Trajectory prediction method, device and vehicle

Abstract

The present application relates to the technical field of self-vehicle trajectory prediction, and discloses a trajectory prediction method, device and vehicle, wherein the present application obtains the state information of the self-vehicle at the current time, lane line information and the state information of the self-vehicle corresponding to the first, second and third control points in the Bezier curve corresponding to the first prediction time period, calculates the second state information of the self-vehicle at the second preset time period from the current time based on the lane line information, then calculates the state information of the fourth control point in the Bezier curve based on the state information corresponding to the first, second and third control points and the second state information of the self-vehicle, and finally performs trajectory prediction on the first prediction time period based on the state information of the self-vehicle corresponding to the first, second, third and fourth control points to obtain the first prediction trajectory segment, thereby combining the self-vehicle information with the lane line information, reducing the dependence on the map, improving the accuracy of the trajectory prediction of the self-vehicle and improving the driving function safety.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of self-vehicle trajectory prediction, in particular to a trajectory prediction method, device and vehicle. BACKGROUND

[0002] The prediction of the trajectory of the self-vehicle, the forward collision warning system can effectively identify the driving intention and driving direction of the driver, play a greater role in the selection of the main target, effectively avoid the false triggering in the forward collision warning and improve the positive triggering, and increase the safety in the driving process.

[0003] However, most of the existing schemes predict the trajectory of the self-vehicle by processing the high-precision map and the obstacles around the self-vehicle, and combining the last frame of trajectory information, so as to improve the accuracy of the trajectory prediction of the self-vehicle, but the accuracy of the predicted trajectory cannot be guaranteed on general roads or under non-high-precision maps. SUMMARY

[0004] Therefore, the present application provides a trajectory prediction method, device and vehicle to solve the problem that the accuracy of the predicted trajectory cannot be guaranteed in the conventional way of relying on high-precision maps.

[0005] In a first aspect, the present application provides a trajectory prediction method, the method comprising: obtaining first ego vehicle state information at a current time, lane line information at the current time, and ego vehicle state information corresponding to a first control point, a second control point, and a third control point in a first Bezier curve of the ego vehicle at a first prediction time period; wherein the first control point is a starting point of the first Bezier curve, the second control point is a tuning point on the side of the starting point, and the third control point is a tuning point on the side of an ending point; determining whether road prediction is valid based on the first ego vehicle state information and the lane line information at the current time; if the road prediction is valid, calculating second ego vehicle state information at a second predetermined time period ahead of the current time based on the lane line information, the second predetermined time period being greater than the first prediction time period and being an integer multiple of the first prediction time period; calculating ego vehicle state information of a fourth control point in the first Bezier curve based on the ego vehicle state information corresponding to the first control point, the second control point, and the third control point, and the second ego vehicle state information; the fourth control point being the ending point of the first Bezier curve; if the road prediction is invalid, calculating the ego vehicle state information of the fourth control point based on the first ego vehicle state information at the current time and the first prediction time period; the first ego vehicle state information comprising at least curvature information, speed, acceleration, and position information of the ego vehicle, the curvature information comprising at least curvature and curvature change rate of the ego vehicle, and the calculation of the ego vehicle state information of the fourth control point based on the first ego vehicle state information at the current time and the first prediction time period comprising: determining whether the curvature of the ego vehicle at the current time is less than a predetermined curvature threshold; when the curvature of the ego vehicle at the current time is less than the predetermined curvature threshold, calculating the ego vehicle state information of the fourth control point based on the position information, speed, acceleration, and first prediction time period of the ego vehicle at the current time; or when the curvature of the ego vehicle at the current time is not less than the predetermined curvature threshold, calculating the ego vehicle state information of the fourth control point based on the curvature and curvature change rate of the ego vehicle, the first prediction time period, and the speed, acceleration, and position information of the ego vehicle at the current time; and performing trajectory prediction on the first prediction time period based on the ego vehicle state information corresponding to the first control point, the second control point, the third control point, and the fourth control point, to obtain a first prediction trajectory segment.

[0006] The trajectory prediction method provided by the embodiment combines the ego vehicle information and the lane line information, reduces the dependence on the map, improves the accuracy of the ego vehicle trajectory prediction, improves the driving function safety, and if it is determined that the road prediction is invalid, the ego vehicle state information can be used for trajectory prediction, thereby ensuring the reliability of the autonomous driving of the trajectory prediction.

[0007] In an optional implementation, the second ego vehicle state information includes a first lateral position, a first lateral speed and a first lateral acceleration of the ego vehicle at a second preset time period ahead of the current time, the lane line information includes at least a lane line orientation angle, a lane line length, a lane line offset and lane line curvature information, and the calculation of the second ego vehicle state information at the second preset time period ahead of the current time based on the lane line information includes: taking the lane line orientation angle, the lane line offset and the lane line curvature information as lane line coefficients in a road polynomial model, taking the lane line length as a variable value, and inputting the road polynomial model to obtain the first lateral position of the ego vehicle at the second preset time period ahead of the current time; determining a vehicle head orientation angle based on the lane line orientation angle, the lane line curvature information and the lane line length; multiplying the vehicle head orientation angle and the longitudinal speed of the ego vehicle to obtain the first lateral speed of the ego vehicle at the second preset time period ahead of the current time; determining a first curvature based on the lane line curvature information and the lane line length; and multiplying the first curvature and the longitudinal speed of the ego vehicle to obtain the first lateral acceleration of the ego vehicle at the second preset time period ahead of the current time.

[0008] The present application converts the lane line features into the coefficients of the road polynomial model, and calculates the second ego vehicle state information at the second preset time period ahead of the current time through the road constraint, so as to identify the lane deviation risk and other problems in advance and reserve sufficient adjustment time.

[0009] In an optional implementation, the first ego vehicle state information at the current moment includes at least a speed, an acceleration and a position information of the ego vehicle, and before the step of calculating the second ego vehicle state information at a second preset time period ahead of the current moment based on the lane line information, the method further includes: detecting whether the speed of the ego vehicle at the current moment is greater than a first preset speed threshold; dividing the speed of the ego vehicle at the current moment by the acceleration of the ego vehicle to obtain a first time length; determining whether the first time length is within a first preset time length range; if the speed of the ego vehicle at the current moment is greater than the first preset speed threshold and the first time length is not within the first preset time length range, determining that the ego vehicle does not meet a stop condition in the first prediction time period, and performing the step of calculating the second ego vehicle state information at the second preset time period ahead of the current moment based on the lane line information.

[0010] The application determines the stop feasibility of the ego vehicle in the first prediction time period through the speed threshold and the time length range, and then calculates the ego vehicle state information of the corresponding control point based on the lane line information for the case that does not meet the stop condition, predicts the safety path in this case, and guarantees the driving safety in different scenarios.

[0011] In an optional implementation, the first ego vehicle state information includes at least a position information of the ego vehicle, and the position information includes a lateral position and a longitudinal position, and the calculating of the ego vehicle state information of the fourth control point in the first Bezier curve based on the ego vehicle state information corresponding to the first control point, the second control point and the third control point respectively and the second ego vehicle state information includes: calculating a lateral position of the fourth control point based on the first lateral acceleration, the first lateral speed, a lateral position of the third control point, a lateral position of the second control point and the first lateral position; calculating a longitudinal position of the fourth control point based on the longitudinal position of the third control point, a starting speed of the first control point, a starting acceleration, the lateral position of the fourth control point and the lateral position of the third control point; and calculating a speed and an acceleration of the fourth control point based on the position information of the second control point, the third control point and the fourth control point.

[0012] The application gradually deduces the information of the fourth control point based on the ego vehicle state information of the previous control point and using the ego vehicle state information at a preset time period as a constraint, improves the continuity and accuracy of the prediction of the fourth control point parameters, and then improves the fitting accuracy of the entire Bezier curve to the preset path.

[0013] In an optional implementation, the method further comprises: if the speed of the ego vehicle at the current time is not greater than a first preset speed threshold, or the first time length is within a first preset time length range, determining that the ego vehicle satisfies the stop condition in the first prediction time period, determining the position information of the fourth control point as the position information of the third control point, and calculating the speed and acceleration of the fourth control point based on the position information of the second control point, the third control point and the fourth control point.

[0014] In an optional implementation, the ego vehicle state information corresponding to the first control point, the second control point and the third control point in the first Bezier curve in the first prediction time period is acquired, comprising: taking the first ego vehicle state information at the current time as the ego vehicle state information of the first control point in the first Bezier curve; when the speed of the ego vehicle at the current time is greater than a first preset speed threshold and the first time length is not within a first preset time length range, determining that the ego vehicle does not satisfy the stop condition in the first prediction time period, and calculating the position information of the ego vehicle at the second control point based on the position information and the speed of the ego vehicle at the first control point; and calculating the position information of the ego vehicle at the third control point based on the position information of the ego vehicle at the second control point, the position information of the ego vehicle at the first control point and the acceleration.

[0015] In an optional implementation, the method further comprises: when the speed of the ego vehicle at the current time is not greater than a first preset speed threshold, or the first time length is within a first preset time length range, determining that the ego vehicle satisfies the stop condition in the first prediction time period, calculating the distance between the ego vehicle and the first control point at the stop time based on the first time length, the speed and the acceleration of the ego vehicle at the first control point; and determining the position information of the second control point and the third control point based on the position information of the ego vehicle at the first control point and the distance between the ego vehicle and the first control point at the stop time.

[0016] The present application directly calculates the ego vehicle state information based on the existing stop condition when the ego vehicle satisfies the stop condition, avoids unnecessary resource consumption such as lane line information analysis, and improves the efficiency of trajectory prediction.

[0017] In an optional implementation, the lane line information at least comprises a lane line length, and the method of judging whether the road prediction is effective based on the first ego vehicle state information and the lane line information at the current time comprises: judging whether the speed of the ego vehicle at the current time is greater than a second preset speed threshold; if the speed of the ego vehicle at the current time is greater than the second preset speed threshold, multiplying the speed of the ego vehicle at the current time by a preset second time length to obtain a first distance; judging whether the first distance is greater than the lane line length; if the first distance is not greater than the lane line length, judging whether the target acceleration of the ego vehicle satisfies an acceleration preset condition; and if the target acceleration of the ego vehicle satisfies the acceleration preset condition, determining that the road prediction is effective.

[0018] The application determines that it is effective to calculate the ego vehicle trajectory based on the lane line information only when the current ego vehicle and road information meet the conditions, and then calculates the ego vehicle state information of the fourth control point based on the lane line information, thereby avoiding invalid calculation or wrong decision, and ensuring the reliability and accuracy of the preview result and the ego vehicle state information calculation.

[0019] In an optional embodiment, the ego vehicle state information of the fourth control point is calculated based on the curvature and the rate of change of curvature of the ego vehicle, the speed, acceleration and position information of the ego vehicle at the current time and the first prediction time period, including: taking the reciprocal of the curvature of the ego vehicle at the current time as the ego vehicle driving radius; multiplying the rate of change of curvature of the ego vehicle at the current time by the first prediction time period to obtain a second curvature; adding the second curvature and the curvature of the ego vehicle at the current time to obtain a curvature sum; taking the reciprocal of the curvature sum as the ego vehicle predicted driving radius; calculating the ego vehicle predicted driving distance based on the speed, acceleration of the ego vehicle at the current time and the first prediction time period; dividing the ego vehicle predicted driving distance by the ego vehicle driving radius to obtain a first angle value; dividing the ego vehicle predicted driving distance by the ego vehicle predicted driving radius to obtain a second angle value; calculating the ego vehicle driving first angle based on the first angle value and the ego vehicle driving radius; calculating the ego vehicle driving second angle based on the second angle value and the ego vehicle predicted driving radius; calculating the position information of the fourth control point based on the ego vehicle driving first angle, the ego vehicle driving second angle, the angle preset coefficient corresponding to each of the angles and the position information of the ego vehicle at the current time; and calculating the speed and acceleration of the fourth control point based on the position information of the second control point, the third control point and the fourth control point.

[0020] The application can determine the steering dynamic trend based on the curvature and the rate of change of curvature of the ego vehicle when it is determined that the vehicle meets the curve driving, and then realize smooth transition through double-angle fusion, realize accurate adaptation to the curve driving demand, and improve the accuracy and smoothness of curve path prediction.

