embodiments, in S2, the lateral uncertainty parameter for each feasible travel path includes an expected lateral acceleration and an end-point lateral distance deviation calculated according to the following equation:

wherein,,and->Expected lateral acceleration and end-point lateral distance deviation, +.>And->Lateral acceleration and average speed obtained from historical track information of traffic participants to be predicted, respectively, = _py For the end point transverse fixed deviation alpha _py 、γ _py 、δ _py The fourth, fifth and sixth influence coefficients are respectively.

In some embodiments, S3 specifically includes:

s31, determining curvature adjustment parameters: for each feasible travel path, a turning curvature ρ based on each feasible travel path ^l Direction of turning theta ^l And a two-sided conflict object number ratio τ ^l Comprehensive determination of curvature adjustment parameters for each predicted trajectoryThe calculation formulas are respectively as follows:

wherein,,and->The starting point and the end point of the first feasible running path face angles respectively; />And->Longitudinal and transverse coordinates of the start point of the first feasible path, respectively, < >>And->Longitudinal and transverse coordinates of the end point of the first feasible travel path, respectively, and provide: if theta is ^l >0, turn left, if θ ^l If less than 0, turn right, if θ ^l =0, then straight; />And->Respectively the first feasible operationThe number of traffic participants with lateral conflict on the left and right sides of the path; alpha _ρ And beta _θ The seventh influence coefficient and the eighth influence coefficient are respectively, and SGN is a sign function; />For the expected lateral acceleration of each possible travel path.

S32, calculating a longitudinal running distance in a prediction time domain: adjusting parameters based on curvature of each predicted trajectoryLongitudinal movement distance +.f. of each predicted track in the prediction time domain is calculated by Newton kinematics respectively>

S33, interpolation is carried out on coordinate points on each feasible running path respectively, and a path coordinate point closest to the coordinates of the traffic participant to be predicted at the current moment is searched;

S34, obtaining the state of the prediction end point along the path direction: starting from the nearest path coordinate point, searching for the end point path coordinate, speed and orientation angle at the predicted distance along the corresponding feasible running path direction, and jointly forming the state of the predicted track end point of the traffic participant to be predicted.

In some embodiments, in S4, a cubic curve is employedAnd fitting and calculating the running track of the traffic participant to be predicted in the prediction time domain, and solving the cubic curve parameters by considering the starting and ending point constraint as follows:

wherein a is ₀ Constant term of cubic curve, a ₁ 、a ₂ And a ₃ Each coefficient of the cubic curve; v _t For the speed of the traffic participant to be predicted in the world coordinate system at time t, v _T The speed of a traffic participant to be predicted in a world coordinate system at the time t+T, wherein T is the duration of a prediction time domain; p is p _t Is the coordinate of the traffic participant to be predicted under the world coordinate system at the moment t, p _T To be the traffic participant at time T + T.

In some embodiments, S5 specifically includes:

and for each feasible running path, adopting an envelope form of geometric shapes according to the end longitudinal distance deviation of each feasible running path, and reversely pushing longitudinal envelope parameters of each prediction moment on the corresponding feasible running path by a linear or nonlinear method to obtain the space-time distribution envelope of the prediction track.

The characteristics and beneficial effects of the present disclosure:

the prediction calculation method for the space-time distribution envelope of the running track of the traffic participant is characterized by adopting a rule type deduction calculation method, fully utilizing road map topological relation information and explicitly carrying out prediction calculation on the space-time distribution envelope of the running track of the traffic participant. Compared with the existing track prediction method, the coordinate type prediction is expanded into space-time distribution envelope prediction, and the method can be directly integrated into the development of an automatic driving control algorithm; in addition, the rule explicit solution also improves the rapid calculation and convenient deployment capability of the method in heterogeneous road traffic scenes, and can provide support for driving safety assistance or automatic driving decision control in complex environments.

Drawings

Fig. 1 is an overall flowchart of a predictive computation method for a traffic participant trajectory spatiotemporal distribution envelope provided by an embodiment of a first aspect of the disclosure.

Fig. 2 is a schematic view for explaining an embodiment of the first aspect of the present disclosure.

Fig. 3 is a schematic structural diagram of an electronic device according to an embodiment of a third aspect of the present disclosure.

The specific embodiment is as follows:

in order to make the objects, technical solutions and advantages of the present application more apparent, the present application is described in further detail below with reference to the accompanying drawings and examples. It should be understood that the specific embodiments described herein are for purposes of illustration only and are not intended to limit the present application.