[0021] In an optional embodiment, the second preset time period is a third prediction time period from the current time, and whether the target acceleration of the ego vehicle satisfies the acceleration preset condition is determined by the following steps: taking the first ego vehicle state information at the current time as the ego vehicle state information of a first control point in the first Bezier curve; taking the second ego vehicle state information as the ego vehicle state information of a fourth control point in a third Bezier curve corresponding to the third prediction time period; calculating the ego vehicle state information of the fourth control point in a second Bezier curve corresponding to the second prediction time period based on the ego vehicle state information of the fourth control point in the third Bezier curve, the acceleration of the ego vehicle at the first control point in the first Bezier curve, and the position information of the fourth control point in the first Bezier curve; calculating the ego vehicle state information of the fourth control point in the first Bezier curve corresponding to the first prediction time period based on the ego vehicle state information of the fourth control point in the third Bezier curve and the acceleration of the ego vehicle at the first control point in the first Bezier curve; wherein the ego vehicle state information of the fourth control point in the first Bezier curve is the ego vehicle state information of the first control point in the second Bezier curve; the ego vehicle state information of the fourth control point in the second Bezier curve is the ego vehicle state information of the first control point in the third Bezier curve; determining the ego vehicle state information of the second control point and the third control point of each Bezier curve based on the ego vehicle state information of the first control point in each Bezier curve and the ego vehicle state information of the fourth control point of each Bezier curve; determining the acceleration value set corresponding to each Bezier curve based on the ego vehicle state information calculated respectively by the second control point, the third control point and the fourth control point of each Bezier curve; combining the acceleration sets corresponding to all Bezier curves into a first acceleration set, and selecting the maximum acceleration in the first acceleration set corresponding to all Bezier curves; determining whether the maximum acceleration is less than a first preset acceleration threshold; subtracting the lateral acceleration of the ego vehicle at the current time from the maximum acceleration in the acceleration set corresponding to the first Bezier curve to obtain a first difference value, and determining whether the first difference value is less than a first preset difference threshold; if the maximum acceleration is less than the first preset acceleration threshold and the first difference value is less than the first preset difference threshold, it is determined that the target acceleration of the ego vehicle satisfies the acceleration preset condition.

[0022] The application determines whether the target acceleration satisfies the acceleration preset condition by constructing and verifying the multi-dimensional acceleration based on the Bezier curve segmentation, which ensures that the acceleration of the vehicle in the preview time period is always within the safety and comfort threshold, and avoids the risk of acceleration exceeding the limit at a single time point.

[0023] In an alternative embodiment, the lane line information at least includes a lane line orientation angle, a lane line length, a lane line offset, a lane line curvature and a lane line curvature change rate, and the method further comprises: obtaining the lane line information corresponding to the second prediction time period by the following steps: calculating the square of the lane line length at the current time to obtain a first square value, and calculating the cube of the lane line length to obtain a first cube value; multiplying the lane line orientation angle at the current time by the lane line length to obtain a first product, and multiplying the lane line curvature at the current time by the first square value to obtain a second product; multiplying the lane line curvature change rate at the current time by the first cube value to obtain a third product; adding the first product, the second product, the third product and the lane line offset at the current time to obtain the lane line offset corresponding to the second prediction time period; multiplying the first preset coefficient, the lane line curvature at the current time and the lane line length to obtain a fourth product; multiplying the second preset coefficient, the lane line curvature change rate at the current time and the first product to obtain a fifth product; adding the fourth product, the fifth product and the orientation angle at the current time to obtain the lane line orientation angle corresponding to the second prediction time period; multiplying the third preset coefficient by the lane line curvature at the current time to obtain a sixth product; multiplying the lane line length at the current time, the fourth preset coefficient and the lane line curvature change rate at the current time to obtain the lane line curvature corresponding to the second prediction time period; taking the lane line curvature change rate at the current time as the lane line curvature change rate corresponding to the second prediction time period; and taking the lane line length at the current time as the lane line length corresponding to the second prediction time period.

[0024] The present application calculates the lane line information corresponding to the next prediction time period based on the lane line information corresponding to the previous prediction time period, to generate continuous lane line information and smooth the transition of lane line features, and to provide homologous and continuous constraint references for the Bezier curves corresponding to different time periods.

[0025] In an alternative embodiment, the ego vehicle state information at least includes the speed, acceleration, position information, curvature and curvature change rate of the ego vehicle, and the method further comprises: obtaining the third ego vehicle state information corresponding to the second prediction time period by the following steps: obtaining the speed of the ego vehicle corresponding to the second prediction time period based on the speed and acceleration of the ego vehicle at the first prediction time period and the current time; taking the acceleration of the ego vehicle at the current time as the acceleration of the ego vehicle corresponding to the second prediction time period; taking the position information of the fourth control point of the ego vehicle in the first Bezier curve as the position information of the ego vehicle corresponding to the second prediction time period; multiplying the curvature change rate at the current time by the first prediction time period to obtain a third curvature; and adding the third curvature and the curvature at the current time to obtain the curvature of the ego vehicle corresponding to the second prediction time period.

[0026] The application updates the ego vehicle state information corresponding to the second prediction time period based on the first ego vehicle state information and the information of the fourth control point in the first Bezier curve, realizes the continuity with the previous trajectory, and the physical consistency among parameters, and provides reliable parameter input for subsequent trajectory segment calculation.

[0027] In an optional embodiment, the method further comprises: updating the first ego vehicle state information at the current time and the lane line information at the current time to the third ego vehicle state information and the lane line information corresponding to the second prediction time period, and performing the step of obtaining the ego vehicle state information corresponding to the first control point, the second control point, and the third control point in the first Bezier curve corresponding to the first prediction time period until the second prediction trajectory segment corresponding to the second prediction time period is obtained.

[0028] The application generates a set of continuous and reliable long-time sequence prediction trajectory system through the iterative generation of multi-segment Bezier curve, combines the road model constraint and real-time state update, and improves the stability of vehicle driving.

[0029] In a second aspect, the present application provides a trajectory prediction device, the device comprising: an information acquisition module, configured to acquire first ego vehicle state information at a current time, lane line information at the current time, and ego vehicle state information corresponding to a first control point, a second control point, and a third control point in a first Bezier curve of the ego vehicle at a first prediction time period; the first control point is a starting point of the first Bezier curve, the second control point is a tuning point on the side of the starting point, and the third control point is a tuning point on the side of an ending point; a road validity judgment module, configured to judge whether road prediction is valid based on the first ego vehicle state information and the lane line information at the current time; a second ego vehicle state information acquisition module, configured to, if the road prediction is valid, calculate second ego vehicle state information at a second preset time period from the current time based on the lane line information; the second preset time period is greater than the first prediction time period, and the second preset time period is an integer multiple of the first prediction time period; a fourth control point information calculation module, configured to calculate ego vehicle state information of a fourth control point of the ego vehicle in the first Bezier curve based on the ego vehicle state information corresponding to the first control point, the second control point, and the third control point and the second ego vehicle state information; the fourth control point is an ending point of the first Bezier curve; an ego vehicle state prediction module, configured to, if the road prediction is invalid, calculate the ego vehicle state information of the fourth control point based on the first ego vehicle state information at the current time and the first prediction time period; the first ego vehicle state information at least includes curvature information, speed, acceleration, and position information of the ego vehicle, and the curvature information at least includes curvature and curvature change rate of the ego vehicle; the ego vehicle state prediction module comprises: a curvature judgment unit, configured to judge whether the curvature of the ego vehicle at the current time is less than a preset curvature threshold; a state information calculation unit, configured to, when the curvature of the ego vehicle at the current time is less than the preset curvature threshold, calculate the ego vehicle state information of the fourth control point based on position information, speed, acceleration, and the first prediction time period of the ego vehicle at the current time; or, when the curvature of the ego vehicle at the current time is not less than the preset curvature threshold, calculate the ego vehicle state information of the fourth control point based on the curvature and the curvature change rate of the ego vehicle, the first prediction time period, and the speed, acceleration, and position information of the ego vehicle at the current time; and a trajectory segment prediction module, configured to perform trajectory prediction on the first prediction time period based on the ego vehicle state information corresponding to the first control point, the second control point, the third control point, and the fourth control point, to obtain a first prediction trajectory segment.

[0030] In a third aspect, the present application provides a vehicle, the vehicle comprising a controller, the controller comprising: a memory and a processor, which are communicatively connected to each other, and the memory stores computer instructions; the processor executes the computer instructions to perform the trajectory prediction method in the first aspect or any one of the corresponding embodiments thereof.

[0031] The present application has the following technical effects:

[0032] The trajectory prediction method provided by the embodiment combines the ego vehicle information and the lane line information, reduces the dependence on the map, improves the accuracy of the ego vehicle trajectory prediction, improves the driving function safety, and if it is determined that the road prediction is invalid, the ego vehicle state information can be used for trajectory prediction, ensuring the reliability of the trajectory prediction of the autonomous driving. BRIEF DESCRIPTION OF DRAWINGS

[0033] In order to more clearly illustrate the specific embodiments of the present application or the technical solutions in the prior art, the following will briefly introduce the drawings needed to be used in the specific embodiments or the prior art description. Obviously, the drawings in the following description are some embodiments of the present application, and those skilled in the art can also obtain other drawings according to these drawings without creative labor.

[0034] Figure 1 is a flowchart of the trajectory prediction method according to the embodiment of the present application;

[0035] Figure 2 is an example diagram of another Bezier curve according to the embodiment of the present application;

[0036] Figure 3 is a flowchart of another trajectory prediction method according to the embodiment of the present application;

[0037] Figure 4 is a calculation method example diagram of calculating the Bezier curve control point information according to the embodiment of the present application;

[0038] Figure 5 is an example diagram of deriving the Bezier curve based on the ego vehicle state information at the second preset time period of the preview and the first ego vehicle state information at the current time according to the embodiment of the present application;

[0039] Figure 6 is an ego vehicle state information prediction example diagram according to the embodiment of the present application;

[0040] Figure 7 is a flowchart of a Bezier curve prediction according to an embodiment of the present application;

[0041] Figure 8 is an example diagram of a four-segment trajectory segment according to an embodiment of the present application;

[0042] Figure 9 is a structural block diagram of a vehicle according to an embodiment of the present application;

[0043] Figure 10 is a structural block diagram of a trajectory prediction device according to an embodiment of the present application;

[0044] Figure 11 is a hardware structural schematic diagram of a controller according to an embodiment of the present application. DETAILED DESCRIPTION

[0045] In order to make the objects, technical solutions and advantages of the embodiments of the present application clearer, the technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are some but not all of the embodiments of the present application. Based on the embodiments in the present application, all other embodiments obtained by a person of ordinary skill in the art without creative work fall within the protection scope of the present application.

[0046] According to an embodiment of the present application, a trajectory prediction method embodiment is provided. It should be noted that the steps shown in the flowchart of the drawings can be executed in a computer system such as a group of computer executable instructions, and although a logical order is shown in the flowchart, in some cases, the steps shown or described herein can be executed in an order different from that shown herein.

[0047] In the present embodiment, a trajectory prediction method is provided, which can be used in a controller in a vehicle, Figure 1 is a flowchart of a trajectory prediction method according to an embodiment of the present application, as Figure 1 shown, the flow includes the following steps:

[0048] In step S101, the first ego vehicle state information at the current time, the lane line information at the current time, and the ego vehicle state information corresponding to the first control point, the second control point and the third control point in the first Bezier curve of the ego vehicle in the first prediction time period are obtained.

[0049] Among them, the first control point is the starting point of the first Bezier curve, the second control point is the adjustment node on the starting point side, and the third control point is the adjustment node on the ending point side.

[0050] The embodiment of the application represents the current time as T0, and first ego vehicle state information and lane line information of T0 can be acquired, wherein the first ego vehicle state information includes but is not limited to the speed, acceleration information, position information and curvature information of the ego vehicle, etc.; the lane line information can be acquired according to a front-view camera of the vehicle, which includes but is not limited to lane line offset, lane line orientation angle, lane line curvature, lane line curvature change rate and lane line length, etc., which are only examples; and a prediction time period can be determined according to a vehicle driving scene, and a short-term prediction (such as 0-1s corresponding to a first prediction time period (T0 is 0s), 1-2s corresponding to a second prediction time period) is used for instant trajectory tracking, which is only an example.

[0051] As shown in Figure 2 The Bezier curve is controlled by a starting point (a first control point P0), an ending point (a fourth control point P3) and intermediate points (a second control point P1 and a third control point P2), and the following formula is referred to:

[0052]

[0053] Wherein B(t) represents the calculation result of the formula; the standard value interval of t is [0, 1], when t=0, the calculation result of the formula corresponds to the starting point of the Bezier curve, and when t=1, the calculation result of the formula corresponds to the ending point of the Bezier curve.

[0054] And the high-order Bezier curve formula can be derived according to the above formula, and the key point of the segmented Bezier curve is the continuity of the segmentation point, and in trajectory prediction, second-order continuous derivable needs to be realized, and the first derivative of the third-order Bezier curve can be represented by the following formula:

[0055]

[0056] The second derivative of the third-order Bezier curve can be represented by the following formula:

[0057]

[0058] In order to ensure the effectiveness of the segmented Bezier curve, when all the control points of a certain segment of the Bezier curve in front are completely determined, the subsequent adjacent Bezier curve needs to be associated with the previous segment through continuity constraint, in order to meet the value continuity, the starting control point P0 of the subsequent segment needs to be directly coincided with the ending control point of the previous segment, and the position thereof is determined by the ending point of the previous segment, so as to achieve the basic continuity condition.

[0059] And in the multi-segment Bezier curve with first-order and second-order continuous derivable, the position, speed and acceleration information of the ending point of the previous segment of the Bezier curve are the same as the position, speed and acceleration information of the starting point of the adjacent next segment of the Bezier curve.

[0060] The embodiment of the present application can directly take the first ego vehicle state at T0 moment as the ego vehicle state information corresponding to the first control point P0 in the first Bezier curve; then the position information of the second control point P1 can be calculated based on the position information, corresponding speed and acceleration of the ego vehicle at T0 moment; and then the position information of the third control point P2 can be calculated based on the position information of the second control point P1, the position information of the first control point P0, and corresponding speed and acceleration, wherein the position information includes lateral position and longitudinal position, which are only examples and are not limited.

[0061] In step S102, whether the road prediction is valid is judged based on the first ego vehicle state information and the lane line information at the current moment.

[0062] The embodiment of the present application can judge the distance walked by the ego vehicle in a certain time period based on the ego vehicle speed in the ego vehicle state information, then obtain the lane line length in the lane line information, and compare the distance walked by the ego vehicle with the lane line length to judge whether the road prediction is valid.

[0063] In step S103, if the road prediction is valid, the second ego vehicle state information at the second preset time period ahead of the current moment is calculated based on the lane line information.

[0064] The second preset time period is greater than the first prediction time period, and the second preset time period is an integer multiple of the first prediction time period.

[0065] If the distance walked by the ego vehicle is less than the lane line length, it means that the road prediction is valid, and when the trajectory prediction for the first prediction time period is performed, the second ego vehicle state information at the second preset time period ahead of T0 moment can be calculated, the second preset time period is 3s, and the second ego vehicle state information can include speed, acceleration and position, which are only examples. Among them, the mathematical model of the lane center line can be determined first, then the obtained lane line information is taken as the coefficient of the mathematical model of the lane line, and the target position (i.e. the position in the second ego vehicle state information) is calculated; the speed of the ego vehicle at the second preset time period ahead is determined based on the acceleration constraint and the curvature of the lane center line, and then the acceleration of the ego vehicle at the second preset time period ahead is calculated based on the speed change and the curvature of the lane line, and finally the second ego vehicle state information is composed of the position, speed and acceleration at the second preset time period ahead, which are only examples.

[0066] In step S104, the ego vehicle state information of the fourth control point in the first Bezier curve is calculated based on the ego vehicle state information corresponding to the first control point, the second control point and the third control point, and the second ego vehicle state information.

[0067] The fourth control point is the end point of the first Bezier curve.

[0068] The embodiment of the application can be based on the first control point P0, the second control point P1, and the third control point P2 corresponding to the self-vehicle state information, combined with the constraint of the second self-vehicle state information (the self-vehicle state at the second preset time period), and the mathematical characteristics of the Bezier curve and the continuity requirement of the self-vehicle motion, to calculate the self-vehicle state information of the fourth control point in the first Bezier curve. For example, based on the Bezier curve parameter equation, the constraint of the position information of the fourth control point P3 and the second self-vehicle state information is associated, the position information of P3 is obtained by reverse deduction based on the position information in the second self-vehicle state information, then the speed of P3 is deduced based on the first derivative of the Bezier curve, and the acceleration of P3 is deduced based on the second derivative of the Bezier curve. Finally, the self-vehicle state information of the fourth control point P3 (at the time T1) is composed of the position information, the speed, and the acceleration of the fourth control point P3, which is only an example.

[0069] In step S105, if the road prediction is invalid, the self-vehicle state information of the fourth control point is calculated based on the first self-vehicle state information at the current time and the first prediction time period.

[0070] The first self-vehicle state information at least includes the curvature information, the speed, the acceleration, and the position information of the self-vehicle, and the curvature information at least includes the curvature and the curvature change rate of the self-vehicle. The self-vehicle state information of the fourth control point is calculated based on the first self-vehicle state information at the current time and the first prediction time period, including: judging whether the curvature of the self-vehicle at the current time is less than a preset curvature threshold; when the curvature of the self-vehicle at the current time is less than the preset curvature threshold, the self-vehicle state information of the fourth control point is calculated based on the position information, the speed, the acceleration of the self-vehicle at the current time, and the first prediction time period; or when the curvature of the self-vehicle at the current time is not less than the preset curvature threshold, the self-vehicle state information of the fourth control point is calculated based on the curvature and the curvature change rate of the self-vehicle, the first prediction time period, and the speed, the acceleration, and the position information of the self-vehicle at the current time.

[0071] If it is determined that the distance traveled by the self-vehicle is greater than the length of the lane line, the road prediction can be determined to be invalid, and the self-vehicle state information of the fourth control point is calculated based on the self-vehicle state information. The position information of the fourth control point can be calculated based on the self-vehicle state information, so as to realize the bottom support of the self-vehicle state information of the fourth control point calculated under the invalid lane line, and ensure the reliability of the self-vehicle state information calculation.

[0072] Specifically, when calculating based on the self-vehicle state information, the self-vehicle state information of the fourth control point is calculated in two ways of straight-line driving and curve driving of the self-vehicle, so as to accurately identify two typical forms of scenes based on the current self-vehicle curvature, match the most suitable input parameters for the fourth control point calculation, and ensure the accuracy of the self-vehicle state information calculation of the fourth control point.

[0073] Specifically, when the curvature of the ego vehicle at the current time is less than the preset curvature threshold, it indicates that the curvature of the ego vehicle at the time T0 satisfies straight-line driving, and the position information of the fourth control point can be directly calculated by using the uniform variable speed straight-line kinematics formula, and if the curvature of the ego vehicle at the current time is greater than the preset curvature threshold, it indicates that the curvature of the ego vehicle at the time T0 satisfies curve driving, and then the position information of the fourth control point of the ego vehicle can be calculated based on the curvature information, and the speed and acceleration of the fourth control point are calculated based on the position information of the fourth control point.

[0074] In step S106, the trajectory of the first prediction time period is predicted based on the ego vehicle state information corresponding to the first control point, the second control point, the third control point and the fourth control point, respectively, to obtain the first prediction trajectory segment.

[0075] The embodiment of the present application can convert the discrete ego vehicle state information of the first control point, the second control point, the third control point and the fourth control point into a continuous trajectory through a mathematical model of a third-order Bezier curve, and then discretely sample the parameter t of the continuous trajectory corresponding to the third-order Bezier curve to generate a trajectory point position, and finally supplement the motion state information of the trajectory point to form the first prediction trajectory segment. The detailed trajectory prediction method will not be described here.

[0076] The trajectory prediction method provided by the embodiment can obtain the first ego vehicle state information at the current time, the lane line information at the current time and the ego vehicle state information corresponding to the first control point, the second control point and the third control point in the first Bezier curve in the first prediction time period, calculate the second ego vehicle state information at a second preset time period from the current time based on the lane line information, and then calculate the ego vehicle state information of the fourth control point in the first Bezier curve based on the ego vehicle state information corresponding to the first control point, the second control point and the third control point and the second ego vehicle state information. Finally, the trajectory of the first prediction time period is predicted based on the ego vehicle state information corresponding to the first control point, the second control point, the third control point and the fourth control point, respectively, to obtain the first prediction trajectory segment. The ego vehicle information and the lane line information are combined to reduce the dependence on the map, improve the accuracy of the ego vehicle trajectory prediction, improve the driving function safety, and if it is determined that the road prediction is invalid, the ego vehicle state information can be used for trajectory prediction to ensure the reliability of the autonomous driving of the trajectory prediction.

[0077] In the embodiment, a trajectory prediction method is provided, which can be used for a controller in a vehicle, Figure 3 The flowchart of the trajectory prediction method according to the embodiment of the present application is shown in Figure 3 The flowchart includes the following steps:

[0078] Step S301: Obtain the current vehicle status information, the current lane line information, and the vehicle status information corresponding to the first control point, the second control point, and the third control point in the first Bézier curve corresponding to the vehicle in the first prediction time period.

[0079] The first control point is the starting point of the first Bézier curve, the second control point is the adjustment point on the starting side, and the third control point is the adjustment point on the ending side.

[0080] Step S302: Based on the first vehicle status information and the lane line information at the current moment, determine whether the road prediction is effective.

[0081] Specifically, based on the first vehicle status information and the lane line information at the current moment, it is determined whether the road prediction is effective, including: determining whether the speed of the vehicle at the current moment is greater than a second preset speed threshold; if the speed of the vehicle at the current moment is greater than the second preset speed threshold, multiplying the speed of the vehicle at the current moment by a preset second duration to obtain a first distance; determining whether the first distance is greater than the lane line length; if the first distance is not greater than the lane line length, determining whether the target acceleration of the vehicle meets the preset acceleration conditions; if the target acceleration of the vehicle meets the preset acceleration conditions, the road prediction is determined to be effective.

[0082] like Figure 4 As shown, before calculating the vehicle state information of the fourth control point based on lane line information (road polynomial model), this embodiment of the invention needs to determine whether the current vehicle and road information meet the conditions for calculating the vehicle state information of the fourth control point based on lane line information. Only when the current vehicle and road information meet the conditions is it determined that the vehicle trajectory calculated based on lane line information is valid (road valid). Only then is the vehicle state information of the fourth control point calculated based on lane line information, avoiding invalid calculations or incorrect decisions, and ensuring the reliability and accuracy of the pre-aiming results and the calculation of vehicle state information.

[0083] Specifically, the embodiment of the present application can determine whether the speed of the ego vehicle at the current time is greater than a second preset speed threshold, and / or can determine whether the first lateral speed of the ego vehicle at the predicted 3s from the current time is greater than the second preset speed threshold (the speed threshold is taken as 35kph as an example), if the speed of the ego vehicle or the predicted ego vehicle speed is greater than 35kph, then the speed of the ego vehicle at the current time can be multiplied by a preset second time length to obtain a first distance, wherein the second time length can be a second preset time period of 3s from the current time, and then compared with the lane line length determined according to the road data output by the front-view camera (the lane line length at this time is the total distance of all lane line segments), if the first distance is less than or equal to the lane line length, it is considered that the road data is valid, and then it can be determined whether the target acceleration of the ego vehicle satisfies the preset acceleration condition, wherein the target acceleration can be the acceleration at the current time or the predicted acceleration of the ego vehicle after 3s, just as an example, if the target acceleration of the ego vehicle also satisfies the preset acceleration condition, it is determined that the ego vehicle state information calculated based on the lane line information of the fourth control point is valid and more accurate, and then the second ego vehicle state information at the second preset time period from the current time can be calculated based on the lane line information.

[0084] In step S303, if the road prediction is valid, the second ego vehicle state information at the second preset time period from the current time is calculated based on the lane line information, the second preset time period is greater than the first prediction time period, and the second preset time period is an integer multiple of the first prediction time period.

[0085] The second self-vehicle state information includes: a first lateral position of the self-vehicle at a second preset time period from the current moment, a first lateral speed and a first lateral acceleration, and the lane line information at least includes a lane line orientation angle, a lane line length, a lane line offset and lane line curvature information; the second self-vehicle state information at the second preset time period from the current moment is calculated based on the lane line information, including: the lane line orientation angle, the lane line offset and the lane line curvature information are taken as lane line coefficients in a road polynomial model, the lane line length is taken as a variable value, and input into the road polynomial model to obtain the first lateral position of the self-vehicle at the second preset time period from the current moment; the lane line orientation angle, the lane line curvature information and the lane line length are used to determine a vehicle head orientation angle; the vehicle head orientation angle is multiplied by the longitudinal speed of the self-vehicle to obtain the first lateral speed of the self-vehicle at the second preset time period from the current moment; the lane line curvature information and the lane line length are used to determine a first curvature; the first curvature is multiplied by the longitudinal speed of the self-vehicle to obtain the first lateral acceleration of the self-vehicle at the second preset time period from the current moment; the lane line length is determined based on road data output by a front-view camera of the vehicle, because the lane line captured is a two-dimensional image, then a deep learning model is used to identify the lane line, the design and training mode of the model determines that the output lane line is segmented, that is, the perception, detection and output mode of the lane line by the autonomous driving system is in units of line segments, so the lane line length obtained can be represented as len1, len2 and len3, only as an example, the lane line length can be segmented according to actual driving scenes and camera parameters.

[0086] In the embodiment of the application, the second preset time period is 3s, the second self-vehicle state information at 3s from the current moment is calculated, when the current moment corresponding to the first prediction time period is 0s, the second self-vehicle state information at 3s is calculated, when the current moment corresponding to the second prediction time period is 1s, the second self-vehicle state information at 4s is calculated, when the current moment corresponding to the fourth prediction time period is 3s, the second self-vehicle state information at 6s is calculated, only as an example.

[0087] The embodiment of the present application can calculate the distance dist1 walked by the ego vehicle 3s, and then compare dist1 with the length of at least one of len1, len2 and len3 to determine the position of the ego vehicle on the lane line segment, and then calculate the lane line length based on the position of the ego vehicle on the lane line segment. For example, when dist1>len1+len2+len3, it indicates that the ego vehicle 3s exceeds the obtained lane line segment, and then len=dist1-(len1+len2+len3) in the following formula can be calculated. If dist1>len1+len2, it indicates that the ego vehicle 3s is located in the third segment of the lane line, and then len=dist1-(len1+len2) in the following formula can be calculated, which represents the distance walked by the ego vehicle in the third segment. If dist1>len1, it indicates that the ego vehicle 3s is located in the second segment of the lane line, and then len=dist1-len1 in the following formula can be calculated, which represents the distance walked by the ego vehicle in the second segment. If dist1<len1, it indicates that the ego vehicle 3s is located in the first segment of the lane line, and then len=dist1 in the following formula can be calculated, which is only an example. Then, the corresponding lane line coefficient and lane line length are selected, the road cubic polynomial model is evaluated, the first derivative is taken to obtain the heading angle, the second derivative is taken to obtain the lane line curvature, and then the vehicle lateral speed and lateral acceleration are converted. After the conversion is completed, the second ego vehicle state information at the preview 3s is calculated according to the following formula.