On the contrary, the application is intended to cover any alternatives, modifications, equivalents, and variations as may be included within the spirit and scope of the application as defined by the appended claims. Further, in the following detailed description of the present invention, certain specific details are set forth in order to provide a better understanding of the present application. The present application will be fully understood by those skilled in the art without a description of these details.

As shown in fig. 1, a method for predicting and calculating a space-time distribution envelope of a traffic participant movement track according to an embodiment of a first aspect of the present disclosure includes the following steps:

s1, screening feasible operation paths: comparing the historical track information of the traffic participant to be predicted with the central line information of the map road, screening and obtaining all feasible central lines of the road and the central lines of the subsequent lanes from the central line information of the map road, and splicing the central lines with topological relation to form a feasible running path set F;

s2, calculating longitudinal and transverse uncertainty parameters: calculating longitudinal and transverse uncertainty parameters of each feasible running path by utilizing running state characteristics extracted from historical track information of a traffic participant to be predicted and a feasible running path set F obtained by screening;

S3, determining a predicted endpoint state: calculating the running state of the predicted track end point of the traffic participant to be predicted, including coordinates, speed, orientation angle and the like, by using the running state characteristics, the longitudinal uncertainty parameters and each feasible running path extracted from the historical track information of the traffic participant to be predicted;

s4, fitting the running track of the traffic participant to be predicted: fitting and calculating the running track of the traffic participant to be predicted in the prediction time domain by adopting a polynomial function, and outputting the running state of each moment in the prediction time domain;

s5, calculating a predicted track space-time distribution envelope: the uncertainty envelope of the running track of the traffic participant to be predicted in the prediction time domain is depicted by adopting geometric shapes by using the transverse and longitudinal uncertainty parameters of each feasible running path.

In some embodiments, when predicting the spatio-temporal distribution envelope of the traffic participant rail at time t, the known information that needs to be acquired includes: historical track information for traffic participantsAnd map road center line information c= { C ^j }, wherein->In the historical time domain [ t-n, t ] for traffic participant numbered i]The track information in the inner part of the track information,x, y are the longitudinal and transverse coordinates of the traffic participant respectively, v is the speed of the traffic participant, phi is the direction angle of the traffic participant, n is the known historical track time domain, i is the number of the traffic participant in the scene, and the number i=0 of the traffic participant to be predicted is set; c ^j As lane center line information numbered j in the map,longitudinal and transverse coordinate vectors, j, of the lane center line, respectively _next Numbering of the lane subsequent to the jth lane centerline, I _signal Marking whether the signal lamp of the lane is red or not, when I _signal When=0, it is expressed as non-red light, when I _signal When=1, it is expressed as red light, I _inter For the mark of whether the path is in the intersection or not, when I _inter When=0, it indicates that the intersection is not located, when I _inter When=1, the indication is in the intersection. For the example of a scenario of the crossroad shown in fig. 2, including traffic participants to be predicted (i=0), and one other vehicle (i=1), one pedestrian (i=2), and one rider (i=3); lane center line information { c } ⁰ ，c ¹ ，c ² ，c ³ ，c ⁴ ，...}. The prediction target is the future time domain [ T, t+T ]]Predicted track status for each time instant in>And corresponding spatio-temporal distribution envelope->Wherein,,for the prediction of the feasible route numbered k, +.>A spatiotemporal distribution envelope of the prediction results for a viable run path numbered k.

In some embodiments, step S1 specifically includes:

s11, preprocessing historical track information of a traffic participant to be predicted, including denoising, extracting running state characteristics of the traffic participant to be predicted and the like;

S12, screening the lane center line where the traffic participant to be predicted is likely to be positioned at the current moment according to the track distance and the distance between the lane center lines by utilizing the preprocessed historical track information, namely selecting the lane center line closest to the historical track of the traffic participant to be predicted;

s13, traversing the screened lane center line where the possible lane is located, and acquiring the subsequent lane center line by using map topology information until the farthest distance exceeds a set distance threshold L;

s14, splicing lane central line information according to the connection relation, removing duplication, and obtaining a feasible running path, so that a feasible running path set F is formed.