[0088] Among them, the lane line curvature information obtained by the embodiment of the present application at least includes lane line curvature and lane line curvature change rate, the designed road polynomial model is a road cubic polynomial model, the lane line heading angle, the lane line offset, the lane line curvature and the lane line curvature change rate are taken as the lane line coefficients of the road cubic polynomial model, and the lane line length len is taken as the variable value. The first lateral position is calculated by the following formula:

[0089]

[0090] Among them, represents the first lateral position; represents the lane line offset; represents the lane line heading angle; represents the distance walked by the ego vehicle on the corresponding lane line segment; represents the lane line curvature; represents the lane line curvature change rate.

[0091] The embodiment of the present application can also calculate the first longitudinal position of the ego vehicle at the preview 3s according to the road polynomial model of the above formula, which will not be described again. Based on the lane line heading angle, the lane line curvature, the lane line curvature change rate and the lane line length, the heading angle of the vehicle head is determined by the following formula:

[0092]

[0093] wherein, represents the first lateral velocity; represents the first longitudinal velocity at the second preset time period.

[0094] The application can determine the first curvature based on the lane line curvature information and the lane line length, multiply the first curvature with the longitudinal velocity of the ego vehicle, and obtain the first lateral acceleration of the ego vehicle at the second preset time period from the current time.

[0095]

[0096] wherein, represents the first lateral acceleration.

[0097] The application converts the lane line features into the coefficients of the road polynomial model, calculates the second ego vehicle state information of the ego vehicle at the second preset time period from the current time through the road constraint, and can identify the lane deviation risk and other problems in advance, and reserve sufficient adjustment time.

[0098] Specifically, the second preset time period is the third prediction time period from the current time. The following steps determine whether the vehicle's target acceleration meets the preset acceleration conditions: The first vehicle state information at the current time is used as the vehicle state information of the first control point in the first Bézier curve; the second vehicle state information is used as the vehicle state information of the fourth control point in the third Bézier curve corresponding to the third prediction time period; based on the vehicle state information of the fourth control point in the third Bézier curve, the vehicle's acceleration at the first control point in the first Bézier curve, and the position information of the fourth control point, the vehicle state information of the fourth control point in the second Bézier curve corresponding to the second prediction time period is calculated; based on the vehicle state information of the fourth control point in the third Bézier curve and the vehicle's acceleration at the first control point in the first Bézier curve, the vehicle state information of the fourth control point in the first prediction time period corresponding to the first Bézier curve is calculated; wherein, the vehicle state information of the fourth control point in the first Bézier curve is the vehicle state information of the first control point in the second Bézier curve; the vehicle state information of the fourth control point in the second Bézier curve is... The status information is the vehicle status information of the first control point in the third Bézier curve; based on the vehicle status information of the first control point and the fourth control point in each Bézier curve, the vehicle status information of the second and third control points of each Bézier curve is determined; based on the vehicle status information calculated from the second, third, and fourth control points of each Bézier curve, the set of acceleration values ​​corresponding to each Bézier curve is determined; the acceleration sets corresponding to all Bézier curves are combined into a first acceleration set, and the maximum acceleration in the first acceleration set corresponding to all Bézier curves is selected; it is determined whether the maximum acceleration is less than a first preset acceleration threshold; the maximum acceleration in the acceleration set corresponding to the first Bézier curve is subtracted from the lateral acceleration of the vehicle at the current moment to obtain a first difference value, and it is determined whether the first difference value is less than a first preset difference threshold value; if the maximum acceleration is less than the first preset acceleration threshold value and the first difference value is less than the first preset difference threshold value, then it is determined that the target acceleration of the vehicle meets the preset acceleration condition.

[0099] In this embodiment of the invention, the lateral position of the vehicle at the current moment can be set to zero, and the speed of the vehicle can be set to zero. Zero and lateral acceleration Set as the product of the vehicle's speed and its yaw rate; for example... Figure 5 As shown, the second vehicle state information, including the first lateral position, is calculated based on the above embodiment at a position 3 seconds away from the current time. First lateral velocity and first lateral acceleration This serves as the vehicle state information for the fourth control point in the third Bézier curve corresponding to the third prediction time period.

[0100] The embodiment of the present application can calculate the lateral position of the fourth control point in the second Bezier curve corresponding to the second prediction time period based on the lateral position of the ego vehicle of the fourth control point in the third Bezier curve, the lateral acceleration, the acceleration of the first control point and the lateral position of the fourth control point in the first Bezier curve according to the following formula:

[0101]

[0102] wherein, represents the lateral position of the fourth control point in the second Bezier curve; represents the lateral acceleration of the ego vehicle of the fourth control point in the third Bezier curve; represents the lateral position of the fourth control point in the first Bezier curve; represents the acceleration of the first control point in the first Bezier curve.

[0103] The embodiment of the present application can calculate the lateral position of the fourth control point in the first Bezier curve corresponding to the first prediction time period based on the lateral position of the ego vehicle of the fourth control point in the third Bezier curve, the acceleration and the speed of the ego vehicle, and the acceleration of the first control point in the first Bezier curve according to the following formula:

[0104]

[0105] wherein, represents the lateral position of the fourth control point in the first Bezier curve; represents the speed of the ego vehicle of the fourth control point in the third Bezier curve.

[0106] The embodiment of the present application can set the ego vehicle state information of the fourth control point in the first Bezier curve as the ego vehicle state information of the first control point in the second Bezier curve, and set the ego vehicle state information of the fourth control point in the second Bezier curve as the ego vehicle state information of the first control point in the third Bezier curve, and then calculate the ego vehicle state information of the first control point, the second control point, the third control point and the fourth control point in the third Bezier curve according to the following formula:

[0107]

[0108] wherein, represents the acceleration of the first control point in the third Bezier curve; represents the acceleration of the fourth control point in the second Bezier curve; represents the speed of the first control point in the third Bezier curve; represents the speed of the fourth control point in the second Bezier curve; represents the lateral position of the first control point in the third Bezier curve; represents a lateral position of the fourth control point in the second Bezier curve; represents a lateral position of the second control point in the third Bezier curve; represents a lateral position of the third control point in the second Bezier curve; represents a lateral position of the third control point in the third Bezier curve; represents a lateral position of the second control point in the second Bezier curve; represents a lateral position of the fourth control point in the third Bezier curve; represents a first lateral position at the pre-look 3s.

[0109] The self-vehicle state information of the first control point, the second control point, the third control point and the fourth control point in the second Bezier curve can be calculated by the following formula combination according to the embodiment of the application:

[0110]

[0111] wherein, represents an acceleration of the first control point in the second Bezier curve; represents an acceleration of the fourth control point in the first Bezier curve; represents a speed of the first control point in the second Bezier curve; represents a speed of the fourth control point in the first Bezier curve; represents a lateral position of the first control point in the second Bezier curve; represents a lateral position of the fourth control point in the first Bezier curve; represents a lateral position of the third control point in the first Bezier curve; represents a lateral position of the second control point in the first Bezier curve; represents a lateral position of the third control point in the first Bezier curve; represents a lateral position of the fourth control point in the first Bezier curve.

[0112] The self-vehicle state information of the first control point, the second control point, the third control point and the fourth control point in the first Bezier curve can be calculated by the following formula combination according to the embodiment of the application:

[0113]

[0114] wherein, represents an acceleration of the first control point in the first Bezier curve; represents a speed of the first control point in the first Bezier curve; represents a lateral position of the first control point in the first Bezier curve.

[0115] The embodiment of the application can further calculate the maximum acceleration of each Bezier curve by the following formula:

[0116]

[0117] wherein, denotes the angle corresponding to the i-th Bezier curve; denotes the current speed of the ego vehicle; denotes the maximum acceleration corresponding to the i-th Bezier curve; denotes the acceleration of the fourth control point in the i-th Bezier curve; denotes the speed of the fourth control point in the i-th Bezier curve; all the accelerations corresponding to the Bezier curves are collected to form a first acceleration set, and the maximum acceleration in the first acceleration set is selected to determine whether the maximum acceleration is less than a first preset acceleration threshold (for example, 2 m / s2); the maximum acceleration corresponding to the first Bezier curve is subtracted from the lateral acceleration of the ego vehicle at the current time to obtain a first difference value, and it is determined whether the first difference value is less than a first preset difference threshold (for example, 0.5 m / s2). If the maximum acceleration is less than the first preset acceleration threshold, the minimum acceleration corresponding to the maximum acceleration is also less than the first preset acceleration threshold, and the first difference value is less than the first preset difference threshold, it is determined that the target acceleration of the ego vehicle satisfies the acceleration preset condition.

[0118] The application determines whether the target acceleration satisfies the acceleration preset condition by the Bezier curve segmentation construction and multi-dimensional acceleration verification, which not only ensures that the acceleration of the vehicle in the preview time period is always within the safety and comfort threshold, but also avoids the risk of acceleration exceeding the limit at a single time point.

[0119] In an optional embodiment, after it is determined that the ego vehicle state information calculated based on the lane line information of the fourth control point is valid, it is further determined whether the speed of the ego vehicle at the current time is greater than a first preset speed threshold; the speed of the ego vehicle at the current time is divided by the acceleration of the ego vehicle to obtain a first time length; it is determined whether the first time length is within a first preset time length range; if the speed of the ego vehicle at the current time is greater than the first preset speed threshold, and the first time length is not within the first preset time length range, it is determined that the ego vehicle does not satisfy the stop condition in the first prediction time period, and the step of calculating the second ego vehicle state information at the second preset time period from the current time based on the lane line information is performed.

[0120] For example, Figure 6 ​As shown, the embodiment of the present application can determine whether the speed of the ego vehicle is greater than a first preset speed threshold (for example, 0.2 m / s), and can also divide the speed of the ego vehicle at the current T0 moment by the acceleration of the ego vehicle to obtain a first time length, and determine whether the first time length is within a first preset time length range (the first preset time length is a first prediction time period). If the speed of the ego vehicle is not greater than the first preset speed threshold, or the first time length is within the first preset time length range, it indicates that the ego vehicle can be stopped in the current prediction time period. The position information (including the longitudinal position and the lateral position) of the fourth control point can be determined as the position information of the third control point, and the speed and acceleration of the fourth control point can be calculated based on the position information of the second control point, the third control point and the fourth control point. For details of how to calculate the speed and acceleration based on the position information, please refer to the above embodiment. Wherein, the speed and acceleration are calculated at t=1 after first-order derivation and second-order derivation based on the position information. When it is determined that the ego vehicle cannot be stopped in the current prediction time period, the lane line information can be used to calculate the second ego vehicle state information, and then the ego vehicle state information of the fourth control point is calculated based on the ego vehicle state information.

[0121] Further, when it is detected that the ego vehicle satisfies the stop condition in the first prediction time period, the distance between the ego vehicle and the first control point at the stopping moment can be calculated based on the first time length, the speed and acceleration of the ego vehicle at the first control point by the following formula:

[0122]

[0123] Wherein, posn represents the distance between the ego vehicle and the first control point at the stopping moment, represents the speed of the ego vehicle in the first ego vehicle state information (i.e., the first control point); t=1; g represents the acceleration in the first ego vehicle state information, which can also be represented by .

[0124] The embodiment of the present application can further add the position information (including the lateral position and the longitudinal position) of the ego vehicle at the first control point and the distance between the ego vehicle and the first control point at the stopping moment to determine the position information of the second control point and the third control point.

[0125] The present application directly calculates the ego vehicle state information based on the existing stopping method when it is determined that the ego vehicle is in a situation that satisfies the stopping condition, avoiding unnecessary resource consumption such as lane line information analysis, and improving the efficiency of trajectory prediction.

[0126] When the speed of the ego vehicle is greater than the first preset speed threshold and the first time length is not within the first preset time length range in the embodiment of the present application, it indicates that the ego vehicle does not satisfy the stopping condition in the current prediction time period, and then the step of calculating the second ego vehicle state information at the second preset time period from the current moment based on the lane line information can be performed.

[0127] In the embodiment of the present application, if the ego vehicle does not satisfy the stop condition in the current prediction period, the position information of the ego vehicle at the second control point can be calculated based on the position information and the speed of the ego vehicle at the first control point; and the position information of the ego vehicle at the third control point can be calculated based on the position information of the ego vehicle at the second control point, the position information of the ego vehicle at the first control point, and the acceleration of the ego vehicle.

[0128] In the embodiment of the present application, the longitudinal position of the ego vehicle at the second control point can be calculated based on the longitudinal position and the longitudinal speed of the ego vehicle at the first control point by the following formula:

[0129]

[0130] wherein, represents the longitudinal position of the ego vehicle at the second control point; represents the longitudinal position of the ego vehicle at the first control point (the longitudinal position of the ego vehicle at the current T0 moment); represents the longitudinal speed of the ego vehicle at the first control point (the longitudinal speed of the ego vehicle at the current T0 moment).