Further, in step S11, the preprocessing of the historical track information data mainly includes denoising, and extracting the running state features such as the average speed, the running direction of the historical track, and the lateral and longitudinal acceleration. In the present embodiment, the window moving average method of formula (1) is first used to track the history information q of the original traffic participant (toQuantitative representation) to perform data denoising to obtain denoised historical track information(in vector form); subsequently extracting the running direction of the denoised history track by using (2) to (5), respectively >Average speed>Lateral acceleration->And longitudinal acceleration->

Wherein n is _w To smooth the window size, t _z To smooth the centre instant of the window, i _w For smoothing the ith in the window _w The time is the same;and->Longitudinal and transverse coordinates of the traffic participant at time t-n after denoising, respectively,/->And->Longitudinal and transverse coordinates of the traffic participant at time t after denoising treatment, respectively, +.>For the traffic participator after denoising treatment in the historical time domain [ t-n, t ]]Speed at each time in>And->The speeds of the traffic participants at the time t-n and the time t after denoising processing are respectively, and delta t is the time interval used for collecting the historical track points.

Further, in step S12, the lane center line information where the history track is likely to be located is screened out in segments by comparing the coordinate position, the running direction of the history track with the lane center line information in the map topology information. In the present embodiment, the history time domain [ t-n, t-n+m ] is considered]Historical time domain [ t-m, t ]]The historical track coordinates at each time in the interior are searched by adopting a traversing mode to find a lane central line set which is closest to and has the same direction as the coordinate points at the time t-n+m, and the lane central line set is recorded as And->Respectively using history time domains [ t-n, t-n+m ]]Historical time domain [ t-m, t ] ]And (3) obtaining a nearest lane center line set by the track coordinates at each moment in time, wherein m is a time step length capable of stably determining the running direction of the vehicle under the condition of considering track data noise. For the scenario shown in fig. 2, m=5 is selected, and C is obtained by screening ^adj ＝{{c ⁰ ，c ⁵ }，{c ¹ ，c ³ }}。

Further, in step S13, a depth-first or breadth-first search algorithm is employed to search for the center line set C from the nearest lane ^adj Starting, obtaining the information of the center lines of all subsequent lanes in the road topology data until the total length of the paths in the depth-first search algorithm or the breadth-first search algorithm is larger than a threshold L _c Form lane center line set C ^alt . In this embodiment, a depth-first search algorithm is used, and the length threshold is set to L _c =100m, for the scenario shown in fig. 2, the searched lane center line set C ^alt ＝{c ⁰ ，c ¹ ，c ² ，...c ⁷ }。

Further, in step S14, corresponding lane center line information is spliced according to the topological connection relationship to form a path set, and then an intersection of path sets corresponding to each section of history track is taken to obtain a feasible running path set F. In the present embodiment, for lane center line set C ^alt Splicing lane center lines with topological connection relationship to obtain a candidate path set F _t--n：t-n+5 ＝{{c ⁰ ，c ¹ ，c ² }，{c ⁰ ，c ³ ，c ⁴ }，{c ⁵ ，c ⁶ }，{c ⁵ ，c ⁷ Sum F _t-5：t ＝{{c ¹ ，c ² }，{c ³ ，c ⁴ -x }; then, taking the intersection of the candidate path sets from t-n to t-n+5 time and from t-5 to t time to obtain a feasible running path set F=F _t-n：t-n+5 ∩F _t-5：t ＝{{c ¹ ，c ² }，{c ³ ，c ⁴ }}＝{f ¹ ，f ² }，Wherein l is the number of the feasible operation path obtained by screening; optional Path +.>And->Can be taken by corresponding to the +.>And->And (3) carrying out Boolean calculation on the value of the obtained product.

In some embodiments, step S2 specifically includes:

s21, calculating longitudinal uncertainty parameters: and (3) considering the longitudinal acceleration of the historical track, whether the historical track passes through an intersection, the state of a signal lamp and other factors on the optional path, and calculating longitudinal uncertainty parameters such as expected longitudinal acceleration, end longitudinal distance deviation and the like for each feasible running path. In the present embodiment, the expected longitudinal acceleration of the first feasible travel path is defined asThe deviation of the longitudinal distance of the end point of each feasible travel path is defined as +.>The end longitudinal distance deviation of each feasible running path is equal, is a function related to the average speed of the historical track, and has the following calculation formulas:

wherein alpha is _px And beta _px The values are respectively a first influence coefficient and a second influence coefficient, and are 0.1 and 0.2; gamma ray _px And the third influence coefficient takes a value of 0.2.

S22, calculating a transverse uncertainty parameter: and calculating the expected lateral acceleration, the end point lateral distance deviation and other lateral uncertainty parameters of each feasible running path by considering the factors such as the speed, the lateral acceleration and the like of the historical track. In the present embodiment, the expected lateral acceleration of each feasible travel path is defined as The expected lateral acceleration of each feasible travel path is equal and is a function related to the historical track lateral acceleration, and the end lateral distance deviation of each feasible travel path is defined as +.>The end point transverse distance deviation of each feasible running path is equal, is a function related to the average speed and the transverse acceleration of the historical track, and has the following calculation formulas:

wherein alpha is _py The value of the fourth influence coefficient is 0.7; beta _py For the end point transverse fixed deviation, 1m, gamma can be taken _py And delta _py The fifth and sixth influence coefficients are respectively 0.01 and 0.1.