[0131] In the embodiment of the present application, the lateral position of the ego vehicle at the second control point can be calculated based on the lateral position and the lateral speed of the ego vehicle at the first control point by the following formula:

[0132]

[0133] wherein, represents the lateral position of the ego vehicle at the second control point; represents the lateral position of the ego vehicle at the first control point (the lateral position of the ego vehicle at the current T0 moment); represents the lateral speed of the ego vehicle at the first control point (the lateral speed of the ego vehicle at the current T0 moment).

[0134] In the embodiment of the present application, the longitudinal position of the ego vehicle at the third control point can be calculated based on the longitudinal position of the ego vehicle at the second control point, the longitudinal position of the ego vehicle at the first control point, and the acceleration of the ego vehicle by the following formula:

[0135]

[0136] wherein, represents the longitudinal position of the ego vehicle at the third control point; represents the acceleration of the ego vehicle at the first control point (the acceleration of the ego vehicle at the T0 moment).

[0137] In the embodiment of the present application, the lateral position of the ego vehicle at the third control point can be calculated based on the lateral position of the ego vehicle at the second control point, the lateral position of the ego vehicle at the first control point, and the acceleration of the ego vehicle by the following formula:

[0138]

[0139] wherein, represents the lateral position of the ego vehicle at the third control point.

[0140] The application determines the stopping feasibility of the ego vehicle in the first prediction time period through the speed threshold and the time length range, and then calculates the ego vehicle state information of the corresponding control point based on the lane line information for the case that does not meet the stopping condition, predicts the safety path in this case, and guarantees the driving safety in different scenarios.

[0141] In step S304, the ego vehicle state information of the fourth control point in the first Bezier curve is calculated based on the ego vehicle state information corresponding to the first control point, the second control point and the third control point respectively and the second ego vehicle state information.

[0142] Specifically, when the ego vehicle cannot stop in the current prediction time period, the ego vehicle state information of the fourth control point can be calculated based on the second ego vehicle state information calculated by using the lane line information, and the position information includes the lateral position and the longitudinal position, and the above step S304 includes:

[0143] In step S3041, the lateral position of the fourth control point is calculated based on the first lateral acceleration, the first lateral speed, the lateral position of the third control point, the lateral position of the second control point and the first lateral position.

[0144] In step S3042, the longitudinal position of the fourth control point is calculated based on the longitudinal position of the third control point, the initial speed of the first control point, the initial acceleration, the lateral position of the fourth control point and the lateral position of the third control point.

[0145] In step S3043, the speed and acceleration of the fourth control point are calculated based on the position information of the second control point, the third control point and the fourth control point.

[0146] The embodiment of the application can calculate the lateral position of the fourth control point based on the first lateral acceleration, the first lateral speed, the lateral position of the third control point, the lateral position of the second control point and the first lateral position by the following formula:

[0147]

[0148] wherein, represents the lateral position of the ego vehicle at the fourth control point; The lateral position of the ego vehicle at the third control point can also be represented by the above represents; represents the lateral position of the ego vehicle at the second control point, which can also be represented by the above represents.

[0149] The embodiment of the present application can calculate the longitudinal position of the fourth control point based on the longitudinal position of the third control point, the initial speed of the first control point, the initial acceleration, the lateral position of the fourth control point and the lateral position of the third control point according to the following formula:

[0150]

[0151] wherein, represents the longitudinal position of the ego vehicle at the fourth control point; represents the longitudinal position of the third control point, which can also be represented by the above-mentioned ; represents the speed of the ego vehicle at T0; represents the acceleration of the ego vehicle at T0, which can also be represented by the above-mentioned .

[0152] The present application is based on the ego vehicle state information of the previous control point, and uses the ego vehicle state information of the preset time period of the preview as a constraint to gradually derive the information of the fourth control point, thereby improving the continuity and accuracy of the parameter prediction of the fourth control point, and further improving the fitting accuracy of the entire Bezier curve to the preset path.

[0153] In step S305, if the road prediction is invalid, the ego vehicle state information of the fourth control point is calculated based on the first ego vehicle state information at the current time and the first prediction time period.

[0154] The first ego vehicle state information at least includes the curvature information, the acceleration and the position information of the ego vehicle, and the curvature information at least includes the curvature and the curvature change rate of the ego vehicle. The ego vehicle state information of the fourth control point is calculated based on the first ego vehicle state information at the current time and the first prediction time period, including: judging whether the curvature of the ego vehicle at the current time is less than a preset curvature threshold; when the curvature of the ego vehicle at the current time is less than the preset curvature threshold, calculating the ego vehicle state information of the fourth control point based on the first ego vehicle state information at the current time and the first prediction time period; or when the curvature of the ego vehicle at the current time is not less than the preset curvature threshold, calculating the ego vehicle state information of the fourth control point based on the curvature and the curvature change rate of the ego vehicle, the first prediction time period and the speed, the acceleration and the position information of the ego vehicle at the current time.

[0155] When the speed of the ego vehicle at the current time is not greater than a second preset speed threshold, or the first distance is greater than the length of the lane line, or the target acceleration of the ego vehicle does not satisfy the acceleration preset condition, it is indicated that the road prediction is invalid, and whether the curvature of the ego vehicle at the current time is less than a preset curvature threshold can be determined; when the curvature of the ego vehicle at the current time is less than the preset curvature threshold, ego vehicle state information of the fourth control point is calculated based on the first ego vehicle state information at the current time and the first prediction time period; or when the curvature of the ego vehicle at the current time is not less than the preset curvature threshold, the ego vehicle state information of the fourth control point is calculated based on the curvature and the curvature change rate of the ego vehicle, the first prediction time period, and the speed, acceleration and position information of the ego vehicle at the current time.

[0156] As shown in Figure 4 If it is determined that the speed of the ego vehicle is not greater than a second preset speed threshold, or the first distance is greater than the length of the lane line, or the target acceleration of the ego vehicle does not satisfy the acceleration preset condition, it is indicated that the calculation of the position information of the fourth control point based on the lane line information is invalid, and then the position information of the fourth control point can be calculated based on the ego vehicle state information, so as to realize bottom handling of the ego vehicle state information of the fourth control point through the ego vehicle state information calculation under the problem of calculation failure caused by invalid lane lines, and ensure the reliability of the ego vehicle state information calculation.

[0157] Specifically, when the ego vehicle state information is calculated, the ego vehicle state information of the fourth control point is calculated in two ways of straight-line driving and curve driving of the ego vehicle, so as to accurately identify two typical scene forms based on the curvature of the ego vehicle at the current time, match the most suitable input parameters for the fourth control point calculation, and ensure the accuracy of the ego vehicle state information calculation of the fourth control point.

[0158] Specifically, the curvature information at least includes the curvature and the curvature change rate of the ego vehicle, and when the curvature of the ego vehicle at the current time is less than a preset curvature threshold, it is indicated that the curvature of the ego vehicle at T0 satisfies straight-line driving, and the position information of the fourth control point can be directly calculated by using a uniform variable speed straight-line kinematics formula, so as to simplify the parameter input and operation logic of the straight-line scene and improve the trajectory prediction efficiency, and then the position information of the fourth control point can be calculated based on the following formula:

[0159]

[0160] wherein, represents the position information of the ego vehicle at the fourth control point; represents the position information of the ego vehicle at T0; represents the acceleration of the ego vehicle at T0; represents the speed of the ego vehicle at T0; represents the time length of the fourth control point from the first control point.

[0161] Specifically, when the curvature of the ego vehicle at the current time is greater than a preset curvature threshold, it indicates that the curvature of the ego vehicle at the time T0 satisfies curve driving, and then the reciprocal of the curvature of the ego vehicle at the current time can be taken as the ego vehicle driving radius; the curvature change rate of the ego vehicle at the current time is multiplied by the first prediction time period to obtain a second curvature; the second curvature is added to the curvature of the ego vehicle at the current time to obtain a curvature sum; the reciprocal of the curvature sum is taken as the ego vehicle predicted driving radius; the ego vehicle predicted driving distance is calculated based on the speed, acceleration of the ego vehicle at the current time and the first prediction time period; the first angle value is obtained by dividing the ego vehicle predicted driving distance by the ego vehicle driving radius; the second angle value is obtained by dividing the ego vehicle predicted driving distance by the ego vehicle predicted driving radius; the first angle of the ego vehicle driving is calculated based on the first angle value and the ego vehicle driving radius; the second angle of the ego vehicle driving is calculated based on the second angle value and the ego vehicle predicted driving radius; and the position information of the fourth control point is calculated based on the first angle of the ego vehicle driving, the second angle of the ego vehicle driving, the angle preset coefficients corresponding to the first angle and the second angle respectively, and the position information of the ego vehicle at the current time.

[0162] The embodiment of the present application can take the reciprocal of the curvature of the ego vehicle at the current time as the ego vehicle driving radius, i.e. , wherein, represents the ego vehicle driving radius; represents the curvature of the ego vehicle; then the curvature change rate of the ego vehicle at the current time can be multiplied by the first prediction time period to obtain a second curvature, and then the second curvature is added to the curvature of the ego vehicle at the current time to obtain a curvature sum, and finally the reciprocal of the curvature sum is taken as the ego vehicle predicted driving radius, i.e. , wherein, represents the ego vehicle predicted driving radius; represents the curvature change rate of the ego vehicle; t represents the first time length, t=1.

[0163] The embodiment of the present application can calculate the ego vehicle predicted driving distance based on , wherein, represents the ego vehicle predicted driving distance; represents the speed of the ego vehicle; represents the acceleration of the ego vehicle.

[0164] The embodiment of the present application can divide the ego vehicle predicted driving distance by the ego vehicle driving radius to obtain the first angle value, i.e. , wherein, represents the first angle value; and the embodiment of the present application can further divide the ego vehicle predicted driving distance by the ego vehicle predicted driving radius to obtain the second angle value, i.e.

[0165] .

[0166] The embodiment of the present application can further calculate the first angle of the ego vehicle driving based on the first angle value and the ego vehicle driving radius, i.e. wherein, represents the first angle of the ego vehicle driving; the second angle of the ego vehicle driving is calculated based on the second angle value and the predicted driving radius of the ego vehicle, wherein, represents the second angle of the ego vehicle driving; finally, the position information of the fourth control point can be calculated based on the first angle of the ego vehicle driving, the second angle of the ego vehicle driving, the angle preset coefficients corresponding to the first angle and the second angle respectively, and the position information of the ego vehicle at the current time, wherein, represents the position information (including the lateral position and the longitudinal position) of the fourth control point; represents the position information of the first control point, the angle preset coefficient corresponding to the first angle of the ego vehicle driving is 0.7, and the angle preset coefficient corresponding to the second angle of the ego vehicle driving is 0.3. The value of the angle preset coefficient can be set according to the actual driving road scene, and is only taken as an example.

[0167] When it is determined that the vehicle meets the curve driving, the steering dynamic trend can be determined based on the curvature and the curvature change rate of the ego vehicle, and then the smooth transition is realized through the double-angle fusion, so that the curve driving demand is accurately adapted, and the accuracy and smoothness of the curve path prediction are improved.

[0168] In step S306, the trajectory of the first prediction time period is predicted based on the ego vehicle state information corresponding to the first control point, the second control point, the third control point and the fourth control point respectively, to obtain the first prediction trajectory segment. For details, please refer to the above embodiments, which will not be repeated here.

[0169] The embodiment of the application can obtain the lane line information corresponding to the second prediction time period through the following steps: calculating the square of the lane line length at the current time to obtain a first square value, and calculating the cube of the lane line length to obtain a first cube value; multiplying the lane line orientation angle at the current time by the lane line length to obtain a first product, and multiplying the lane line curvature at the current time by the first square value to obtain a second product; multiplying the lane line curvature change rate at the current time by the first cube value to obtain a third product; adding the first product, the second product, the third product and the lane line offset at the current time to obtain the lane line offset corresponding to the second prediction time period; multiplying the first preset coefficient, the lane line curvature at the current time and the lane line length to obtain a fourth product; multiplying the second preset coefficient, the lane line curvature change rate at the current time and the first product to obtain a fifth product; adding the fourth product, the fifth product and the orientation angle at the current time to obtain the lane line orientation angle corresponding to the second prediction time period; multiplying the third preset coefficient by the lane line curvature at the current time to obtain a sixth product; multiplying the lane line length at the current time, the fourth preset coefficient and the lane line curvature change rate at the current time to obtain the lane line curvature corresponding to the second prediction time period; taking the lane line curvature change rate at the current time as the lane line curvature change rate corresponding to the second prediction time period; and taking the lane line length at the current time as the lane line length corresponding to the second prediction time period.

[0170] The embodiment of the application can calculate the lane line offset corresponding to the second prediction time period through the following formula:

[0171]

[0172] wherein, represents the lane line offset corresponding to the second prediction time period; represents the lane line offset corresponding to the first prediction time period; represents the lane line length corresponding to the first prediction time period; represents the lane line orientation angle corresponding to the first prediction time period; represents the lane line curvature corresponding to the first prediction time period; represents the lane line curvature change rate corresponding to the first prediction time period.

[0173] The embodiment of the application can calculate the orientation angle corresponding to the second prediction time period through the following formula:

[0174]

[0175] wherein, represents the orientation angle corresponding to the second prediction time period, the first preset coefficient is 2, the second preset coefficient is 3, and the specific preset coefficient is calculated in real time according to the road polynomial model and is adjusted based on the bending characteristics of the road.

[0176] The embodiment of the present application can multiply the third preset coefficient and the lane line curvature at the current moment to obtain a sixth product; multiply the lane line length at the current moment, the fourth preset coefficient, and the lane line curvature change rate at the current moment to obtain the lane line curvature corresponding to the second prediction time period , that is

[0177]

[0178] The third preset coefficient is 2, the fourth preset coefficient is 6, and the specific preset coefficient is also calculated in real time according to the road polynomial model. The road polynomial model can accurately fit the continuous geometric characteristics of the road, derive the theoretical value of the curvature and the curvature change rate through the polynomial coefficient, and further derive the preset coefficient matched with the current road characteristics.

[0179] The embodiment of the present application can take the lane line curvature change rate at the current moment as the lane line curvature change rate corresponding to the second prediction time period, that is ; or take the lane line length at the current moment as the lane line length at the second prediction time period .

[0180] The present application calculates the lane line information corresponding to the next prediction time period based on the lane line information corresponding to the last prediction time period, to generate continuous lane line information, smoothly transition the lane line characteristics, and provide a homologous and continuous constraint reference for the Bezier curve corresponding to different time periods.

[0181] The embodiment of the present application can obtain the third ego vehicle state information corresponding to the second prediction time period by the following steps: based on the speed and acceleration of the ego vehicle at the first prediction time period and the current moment, the speed of the ego vehicle corresponding to the second prediction time period is obtained; the acceleration of the ego vehicle at the current moment is taken as the acceleration of the ego vehicle corresponding to the second prediction time period; the position information of the fourth control point of the ego vehicle in the first Bezier curve is taken as the position information of the ego vehicle corresponding to the second prediction time period; the curvature change rate at the current moment is multiplied by the first prediction time period to obtain a third curvature; the third curvature is added to the curvature at the current moment to obtain the curvature of the ego vehicle corresponding to the second prediction time period.

[0182] The embodiment of the application can calculate the speed of the ego vehicle corresponding to the second prediction time period by spd_t1 = m_motion_t0.speed + m_motion_t0.A, wherein spd_t1 represents the speed of the ego vehicle corresponding to the second prediction time period, m_motion_t0.speed represents the speed of the ego vehicle at T0, m_motion_t0.A represents the acceleration of the ego vehicle at T0, t = 1, and a limiting function can be added to limit the maximum value; the embodiment of the application can initialize the acceleration of the ego vehicle corresponding to the second prediction time period as the acceleration of the ego vehicle at T0, and if there is no stop information indicating that the ego vehicle is decelerating, the current acceleration can be kept, otherwise, the acceleration of the ego vehicle corresponding to the second prediction time period is set to zero, and the speed and curvature of the ego vehicle corresponding to the second prediction time period can be used to calculate the longitudinal and lateral accelerations of the ego vehicle; the position information of the fourth control point in the first Bezier curve can be used as the position information of the ego vehicle corresponding to the second prediction time period; the average speed of the ego vehicle corresponding to the second prediction time period in the longitudinal and lateral directions can be calculated; the heading angle of the ego vehicle corresponding to the second prediction time period can be calculated by an arctangent function, i.e., hd_ag_t1 = atan2f(vlat, vlgt), wherein hd_ag_t1 represents the heading angle of the ego vehicle corresponding to the second prediction time period, vlat and vlgt represent the lateral speed and longitudinal speed at T0 respectively, and if the change of the heading angle is within a threshold, the heading angle at the previous time is used, otherwise, the current calculated heading angle is kept; the curvature change rate at the current time can be multiplied by the first prediction time period to obtain a third curvature, and then the third curvature is added to the curvature at the current time to obtain the curvature of the ego vehicle corresponding to the second prediction time period, and a saturation function is used to limit the curvature range; finally, the motion state of the ego vehicle at T = 1s corresponding to the second prediction time period can be updated, and the speed, position, speed and acceleration components of the ego vehicle at T1 are calculated to update all the motion states of the ego vehicle at T = 1s, which are used for the second prediction trajectory segment corresponding to the second prediction time period.

[0183] The embodiment of the application updates the ego vehicle state information corresponding to the second prediction time period based on the first ego vehicle state information and the information of the fourth control point in the first Bezier curve, realizes the continuity with the previous trajectory and the physical consistency among the parameters, and provides reliable parameter input for subsequent trajectory segment calculation.

[0184] Finally, as Figure 7As shown, when it is determined that the lane line information is calculated and the fourth control point of the self-vehicle state information is valid, the predicted lateral and longitudinal positions of the P0, P1, P2 and P3 control points in the first Bezier curve corresponding to the first prediction time period are calculated based on the road model (which can be a road cubic polynomial model), it is judged whether the current Bezier curve is the last Bezier curve, if not, the first self-vehicle state information at the current time and the lane line information at the current time are updated as the third self-vehicle state information and the lane line information corresponding to the second prediction time period, the steps of obtaining the self-vehicle state information corresponding to the first control point, the second control point and the third control point in the first Bezier curve corresponding to the first prediction time period are executed until the second prediction trajectory segment corresponding to the second prediction time period is obtained, and then the above embodiment is repeatedly executed until the four prediction trajectory segments as shown are obtained, so that the vehicle is driven based on the predicted trajectory segments, and the safety and reliability of vehicle driving are improved. Figure 8 As shown, the four prediction trajectory segments are obtained, so that the vehicle is driven based on the predicted trajectory segments, and the safety and reliability of vehicle driving are improved.

[0185] The present application generates a set of continuous and reliable long-time sequence prediction trajectory system by iterative generation of multi-segment Bezier curve, combined with road model constraint and real-time state update, and improves the stability of vehicle driving.

[0186] In the present embodiment, a vehicle is also provided, as shown, Figure 9 The vehicle includes a controller 91, the controller includes a memory and a processor, the memory and the processor are connected to each other in communication, the memory stores computer instructions, and the processor executes the computer instructions to perform the trajectory prediction method described above.

[0187] In the present embodiment, a trajectory prediction device is also provided, which is used to implement the above embodiments and preferred embodiments, and has been described above. As used below, the term "module" can be a combination of software and / or hardware that implements a predetermined function. Although the device described in the following embodiments is preferably implemented in software, hardware or a combination of software and hardware is also possible and is contemplated.

[0188] The present embodiment provides a trajectory prediction device, as shown, Figure 10As shown, the method comprises: an information acquisition module 1001, configured to acquire first ego vehicle state information at a current time, lane line information at the current time, and ego vehicle state information corresponding to a first control point, a second control point, and a third control point in a first Bezier curve of the ego vehicle at a first prediction time period; the first control point is the starting point of the first Bezier curve, the second control point is a regulation node on the starting point side, and the third control point is a regulation node on the ending point side; a road validity judgment module 1002, configured to judge whether the road prediction is valid based on the first ego vehicle state information and the lane line information at the current time; a second ego vehicle state information acquisition module 1003, configured to, if the road prediction is valid, calculate second ego vehicle state information at a second preset time period from the current time based on the lane line information, the second preset time period being greater than the first prediction time period and being an integer multiple of the first prediction time period; a fourth control point information calculation module 1004, configured to calculate ego vehicle state information of a fourth control point in the first Bezier curve based on the ego vehicle state information corresponding to the first control point, the second control point, and the third control point and the second ego vehicle state information; the fourth control point is the ending point of the first Bezier curve; an ego vehicle state prediction module 1005, configured to, if the road prediction is invalid, calculate the ego vehicle state information of the fourth control point based on the first ego vehicle state information at the current time and the first prediction time period; the first ego vehicle state information at least includes curvature information, speed, acceleration, and position information of the ego vehicle, the curvature information at least includes curvature and curvature change rate of the ego vehicle, the ego vehicle state prediction module 1005 comprises: a curvature judgment unit, configured to judge whether the curvature of the ego vehicle at the current time is less than a preset curvature threshold; a state information calculation unit, configured to, when the curvature of the ego vehicle at the current time is less than the preset curvature threshold, calculate the ego vehicle state information of the fourth control point based on the position information, the speed, the acceleration of the ego vehicle at the current time, and the first prediction time period; or, when the curvature of the ego vehicle at the current time is not less than the preset curvature threshold, calculate the ego vehicle state information of the fourth control point based on the curvature and the curvature change rate of the ego vehicle, the first prediction time period, and the speed, the acceleration, and the position information of the ego vehicle at the current time; a trajectory segment prediction module 1006, configured to perform trajectory prediction on the first prediction time period based on the ego vehicle state information corresponding to the first control point, the second control point, the third control point, and the fourth control point, to obtain a first prediction trajectory segment.

[0189] In some optional embodiments, the second ego vehicle state information includes: a first lateral position, a first lateral speed and a first lateral acceleration of the ego vehicle at a second preset time period ahead of the current time, and the lane line information at least includes a lane line orientation angle, a lane line length, a lane line offset and lane line curvature information, and the second ego vehicle state information acquisition module 1003 includes: a first lateral position calculation unit configured to input the lane line orientation angle, the lane line offset and the lane line curvature information as lane line coefficients in a road polynomial model, and the lane line length as a variable value into the road polynomial model, to obtain the first lateral position of the ego vehicle at the second preset time period ahead of the current time; an orientation angle calculation unit configured to determine a vehicle head orientation angle based on the lane line orientation angle, the lane line curvature information and the lane line length; a first lateral speed calculation unit configured to multiply the vehicle head orientation angle by a longitudinal speed of the ego vehicle to obtain the first lateral speed of the ego vehicle at the second preset time period ahead of the current time; a first curvature calculation unit configured to determine a first curvature based on the lane line curvature information and the lane line length; and a first lateral acceleration calculation unit configured to multiply the first curvature by the longitudinal speed of the ego vehicle to obtain the first lateral acceleration of the ego vehicle at the second preset time period ahead of the current time.

[0190] In some optional embodiments, the first ego vehicle state information at the current time at least includes speed, acceleration and position information of the ego vehicle, and before the step of calculating the second ego vehicle state information at the second preset time period ahead of the current time based on the lane line information, the trajectory prediction apparatus further includes: a speed detection unit configured to detect whether the speed of the ego vehicle at the current time is greater than a first preset speed threshold; a first time length calculation unit configured to divide the speed of the ego vehicle at the current time by the acceleration of the ego vehicle to obtain a first time length; a time length judgment unit configured to judge whether the first time length is within a first preset time length range; and a step execution unit configured to, if the speed of the ego vehicle at the current time is greater than the first preset speed threshold and the first time length is not within the first preset time length range, determine that the ego vehicle does not meet the stop condition within the first prediction time period, and execute the step of calculating the second ego vehicle state information at the second preset time period ahead of the current time based on the lane line information.

[0191] In some optional embodiments, the first ego vehicle state information at least includes position information of the ego vehicle, the position information including: a lateral position and a longitudinal position, the fourth control point information calculation module 1004 includes: a lateral position calculation unit, configured to calculate the lateral position of the fourth control point based on the first lateral acceleration, the first lateral speed, the lateral position of the third control point, the lateral position of the second control point and the first lateral position; a longitudinal position calculation unit, configured to calculate the longitudinal position of the fourth control point based on the longitudinal position of the third control point, the initial speed of the first control point, the initial acceleration, the lateral position of the fourth control point and the lateral position of the third control point; and a speed calculation unit, configured to calculate the speed and acceleration of the fourth control point based on the position information of the second control point, the third control point and the fourth control point.

[0192] In some optional embodiments, the trajectory prediction apparatus further includes an information determination module, configured to: if the speed of the ego vehicle at the current time is not greater than a first preset speed threshold, or the first time length is within a first preset time length range, determine that the ego vehicle satisfies the stop condition in the first prediction time period, determine the position information of the fourth control point as the position information of the third control point, and calculate the speed and acceleration of the fourth control point based on the position information of the second control point, the third control point and the fourth control point.

[0193] In some optional embodiments, the information acquisition module 1001 includes: a state information determination unit, configured to determine the first ego vehicle state information at the current time as the ego vehicle state information of the first control point in the first Bezier curve; a second position information calculation unit, configured to: when the speed of the ego vehicle at the current time is greater than a first preset speed threshold, and the first time length is not within a first preset time length range, determine that the ego vehicle does not satisfy the stop condition in the first prediction time period, and calculate the position information of the ego vehicle at the second control point based on the position information and speed of the ego vehicle at the first control point; and a third position information calculation unit, configured to calculate the position information of the ego vehicle at the third control point based on the position information of the ego vehicle at the second control point, the position information of the ego vehicle at the first control point and the acceleration.

[0194] In some optional embodiments, the trajectory prediction apparatus further includes: a distance calculation module, configured to: when the speed of the ego vehicle at the current time is not greater than a first preset speed threshold, or the first time length is within a first preset time length range, determine that the ego vehicle satisfies the stop condition in the first prediction time period, calculate the distance between the ego vehicle and the first control point at the stop time based on the first time length, the speed of the ego vehicle at the first control point and the acceleration; and a position information calculation module, configured to determine the position information of the second control point and the third control point based on the position information of the ego vehicle at the first control point and the distance between the ego vehicle and the first control point at the stop time.