In some embodiments, step S3 specifically includes:

s31, determining curvature adjustment parameters: considering the driver's tendency to follow the inside of a curveRunning, calculating turning curvature rho of each feasible running path according to each feasible running path ^l Direction of turning theta ^l Two-sided conflict object number ratio τ ^l Etc. to determine curvature adjustment parameters for each predicted trajectoryIn the present embodiment, the turning curvature ρ of the path ^l The method is characterized by approximate difference of the starting point and the ending point of the path and the direction angle of the turning direction theta ^l The method adopts the calculation of the starting point orientation vector and the starting and ending point connecting line orientation vector outer product to obtain the number ratio tau of the conflict objects at two sides ^l The curvature adjustment parameter of the predicted track is calculated by the number of the transverse conflict traffic participants on the left side and the right side of the screening path>By comprehensively considering ρ ^l 、θ ^l And τ ^l The calculation formula is calculated as follows:

wherein,,and->The starting point and the end point of the first feasible running path face angles respectively; />And->Longitudinal and transverse coordinates of the start point of the first feasible path, respectively, < >>And->Longitudinal and transverse coordinates of the end point of the first feasible travel path, respectively, and provide: if theta is ^l >0, turn left, if θ ^l If less than 0, turn right, if θ ^l =0, then straight; />And->The number of traffic participants with transverse conflict on the left side and the right side of the first feasible running path respectively; alpha _ρ And beta _θ The seventh and eighth influence coefficients are respectively 0.2 and 0.1, and SGN is a sign function.

S32, calculating a longitudinal running distance in a prediction time domain: adjusting parameters based on curvature of each predicted trajectoryLongitudinal movement distance +/of each predicted trajectory was calculated separately using Newton kinematics>In the present embodiment, the longitudinal movement distance of the predicted trajectory is +.>The calculation is performed as follows:

S33, interpolation is carried out on each feasible running path, and the mapping point of the traffic participant to be predicted at the current moment on the path is obtained: and respectively interpolating coordinate points on each feasible running path, and searching a path coordinate point closest to the coordinates of the traffic participant to be predicted at the moment t, and taking the path coordinate point as a mapping point of the traffic participant to be predicted at the current moment on the path. In the present embodiment, interpolation is performed on the feasible paths at interpolation interval Δd=0.1m, and then the interpolation points are traversed to find the coordinates (x _t ，y _t ) Nearest path coordinate point

S34, obtaining state information of a prediction end point along the path direction: from the nearest path coordinate pointInitially, along the corresponding feasible travel path direction, the end point path coordinates and other state information at the predicted distance are searched. In the embodiment, a traversal mode is adopted to iteratively calculate the distance +.> Wherein z is the number of coordinate points on the path until the condition L is satisfied _cum ≥L _p Corresponding coordinate point +.>Regarding as the predicted end point coordinates, the running state at this point is calculated using the following formula>

In some embodiments, step S4 specifically includes:

since a cubic curve can be associated with a kinematic law and can be explicitly solved using the starting and ending point states of the predicted trajectory, a cubic curve fit is employed. Fitting the function: Considering the starting and ending point constraint, the cubic curve parameters can be obtained by solving:

wherein a is ₀ Constant term of cubic curve, a ₁ 、a ₂ And a ₃ Each coefficient of the cubic curve; v _t ＝[v _x ，v _y ] _t Is the velocity vector v of the traffic participant to be predicted in the world coordinate system at the moment t _T ＝[v _x ，v _y ] _T Is the velocity vector v of the traffic participant to be predicted in the world coordinate system at the time t+T _x And v _y The component speeds in the transverse and longitudinal directions are respectively; p is p _t Taking p as the coordinate vector of the traffic participant to be predicted under the world coordinate system at the moment t _t ＝[x，y] _t ，p _T Taking p as the coordinate vector of the traffic participant to be predicted at the time t+T _T ＝[x，y] _T X and y are respectively the sub-coordinates in the transverse and longitudinal directions; (. Cndot.) represents the first derivative with respect to time. Accordingly, the orientation angle phi of each point on the predicted track is further calculated, and finally the predicted track state of each moment in the predicted time domain is obtained For the prediction result of the kth feasible route,/-Can->