[0195] In some optional embodiments, the road validity determination module 1002 comprises: a speed determination unit configured to determine whether the speed of the ego vehicle at the current time is greater than a second preset speed threshold; a distance calculation unit configured to, if the speed of the ego vehicle at the current time is greater than the second preset speed threshold, multiply the speed of the ego vehicle at the current time by a preset second time length to obtain a first distance; a distance determination unit configured to determine whether the first distance is greater than the length of the lane line; an acceleration determination unit configured to, if the first distance is not greater than the length of the lane line, determine whether the target acceleration of the ego vehicle satisfies an acceleration preset condition; and a road validity determination unit configured to, if the target acceleration of the ego vehicle satisfies the acceleration preset condition, determine that the road prediction is valid.

[0196] In some optional embodiments, the ego vehicle state information of the fourth control point is calculated based on the curvature and the rate of change of curvature of the ego vehicle, the speed, the acceleration, and the position information of the ego vehicle at the current time, and the first prediction time period, and comprises: taking the reciprocal of the curvature of the ego vehicle at the current time as an ego vehicle driving radius; multiplying the rate of change of curvature of the ego vehicle at the current time by the first prediction time period to obtain a second curvature; adding the second curvature and the curvature of the ego vehicle at the current time to obtain a curvature sum; taking the reciprocal of the curvature sum as an ego vehicle predicted driving radius; calculating an ego vehicle predicted driving distance based on the speed, the acceleration of the ego vehicle at the current time, and the first prediction time period; dividing the ego vehicle predicted driving distance by the ego vehicle driving radius to obtain a first angle value; dividing the ego vehicle predicted driving distance by the ego vehicle predicted driving radius to obtain a second angle value; calculating an ego vehicle driving first angle based on the first angle value and the ego vehicle driving radius; calculating an ego vehicle driving second angle based on the second angle value and the ego vehicle predicted driving radius; calculating the position information of the fourth control point based on the ego vehicle driving first angle, the ego vehicle driving second angle, and the angle preset coefficients corresponding to the ego vehicle driving first angle and the ego vehicle driving second angle, and the position information of the ego vehicle at the current time; and calculating the speed and the acceleration of the fourth control point based on the position information of the second control point, the third control point, and the fourth control point.

[0197] In some optional embodiments, the second preset time period is a third prediction time period from the current time, and whether the target acceleration of the ego vehicle satisfies the acceleration preset condition is determined by the following steps: taking the first ego vehicle state information at the current time as the ego vehicle state information of a first control point in a first Bezier curve; taking the second ego vehicle state information as the ego vehicle state information of a fourth control point in a third Bezier curve corresponding to the third prediction time period; calculating the ego vehicle state information of the fourth control point in a second Bezier curve corresponding to the second prediction time period based on the ego vehicle state information of the fourth control point in the third Bezier curve, the acceleration of the ego vehicle at the first control point in the first Bezier curve, and the position information of the fourth control point in the first Bezier curve; calculating the ego vehicle state information of the fourth control point in the first Bezier curve corresponding to the first prediction time period based on the ego vehicle state information of the fourth control point in the third Bezier curve and the acceleration of the ego vehicle at the first control point in the first Bezier curve; wherein the ego vehicle state information of the fourth control point in the first Bezier curve is the ego vehicle state information of the first control point in the second Bezier curve, and the ego vehicle state information of the fourth control point in the second Bezier curve is the ego vehicle state information of the first control point in the third Bezier curve; determining the ego vehicle state information of the second control point and the third control point of each Bezier curve based on the ego vehicle state information of the first control point in each Bezier curve and the ego vehicle state information of the fourth control point of each Bezier curve; determining the acceleration value set corresponding to each Bezier curve based on the ego vehicle state information calculated respectively by the second control point, the third control point, and the fourth control point of each Bezier curve; combining the acceleration sets corresponding to all Bezier curves into a first acceleration set, and selecting the maximum acceleration in the first acceleration set corresponding to all Bezier curves; determining whether the maximum acceleration is less than a first preset acceleration threshold; subtracting the lateral acceleration of the ego vehicle at the current time from the maximum acceleration in the acceleration set corresponding to the first Bezier curve to obtain a first difference value, and determining whether the first difference value is less than a first preset difference threshold; if the maximum acceleration is less than the first preset acceleration threshold, and the first difference value is less than the first preset difference threshold, it is determined that the target acceleration of the ego vehicle satisfies the acceleration preset condition.

[0198] In some optional embodiments, the lane line information at least includes a lane line orientation angle, a lane line length, a lane line offset, a lane line curvature, and a lane line curvature change rate, and the method further comprises: obtaining the lane line information corresponding to the second prediction time period by the following steps: calculating a square of the lane line length at the current time to obtain a first square value, and calculating a cube of the lane line length to obtain a first cube value; multiplying the lane line orientation angle at the current time by the lane line length to obtain a first product, and multiplying the lane line curvature at the current time by the first square value to obtain a second product; multiplying the lane line curvature change rate at the current time by the first cube value to obtain a third product; adding the first product, the second product, the third product, and the lane line offset at the current time to obtain the lane line offset corresponding to the second prediction time period; multiplying a first preset coefficient, the lane line curvature at the current time, and the lane line length to obtain a fourth product; multiplying a second preset coefficient, the lane line curvature change rate at the current time, and the first product to obtain a fifth product; adding the fourth product, the fifth product, and the orientation angle at the current time to obtain the lane line orientation angle corresponding to the second prediction time period; multiplying a third preset coefficient by the lane line curvature at the current time to obtain a sixth product; multiplying the lane line length at the current time, a fourth preset coefficient, and the lane line curvature change rate at the current time to obtain the lane line curvature corresponding to the second prediction time period; taking the lane line curvature change rate at the current time as the lane line curvature change rate corresponding to the second prediction time period; and taking the lane line length at the current time as the lane line length corresponding to the second prediction time period.

[0199] In some optional embodiments, the ego vehicle state information at least includes a speed, an acceleration, position information, a curvature, and a curvature change rate of the ego vehicle, and the method further comprises: obtaining the third ego vehicle state information corresponding to the second prediction time period by the following steps: based on the speed and the acceleration of the ego vehicle at the first prediction time period and the current time, obtaining the speed of the ego vehicle corresponding to the second prediction time period; taking the acceleration of the ego vehicle at the current time as the acceleration of the ego vehicle corresponding to the second prediction time period; taking the position information of the fourth control point of the ego vehicle in the first Bezier curve as the position information of the ego vehicle corresponding to the second prediction time period; multiplying the curvature change rate at the current time by the first prediction time period to obtain a third curvature; adding the third curvature and the curvature at the current time to obtain the curvature of the ego vehicle corresponding to the second prediction time period.

[0200] In some optional embodiments, the device further comprises: a trajectory prediction module configured to update the first ego vehicle state information at the current time and the lane line information at the current time to the third ego vehicle state information and the lane line information corresponding to the second prediction time period, and perform the step of obtaining the ego vehicle state information corresponding to the first control point, the second control point, and the third control point of the ego vehicle in the first Bezier curve corresponding to the first prediction time period until the second prediction trajectory segment corresponding to the second prediction time period is obtained.

[0201] Further functional descriptions of the above modules and units are the same as those in the corresponding embodiments described above, and will not be repeated here.

[0202] In this embodiment, the trajectory prediction device is presented in the form of a functional unit. Here, a unit refers to an ASIC (Application Specific Integrated Circuit) circuit, a processor and memory that execute one or more software or fixed programs, and / or other devices that can provide the above functions.

[0203] This invention also provides a controller having the above-described features. Figure 10 The trajectory prediction device shown.

[0204] Please see Figure 11 , Figure 11 This is a schematic diagram of the structure of a controller in a vehicle provided by an optional embodiment of the present invention, such as... Figure 11 As shown, the controller includes one or more processors 10, memory 20, and interfaces for connecting the components, including high-speed interfaces and low-speed interfaces. The components communicate with each other via different buses and can be mounted on a common motherboard or otherwise as required. The processors can process instructions executed within the controller, including instructions stored in or on memory to display graphical information of a GUI on external input / output devices (such as display devices coupled to the interfaces). In some alternative implementations, multiple processors and / or multiple buses can be used with multiple memories and multiple memory modules, if desired. Similarly, multiple controllers can be connected, each providing some of the necessary operations (e.g., as a server array, a group of blade servers, or a multiprocessor system). Figure 11 Take a processor 10 as an example.

[0205] Processor 10 may be a central processing unit, a network processor, or a combination thereof. Processor 10 may further include a hardware chip. The hardware chip may be an application-specific integrated circuit (ASIC), a programmable logic device (PLD), or a combination thereof. The programmable logic device may be a complex programmable logic device (CAMP), a field-programmable gate array (FPGA), a general-purpose array logic (GDA), or any combination thereof.

[0206] The memory 20 stores instructions executable by at least one processor 10 to cause at least one processor 10 to perform the method shown in the above embodiments.

[0207] The memory 20 can include a program storage area and a data storage area. The program storage area can store an operating system and application programs required for at least one function. The data storage area can store data created according to the use of the controller, etc. In addition, the memory 20 can include a high-speed random access memory, and can also include a non-transitory memory such as at least one disk storage device, a flash memory device, or other non-transitory solid state memory device. In some alternative embodiments, the memory 20 can optionally include memory that is remotely located with respect to the processor 10, and these remote memories can be connected to the controller through a network. Examples of such networks include, but are not limited to, the Internet, an intranet, a local area network, a mobile communications network, and combinations thereof.

[0208] The memory 20 can include a volatile memory such as a random access memory, and can also include a non-volatile memory such as a flash memory, a hard disk, or a solid state disk. The memory 20 can also include a combination of the above-mentioned types of memory.

[0209] The controller also includes a communication interface 30 for communication of the controller with other devices or communication networks.

[0210] The embodiments of the present application also provide a computer readable storage medium. The above-mentioned method according to the embodiments of the present application can be implemented in hardware, firmware, or as computer code recorded on a storage medium, or be implemented by downloading a computer program from a network and stored in a remote storage medium or a non-transitory machine readable storage medium and then stored in a local storage medium, so that the method described herein can be stored on such a software processing of a storage medium using a general purpose computer, a special purpose processor, or programmable or special purpose hardware. The storage medium can be a disk, a compact disk, a read-only memory, a random access memory, a flash memory, a hard disk, or a solid state disk, etc. Further, the storage medium can also include a combination of the above-mentioned types of memory. It can be understood that the computer, processor, microprocessor controller, or programmable hardware includes a storage component that can store or receive software or computer code, when the software or computer code is accessed and executed by the computer, processor, or hardware, the method shown in the above embodiments is implemented.

[0211] Although the embodiments of the present application have been described in conjunction with the accompanying drawings, various modifications and changes can be made by those skilled in the art without departing from the spirit and scope of the present application, and such modifications and changes fall within the scope defined by the appended claims.

Claims

1. A trajectory prediction method characterized by, The method comprises: acquiring first ego vehicle state information at a current time, lane line information at the current time, and ego vehicle state information corresponding to first, second, and third control points in a first Bezier curve of the ego vehicle at a first prediction time period, wherein the first control point is the starting point of the first Bezier curve, the second control point is a tuning point on the starting point side, and the third control point is a tuning point on the ending point side; judging whether the road prediction is valid based on the first ego vehicle state information and the lane line information at the current time; if the road prediction is valid, calculating second ego vehicle state information at a second prediction time period ahead of the current time based on the lane line information, wherein the second prediction time period is longer than the first prediction time period and is an integer multiple of the first prediction time period; calculating ego vehicle state information of a fourth control point in the first Bezier curve based on the ego vehicle state information corresponding to the first, second, and third control points and the second ego vehicle state information, wherein the fourth control point is the ending point of the first Bezier curve; if the road prediction is invalid, calculating the ego vehicle state information of the fourth control point based on the first ego vehicle state information at the current time and the first prediction time period; the first ego vehicle state information at least includes curvature information, speed, acceleration, and position information of the ego vehicle, the curvature information at least includes curvature and curvature change rate of the ego vehicle, and the calculation of the ego vehicle state information of the fourth control point based on the first ego vehicle state information at the current time and the first prediction time period comprises: judging whether the curvature of the ego vehicle at the current time is less than a preset curvature threshold; when the curvature of the ego vehicle at the current time is less than the preset curvature threshold, calculating the ego vehicle state information of the fourth control point based on the position information, speed, acceleration, and first prediction time period of the ego vehicle at the current time; or, when the curvature of the ego vehicle at the current time is not less than the preset curvature threshold, calculating the ego vehicle state information of the fourth control point based on the curvature and curvature change rate of the ego vehicle, the first prediction time period, and the speed, acceleration, and position information of the ego vehicle at the current time; performing trajectory prediction on the first prediction time period based on the ego vehicle state information corresponding to the first, second, third, and fourth control points to obtain a first prediction trajectory segment.