In some embodiments, step S5 specifically includes:

according to the deviation of the end longitudinal distance of each feasible running pathThe longitudinal envelope parameters of each predicted moment are reversely deduced through a linear or nonlinear method by adopting envelope forms of circles, ellipses, rectangles or other geometric shapes. The envelope form is selected according to a track prediction follow-up processing mode, wherein the circle is selected conveniently if the distance calculation is carried out, the ellipse is selected appropriately if the probability calculation is carried out, and the rectangle is used appropriately if the accuracy and the calculation simplicity are considered. In this embodiment, a rectangular envelope form is adopted, that is, the predicted track coordinates are taken as rectangular center points, the orientation angles are taken as rectangular orientations, and a linear manner is adopted to perform back-pushing, so as to obtain a space-time distribution envelope rectangular length of +. >Width is->Wherein m is _p ∈[1，T/Δt]For the prediction time step starting from the time t, the spatial-temporal distribution envelope of the prediction track is thus obtained for each possible path of travel>A space-time distribution envelope for the predicted result of the kth feasible path.

An embodiment of a second aspect of the present disclosure provides a prediction calculation device for a traffic participant trajectory spatiotemporal distribution envelope, including:

the first module is configured to screen and obtain all feasible road center lines and the following lane center lines thereof from the map road center line information by comparing the historical track information of the traffic participants to be predicted with the map road center line information, and splice the center lines with topological relation to form a feasible running path set;

a second module configured to calculate longitudinal and lateral uncertainty parameters of each feasible travel path, respectively, using travel state features extracted from historical track information of traffic participants to be predicted, and the set of feasible travel paths obtained by screening;

a third module configured to calculate an operational state of a predicted trajectory end point of the traffic participant to be predicted, including coordinates, speed, and orientation angle, using operational state features, longitudinal uncertainty parameters, and each of the viable operational paths extracted from historical trajectory information of the traffic participant to be predicted;

A fourth module configured to fit and calculate the running track of the traffic participant to be predicted in the prediction time domain and output the running state of each moment in the prediction time domain;

and a fifth module configured to draw an uncertainty envelope of the travel track of the traffic participant to be predicted in the prediction time domain by using the transverse and longitudinal uncertainty parameters of each feasible travel path and adopting a geometric shape to obtain a space-time distribution envelope of the prediction track.

In order to implement the above embodiments, the present disclosure further proposes a computer-readable storage medium having stored thereon a computer program that is executed by a processor for executing the parameter calibration method of the driving risk model of an automatic driving automobile of the above embodiments.

Referring now to fig. 3, a schematic diagram of an electronic device suitable for use in implementing embodiments of the present disclosure is shown. It should be noted that, the electronic device in the embodiment of the present disclosure may include, but is not limited to, a mobile terminal such as a mobile phone, a notebook computer, a digital broadcast receiver, a PDA (personal digital assistant), a PAD (tablet computer), a PMP (portable multimedia player), a vehicle-mounted terminal (e.g., a vehicle-mounted navigation terminal), and the like, and a fixed terminal such as a digital TV, a desktop computer, a server, and the like. The electronic device shown in fig. 3 is merely an example and should not be construed to limit the functionality and scope of use of the disclosed embodiments.

As shown in fig. 3, the electronic device may include a processing means (e.g., a central processing unit, a graphics processor, etc.) 101, which may perform various appropriate actions and processes according to a program stored in a Read Only Memory (ROM) 102 or a program loaded from a storage means 108 into a Random Access Memory (RAM) 103. In the RAM103, various programs and data required for the operation of the electronic device are also stored. The processing device 101, ROM102, and RAM103 are connected to each other by a bus 104. An input/output (I/O) interface 105 is also connected to bus 104.

In general, the following devices may be connected to the I/O interface 105: input devices 106 including, for example, a touch screen, touchpad, keyboard, mouse, camera, microphone, etc.; an output device 107 including, for example, a Liquid Crystal Display (LCD), a speaker, a vibrator, and the like; storage devices 108 including, for example, magnetic tape, hard disk, etc.; and a communication device 109. The communication means 109 may allow the electronic device to communicate with other devices wirelessly or by wire to exchange data. While fig. 3 shows an electronic device having various means, it is to be understood that not all of the illustrated means are required to be implemented or provided. More or fewer devices may be implemented or provided instead.

In particular, according to embodiments of the present disclosure, the processes described above with reference to flowcharts may be implemented as computer software programs. For example, the present embodiment includes a computer program product comprising a computer program loaded on a computer readable medium, the computer program comprising program code for performing the method shown in the flowchart. In such an embodiment, the computer program may be downloaded and installed from a network via the communication means 109, or from the storage means 108, or from the ROM 102. The above-described functions defined in the methods of the embodiments of the present disclosure are performed when the computer program is executed by the processing device 101.

It should be noted that the computer readable medium described in the present disclosure may be a computer readable signal medium or a computer readable storage medium, or any combination of the two. The computer readable storage medium can be, for example, but not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or a combination of any of the foregoing. More specific examples of the computer-readable storage medium may include, but are not limited to: an electrical connection having one or more wires, a portable computer diskette, a hard disk, a Random Access Memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing. In the context of this disclosure, a computer-readable storage medium may be any tangible medium that can contain, or store a program for use by or in connection with an instruction execution system, apparatus, or device. In the present disclosure, however, the computer-readable signal medium may include a data signal propagated in baseband or as part of a carrier wave, with the computer-readable program code embodied therein. Such a propagated data signal may take any of a variety of forms, including, but not limited to, electro-magnetic, optical, or any suitable combination of the foregoing. A computer readable signal medium may also be any computer readable medium that is not a computer readable storage medium and that can communicate, propagate, or transport a program for use by or in connection with an instruction execution system, apparatus, or device. Program code embodied on a computer readable medium may be transmitted using any appropriate medium, including but not limited to: electrical wires, fiber optic cables, RF (radio frequency), and the like, or any suitable combination of the foregoing.

The computer readable medium may be contained in the electronic device; or may exist alone without being incorporated into the electronic device.

The computer readable medium carries one or more programs which, when executed by the electronic device, cause the electronic device to: comparing the historical track information of the traffic participant to be predicted with the central line information of the map road, screening and obtaining all feasible central lines of the road and the central lines of the subsequent lanes from the central line information of the map road, and splicing the central lines with topological relation to form a feasible running path set; calculating longitudinal and transverse uncertainty parameters of each feasible running path by utilizing running state characteristics extracted from historical track information of a traffic participant to be predicted and the feasible running path set obtained by screening; calculating the running state of the predicted track end point of the traffic participant to be predicted, including coordinates, speed and orientation angle, by using the running state characteristics, longitudinal uncertainty parameters and each feasible running path extracted from the historical track information of the traffic participant to be predicted; fitting and calculating the running track of the traffic participant to be predicted in the prediction time domain and outputting the running state of each moment in the prediction time domain; and drawing an uncertainty envelope of the running track of the traffic participant to be predicted in the prediction time domain by using the transverse and longitudinal uncertainty parameters of each feasible running path and adopting a geometric shape to obtain a space-time distribution envelope of the prediction track.

Computer program code for carrying out operations of the present disclosure may be written in one or more programming languages, including an object oriented programming language such as Java, smalltalk, C ++, python and conventional procedural programming languages, such as the "C" programming language or similar programming languages. The program code may execute entirely on the user's computer, partly on the user's computer, as a stand-alone software package, partly on the user's computer and partly on a remote computer or entirely on the remote computer or server. In the case of a remote computer, the remote computer may be connected to the user's computer through any kind of network, including a Local Area Network (LAN) or a Wide Area Network (WAN), or may be connected to an external computer (for example, through the Internet using an Internet service provider).

In the description of the present specification, a description referring to terms "one embodiment," "some embodiments," "examples," "specific examples," or "some examples," etc., means that a particular feature, structure, material, or characteristic described in connection with the embodiment or example is included in at least one embodiment or example of the present application. In this specification, schematic representations of the above terms are not necessarily directed to the same embodiment or example. Furthermore, the particular features, structures, materials, or characteristics described may be combined in any suitable manner in any one or more embodiments or examples. Furthermore, the different embodiments or examples described in this specification and the features of the different embodiments or examples may be combined and combined by those skilled in the art without contradiction.

Furthermore, the terms "first," "second," and the like, are used for descriptive purposes only and are not to be construed as indicating or implying a relative importance or implicitly indicating the number of technical features indicated. Thus, a feature defining "a first" or "a second" may explicitly or implicitly include at least one such feature. In the description of the present application, the meaning of "plurality" is at least two, such as two, three, etc., unless explicitly defined otherwise.

Any process or method descriptions in flow charts or otherwise described herein may be understood as representing modules, segments, or portions of code which include one or more executable instructions for implementing specific logical functions or steps of the process, and further implementations are included within the scope of the preferred embodiment of the present application in which functions may be executed out of order from that shown or discussed, including substantially concurrently or in reverse order, depending on the functionality involved, as would be understood by those reasonably skilled in the art of the embodiments of the present application.