2. The method of claim 1, wherein, the second ego vehicle state information includes a first lateral position, a first lateral speed, and a first lateral acceleration of the ego vehicle at the second prediction time period ahead of the current time, the lane line information at least includes a lane line orientation angle, a lane line length, a lane line offset, and lane line curvature information, and the calculation of the second ego vehicle state information at the second prediction time period ahead of the current time based on the lane line information comprises: inputting the lane line orientation angle, the lane line offset, and the lane line curvature information as lane line coefficients in a road polynomial model and the lane line length as a variable value into the road polynomial model to obtain the first lateral position of the ego vehicle at the second prediction time period ahead of the current time. determining a vehicle head orientation angle based on the lane line orientation angle, the lane line curvature information, and the lane line length; multiplying the vehicle head orientation angle by a longitudinal speed of the ego vehicle to obtain a first lateral speed of the ego vehicle at a second preset time period ahead of a current time; determining a first curvature based on the lane line curvature information and the lane line length; multiplying the first curvature by the longitudinal speed of the ego vehicle to obtain a first lateral acceleration of the ego vehicle at the second preset time period ahead of the current time.

3. The method of claim 1, wherein, The first ego vehicle state information at the current time at least includes speed, acceleration, and position information of the ego vehicle, and before calculating second ego vehicle state information at the second preset time period ahead of the current time based on the lane line information, the method further comprises: detecting whether the speed of the ego vehicle at the current time is greater than a first preset speed threshold; dividing the speed of the ego vehicle at the current time by the acceleration of the ego vehicle to obtain a first time length; determining whether the first time length is within a first preset time length range; If the speed of the ego vehicle at the current time is greater than the first preset speed threshold, and the first time length is not within the first preset time length range, it is determined that the ego vehicle does not satisfy the stop condition within the first prediction time period, and the step of calculating the second ego vehicle state information at the second preset time period ahead of the current time based on the lane line information is executed.

4. The method of claim 2, wherein, The first ego vehicle state information at least includes position information of the ego vehicle, and the position information includes lateral position and longitudinal position. The calculation of the ego vehicle state information of the fourth control point in the first Bezier curve based on the first control point, the second control point, the third control point, and the second ego vehicle state information comprises: calculating the lateral position of the fourth control point based on the first lateral acceleration, the first lateral speed, the lateral position of the third control point, the lateral position of the second control point, and the first lateral position; calculating the longitudinal position of the fourth control point based on the longitudinal position of the third control point, the initial speed of the first control point, the initial acceleration, the lateral position of the fourth control point, and the lateral position of the third control point; calculating the speed and acceleration of the fourth control point based on the position information of the second control point, the third control point, and the fourth control point.

5. The method of claim 3, wherein, The method further comprises: If the speed of the ego vehicle at the current time is not greater than the first preset speed threshold, or the first time length is within the first preset time length range, it is determined that the ego vehicle satisfies the stop condition within the first prediction time period, the position information of the fourth control point is determined as the position information of the third control point, and the speed and acceleration of the fourth control point are calculated based on the position information of the second control point, the third control point, and the fourth control point.

6. The method of claim 3, wherein, The method further comprises: taking the first ego vehicle state information at the current time as the ego vehicle state information of the first control point in the first Bezier curve; When the speed of the ego vehicle at the current time is greater than the first preset speed threshold, and the first time length is not within the first preset time length range, it is determined that the ego vehicle does not satisfy the stop condition in the first prediction time period, and then the position information of the ego vehicle at the second control point is calculated based on the position information and the speed of the ego vehicle at the first control point. The position information of the ego vehicle at the third control point is calculated based on the position information of the ego vehicle at the second control point, the position information of the ego vehicle at the first control point, and the acceleration.

7. The method of claim 6, wherein, The method further comprises: When the speed of the ego vehicle at the current time is not greater than the first preset speed threshold, or the first time length is within the first preset time length range, it is determined that the ego vehicle satisfies the stop condition in the first prediction time period, and then the distance between the ego vehicle and the first control point at the stop time is calculated based on the first time length, the speed and the acceleration of the ego vehicle at the first control point. The position information of the second control point and the third control point is determined based on the position information of the ego vehicle at the first control point and the distance between the ego vehicle and the first control point at the stop time.

8. The method of claim 1, wherein, The lane line information at least includes a lane line length, and the determination of whether the road prediction is valid based on the first ego vehicle state information and the lane line information at the current time comprises: determining whether the speed of the ego vehicle at the current time is greater than a second preset speed threshold; if the speed of the ego vehicle at the current time is greater than the second preset speed threshold, multiplying the speed of the ego vehicle at the current time by a preset second time length to obtain a first distance; determining whether the first distance is greater than the lane line length; if the first distance is not greater than the lane line length, determining whether the target acceleration of the ego vehicle satisfies an acceleration preset condition; if the target acceleration of the ego vehicle satisfies the acceleration preset condition, determining that the road prediction is valid.

9. The method of claim 1, wherein, The calculation of the ego vehicle state information at the fourth control point based on the curvature and the curvature change rate of the ego vehicle, the first prediction time period, and the speed, the acceleration and the position information of the ego vehicle at the current time comprises: taking the reciprocal of the curvature of the ego vehicle at the current time as the ego vehicle driving radius; multiplying the curvature change rate of the ego vehicle at the current time by the first prediction time period to obtain a second curvature; adding the second curvature to the curvature of the ego vehicle at the current time to obtain a curvature sum; taking the reciprocal of the curvature sum as the ego vehicle predicted driving radius; calculating the ego vehicle predicted driving distance based on the speed, the acceleration of the ego vehicle at the current time, and the first prediction time period; dividing the ego vehicle predicted driving distance by the ego vehicle driving radius to obtain a first angle value; dividing the ego vehicle predicted driving distance by the ego vehicle predicted driving radius to obtain a second angle value; calculating the ego vehicle driving first angle based on the first angle value and the ego vehicle driving radius; calculating the ego vehicle driving second angle based on the second angle value and the ego vehicle predicted driving radius; calculating the position information of the fourth control point based on the ego vehicle driving first angle, the ego vehicle driving second angle, the angle preset coefficients corresponding to the ego vehicle driving first angle and the ego vehicle driving second angle, and the position information of the ego vehicle at the current time; calculating the speed and the acceleration of the fourth control point based on the position information of the second control point, the third control point and the fourth control point.

10. The method of claim 8, wherein, The second preset time period is a third prediction time period from the current time, and whether the target acceleration of the ego vehicle satisfies the acceleration preset condition is determined by the following steps: The first ego vehicle state information at the current time is taken as the ego vehicle state information of a first control point in the first Bezier curve; The second ego vehicle state information is taken as the ego vehicle state information of a fourth control point in a third Bezier curve corresponding to the third prediction time period; The ego vehicle state information of the fourth control point in the third Bezier curve, the acceleration of the ego vehicle at the first control point in the first Bezier curve, and the position information of the fourth control point in the first Bezier curve are used to calculate the ego vehicle state information of the fourth control point in the second Bezier curve corresponding to the second prediction time period; The ego vehicle state information of the fourth control point in the third Bezier curve, the acceleration of the ego vehicle at the first control point in the first Bezier curve, and the position information of the fourth control point in the first Bezier curve are used to calculate the ego vehicle state information of the fourth control point in the first Bezier curve corresponding to the first prediction time period; wherein the ego vehicle state information of the fourth control point in the first Bezier curve is the ego vehicle state information of the first control point in the second Bezier curve; and the ego vehicle state information of the fourth control point in the second Bezier curve is the ego vehicle state information of the first control point in the third Bezier curve; The ego vehicle state information of the first control point in each Bezier curve and the ego vehicle state information of the fourth control point in each Bezier curve are used to determine the ego vehicle state information of the second control point and the third control point in each Bezier curve; The ego vehicle state information respectively calculated based on the second control point, the third control point and the fourth control point in each Bezier curve is used to determine an acceleration value set corresponding to each Bezier curve; The acceleration sets corresponding to all the Bezier curves are combined into a first acceleration set, and the maximum acceleration in the first acceleration set corresponding to all the Bezier curves is selected; It is determined whether the maximum acceleration is less than a first preset acceleration threshold; The maximum acceleration in the acceleration set corresponding to the first Bezier curve is subtracted by the lateral acceleration of the ego vehicle at the current time to obtain a first difference value, and it is determined whether the first difference value is less than a first preset difference threshold; If the maximum acceleration is less than the first preset acceleration threshold and the first difference value is less than the first preset difference threshold, it is determined that the target acceleration of the ego vehicle satisfies the acceleration preset condition.

11. The method of claim 10, wherein, The lane line information at least includes a lane line orientation angle, a lane line length, a lane line offset, a lane line curvature and a lane line curvature change rate, and the method further comprises: obtaining the lane line information corresponding to the second prediction time period by the following steps: The square of the lane line length at the current time is calculated to obtain a first square value, and the cube of the lane line length is calculated to obtain a first cube value; The lane line orientation angle at the current time is multiplied by the lane line length to obtain a first product, and the lane line curvature at the current time is multiplied by the first square value to obtain a second product; The lane line curvature change rate at the current time is multiplied by the first cube value to obtain a third product; The first product, the second product, the third product and the lane line offset at the current time are added to obtain the lane line offset corresponding to the second prediction time period. multiplying the first preset coefficient, the lane line curvature at the current moment, and the lane line length to obtain a fourth product; multiplying the second preset coefficient, the lane line curvature change rate at the current moment, and the first product to obtain a fifth product; adding the fourth product, the fifth product, and the heading angle at the current moment to obtain a lane line heading angle corresponding to the second prediction time period; multiplying the third preset coefficient and the lane line curvature at the current moment to obtain a sixth product; multiplying the lane line length at the current moment, the fourth preset coefficient, and the lane line curvature change rate at the current moment to obtain a lane line curvature corresponding to the second prediction time period; taking the lane line curvature change rate at the current moment as a lane line curvature change rate corresponding to the second prediction time period; taking the lane line length at the current moment as a lane line length corresponding to the second prediction time period.

12. The method of claim 11, wherein, The self-vehicle state information at least includes a speed, an acceleration, position information, a curvature, and a curvature change rate of the self-vehicle, and the method further includes: obtaining third self-vehicle state information corresponding to the second prediction time period by the following steps: obtaining a speed of the self-vehicle corresponding to the second prediction time period based on the speed and the acceleration of the self-vehicle at the first prediction time period and the current moment; taking the acceleration of the self-vehicle at the current moment as an acceleration of the self-vehicle corresponding to the second prediction time period; taking position information of the fourth control point of the self-vehicle in the first Bezier curve as position information of the self-vehicle corresponding to the second prediction time period; multiplying the curvature change rate at the current moment by the first prediction time period to obtain a third curvature; adding the third curvature and the curvature at the current moment to obtain a curvature of the self-vehicle corresponding to the second prediction time period.

13. The method of claim 12, wherein, The method further includes: updating the first self-vehicle state information at the current moment and the lane line information at the current moment to third self-vehicle state information and lane line information corresponding to the second prediction time period, and performing the step of obtaining the self-vehicle state information corresponding to the first control point, the second control point, and the third control point of the self-vehicle in the first Bezier curve corresponding to the first prediction time period until the second prediction trajectory segment corresponding to the second prediction time period is obtained.

14. A trajectory prediction device, characterized by, The device includes: an information acquisition module configured to acquire first self-vehicle state information at a current moment, lane line information at the current moment, and self-vehicle state information corresponding to a first control point, a second control point, and a third control point of the self-vehicle in a first Bezier curve corresponding to a first prediction time period; wherein the first control point is a starting point of the first Bezier curve, the second control point is a control point on the side of the starting point, and the third control point is a control point on the side of an end point; a road validity judgment module configured to judge whether a road prediction is valid based on the first self-vehicle state information and the lane line information at the current moment; a second self-vehicle state information acquisition module configured to, if the road prediction is valid, calculate second self-vehicle state information at a second prediction time period away from the current moment based on the lane line information, the second prediction time period being greater than the first prediction time period and being an integer multiple of the first prediction time period. a fourth control point information calculation module, configured to calculate ego vehicle state information of a fourth control point in the first Bezier curve based on ego vehicle state information corresponding to the first control point, the second control point and the third control point respectively and the second ego vehicle state information, the fourth control point being an end point of the first Bezier curve; an ego vehicle state prediction module, configured to calculate ego vehicle state information of the fourth control point based on the first ego vehicle state information of the current time and the first prediction time period if the road prediction is invalid; the first ego vehicle state information at least includes curvature information, speed, acceleration and position information of the ego vehicle, the curvature information at least includes curvature and curvature change rate of the ego vehicle, and the ego vehicle state prediction module comprises: a curvature judgment unit, configured to judge whether the curvature of the ego vehicle at the current time is less than a preset curvature threshold; a state information calculation unit, configured to calculate ego vehicle state information of the fourth control point based on position information, speed, acceleration of the ego vehicle at the current time and the first prediction time period when the curvature of the ego vehicle at the current time is less than the preset curvature threshold, or calculate ego vehicle state information of the fourth control point based on the curvature and the curvature change rate of the ego vehicle, the first prediction time period and speed, acceleration and position information of the ego vehicle at the current time when the curvature of the ego vehicle at the current time is not less than the preset curvature threshold; a trajectory segment prediction module, configured to perform trajectory prediction on the first prediction time period based on ego vehicle state information corresponding to the first control point, the second control point, the third control point and the fourth control point respectively to obtain a first prediction trajectory segment.

15. A vehicle characterized by comprising: The vehicle comprises a controller, the controller comprises a memory and a processor, the memory and the processor are communicatively connected with each other, the memory stores computer instructions, and the processor executes the computer instructions to perform the trajectory prediction method in any one of claims 1 to 13.

Images