Logic and/or steps represented in the flowcharts or otherwise described herein, e.g., a ordered listing of executable instructions for implementing logical functions, can be embodied in any computer-readable medium for use by or in connection with an instruction execution system, apparatus, or device, such as a computer-based system, processor-containing system, or other system that can fetch the instructions from the instruction execution system, apparatus, or device and execute the instructions. For the purposes of this description, a "computer-readable medium" can be any means that can contain, store, communicate, propagate, or transport the program for use by or in connection with the instruction execution system, apparatus, or device. More specific examples (a non-exhaustive list) of the computer-readable medium would include the following: an electrical connection (electronic device) having one or more wires, a portable computer diskette (magnetic device), a Random Access Memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber device, and a portable compact disc read-only memory (CDROM). Additionally, the computer-readable medium may even be paper or other suitable medium upon which the program is printed, as the program may be electronically captured, via, for instance, optical scanning of the paper or other medium, then compiled, interpreted or otherwise processed in a suitable manner, if necessary, and then stored in a computer memory.

It is to be understood that portions of the present application may be implemented in hardware, software, firmware, or a combination thereof. In the above-described embodiments, the various steps or methods may be implemented in software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented in hardware, as in another embodiment, may be implemented using any one or combination of the following techniques, as is well known in the art: discrete logic circuits having logic gates for implementing logic functions on data signals, application specific integrated circuits having suitable combinational logic gates, programmable Gate Arrays (PGAs), field Programmable Gate Arrays (FPGAs), and the like.

Those of ordinary skill in the art will appreciate that implementing all or part of the steps carried by the method of the above embodiments may be accomplished by a program to instruct related hardware and the developed program may be stored in a computer readable storage medium, which when executed, includes one or a combination of the steps of the method embodiments.

In addition, each functional unit in each embodiment of the present application may be integrated in one processing module, or each unit may exist alone physically, or two or more units may be integrated in one module. The integrated modules may be implemented in hardware or in software functional modules. The integrated modules may also be stored in a computer readable storage medium if implemented as software functional modules and sold or used as a stand-alone product.

The above-mentioned storage medium may be a read-only memory, a magnetic disk or an optical disk, or the like. Although embodiments of the present application have been shown and described above, it will be understood that the above embodiments are illustrative and not to be construed as limiting the application, and that variations, modifications, alternatives, and variations may be made to the above embodiments by one of ordinary skill in the art within the scope of the application.

Claims (10)

1. A predictive computation method for a traffic participant trajectory spatiotemporal distribution envelope, comprising:

s1, screening feasible operation paths: comparing the historical track information of the traffic participant to be predicted with the central line information of the map road, screening and obtaining all feasible central lines of the road and the central lines of the subsequent lanes from the central line information of the map road, and splicing the central lines with topological relation to form a feasible running path set;

s2, calculating longitudinal and transverse uncertainty parameters: calculating longitudinal and transverse uncertainty parameters of each feasible running path by utilizing running state characteristics extracted from historical track information of a traffic participant to be predicted and the feasible running path set obtained by screening;

s3, determining a predicted endpoint state: calculating the running state of the predicted track end point of the traffic participant to be predicted, including coordinates, speed and orientation angle, by using the running state characteristics, longitudinal uncertainty parameters and each feasible running path extracted from the historical track information of the traffic participant to be predicted;

S4, fitting the running track of the traffic participant to be predicted: fitting and calculating the running track of the traffic participant to be predicted in the prediction time domain and outputting the running state of each moment in the prediction time domain;

s5, calculating a space-time distribution envelope of the predicted track: and drawing an uncertainty envelope of the running track of the traffic participant to be predicted in the prediction time domain by using the transverse and longitudinal uncertainty parameters of each feasible running path and adopting a geometric shape to obtain a space-time distribution envelope of the prediction track.

2. The predictive computation method of claim 1, wherein S1 specifically comprises:

s11, preprocessing historical track information of a traffic participant to be predicted, including denoising and extracting running state characteristics of the traffic participant to be predicted, wherein the running state characteristics include average speed, running direction and transverse and longitudinal acceleration of the historical track of the traffic participant to be predicted;

s12, screening a lane center line closest to the historical track of the traffic participant to be predicted at the current moment by using the preprocessed historical track information to serve as a candidate lane center line;

s13, traversing the center line of the candidate lane, and acquiring the center line of the subsequent lane by using map topology information until the farthest distance exceeds a set distance threshold L;

S14, splicing lane central line information according to the connection relation, removing duplication, and obtaining a feasible running path, so that a feasible running path set is formed.

3. The predictive computation method of claim 2, wherein in S13, a depth-first or breadth-first search algorithm is employed to traverse the candidate lane centerlines.

4. The predictive computing method of claim 1, wherein the historical trajectory information of the traffic participant to be predicted includes longitudinal and lateral coordinates, speed, and heading angle of the traffic participant to be predicted in a historical time domain.

5. The predictive computation method of claim 1, wherein the map road centerline information includes a number of each lane centerline, a number of a lane centerline subsequent to each lane centerline, longitudinal and transverse coordinate vectors of each lane centerline, a traffic light status within each lane, and a sign of whether a lane passes an intersection.

6. The predictive computation method of claim 1, wherein in S2, the longitudinal uncertainty parameter for each feasible travel path includes an expected longitudinal acceleration and an ending longitudinal distance deviation, calculated according to the following equation:

Wherein,,for the expected longitudinal acceleration of the first possible path of travel, +.>Deviation of the longitudinal distance for the end of each possible path of travel,/->And->Longitudinal acceleration and average speed, respectively, derived from historical trajectory information of the traffic participant to be predicted,/->Signal lamp status for the first possible path, +.>For the first feasible path to pass through the sign of the intersection, alpha _px 、β _px 、γ _px The first, second and third influence coefficients are respectively.

7. The predictive computation method of claim 1, wherein in S2, the lateral uncertainty parameter for each feasible travel path includes an expected lateral acceleration and an end-point lateral distance deviation, calculated according to the following equation:

wherein,,and->Expected lateral acceleration and end-point lateral distance deviation, +.>And->Lateral acceleration and average speed, beta, respectively, derived from historical trajectory information of the traffic participant to be predicted _py For the end point transverse fixed deviation alpha _py 、γ _py 、δ _py Respectively, fourth, fifth and sixthInfluence coefficient.

8. The predictive computation method as recited in claim 1, wherein S3 specifically comprises:

s31, determining curvature adjustment parameters: for each feasible travel path, a turning curvature ρ based on each feasible travel path ^l Direction of turning theta ^l And a two-sided conflict object number ratio τ ^l Comprehensive determination of curvature adjustment parameters for each predicted trajectoryThe calculation formulas are respectively as follows:

wherein,,and->The starting point and the end point of the first feasible running path face angles respectively; />And->Longitudinal and transverse coordinates of the start point of the first feasible path, respectively, < >>And->Longitudinal and transverse coordinates of the end point of the first feasible travel path, respectively, and provide: if theta is ^l >0, turn left, if θ ^l <0, turn right if θ ^l =0, then straight; />And->The number of traffic participants with transverse conflict on the left side and the right side of the first feasible running path respectively; alpha _ρ And beta _θ The seventh influence coefficient and the eighth influence coefficient are respectively, and SGN is a sign function; />For the expected lateral acceleration of each possible travel path.

S32, calculating a longitudinal running distance in a prediction time domain: adjusting parameters based on curvature of each predicted trajectoryLongitudinal movement distance +.f. of each predicted track in the prediction time domain is calculated by Newton kinematics respectively>

S33, interpolation is carried out on coordinate points on each feasible running path respectively, and a path coordinate point closest to the coordinates of the traffic participant to be predicted at the current moment is searched;

s34, obtaining the state of the prediction end point along the path direction: starting from the nearest path coordinate point, searching for the end point path coordinate, speed and orientation angle at the predicted distance along the corresponding feasible running path direction, and jointly forming the state of the predicted track end point of the traffic participant to be predicted.

9. The predictive computation method of claim 1, wherein in S4, a cubic curve is used And fitting and calculating the running track of the traffic participant to be predicted in the prediction time domain, and solving the cubic curve parameters by considering the starting and ending point constraint as follows:

wherein a is ₀ Constant term of cubic curve, a ₁ 、a ₂ And a ₃ Each coefficient of the cubic curve; v _t For the speed of the traffic participant to be predicted in the world coordinate system at time t, v _T The speed of a traffic participant to be predicted in a world coordinate system at the time t+T, wherein T is the duration of a prediction time domain; p is p _t Is the coordinate of the traffic participant to be predicted under the world coordinate system at the moment t, p _T To be the traffic participant at time T + T.

10. The predictive computation method of claim 1, wherein S5 specifically comprises:

and for each feasible running path, adopting an envelope form of geometric shapes according to the end longitudinal distance deviation of each feasible running path, and reversely pushing longitudinal envelope parameters of each prediction moment on the corresponding feasible running path by a linear or nonlinear method to obtain the space-time distribution envelope of the prediction track.