Trajectory prediction method, device, vehicle, chip and storage medium

Abstract

The present disclosure proposes a trajectory prediction method, device, vehicle, chip and storage medium, wherein the method includes: obtaining the historical lane change probability distribution of the target vehicle at the previous moment before the current moment; predicting the initial lane change probability distribution of the target vehicle; generating a target lane change probability distribution of the target vehicle based on the historical lane change probability distribution and the initial lane change probability distribution; predicting the driving trajectory of the target vehicle according to the target lane change probability distribution, thereby generating the target lane change probability distribution not only takes into account the past lane change behavior of the target vehicle, but also incorporates the real-time information of the current environment, thereby improving the accuracy of the target lane change probability distribution, thereby improving the accuracy of the driving trajectory prediction.

Description

Technical Field

[0001] The present disclosure relates to the field of autonomous driving technology, and in particular to a trajectory prediction method, device, vehicle, chip, and storage medium. Background Art

[0002] In autonomous driving scenarios, it is very important for autonomous vehicles to predict the driving trajectories of other vehicles. For example, autonomous vehicles predicting the driving trajectories of other vehicles can help identify potential collision risks in advance, optimize path planning, ensure the safety of lane changes and intersections, improve driving efficiency and passenger comfort, while complying with traffic rules and responding to emergencies, thereby comprehensively improving the safety and reliability of the autonomous driving system.

[0003] In the related art, an autonomous driving vehicle predicts its driving trajectory only based on the lane change probability of other vehicles output by a lane change model. However, the lane change model may misjudge or miss the lane changes of other vehicles due to the influence of input data and environmental factors, resulting in low accuracy of the lane change probability of other vehicles output by the lane change model, which in turn causes the autonomous driving vehicle to be unable to accurately predict the driving trajectory of other vehicles. Summary of the invention

[0004] The present disclosure aims to solve one of the technical problems in the related art to a certain extent.

[0005] To this end, the present disclosure proposes a trajectory prediction method, device, vehicle, chip and storage medium. First, by obtaining the historical lane change probability distribution of the target vehicle at the previous moment at the current moment, the past lane change behavior of the target vehicle can be understood, which provides an important reference basis for the prediction of the lane change behavior at the current moment; then, the initial lane change probability distribution of the target vehicle at the current moment is predicted, and the lane change behavior of the target vehicle is preliminarily estimated in combination with the current traffic environment and vehicle status. Then, by combining the historical lane change probability distribution and the initial lane change probability distribution, a target lane change probability distribution is generated, so that the generation of the target lane change probability distribution not only takes into account the past lane change behavior of the target vehicle, but also integrates the real-time information of the current environment, thereby improving the accuracy of the target lane change probability distribution, thereby predicting the target vehicle's driving trajectory based on the target lane change probability distribution, thereby improving the accuracy of the driving trajectory prediction.

[0006] In one aspect, an embodiment of the present disclosure proposes a trajectory prediction method, comprising: obtaining a historical lane change probability distribution of a target vehicle at a previous moment at a current moment; wherein the historical lane change probability distribution is used to indicate the probability of the target vehicle performing different lane change behaviors at the previous moment; predicting an initial lane change probability distribution of the target vehicle at the current moment; generating a target lane change probability distribution of the target vehicle at the current moment based on the historical lane change probability distribution and the initial lane change probability distribution; and predicting the driving trajectory of the target vehicle according to the target lane change probability distribution.

[0007] Another aspect of the present disclosure is directed to a trajectory prediction device, comprising: an acquisition module for acquiring a historical lane change probability distribution of a target vehicle at a previous moment before a current moment; wherein the historical lane change probability distribution is used to indicate the probability of the target vehicle performing different lane change behaviors at the previous moment; a first prediction module for predicting an initial lane change probability distribution of the target vehicle at the current moment; a generation module for generating a target lane change probability distribution of the target vehicle at the current moment based on the historical lane change probability distribution and the initial lane change probability distribution; and a second prediction module for predicting the driving trajectory of the target vehicle according to the target lane change probability distribution.

[0008] Another aspect of the present disclosure provides a vehicle, comprising: a processor, and a memory communicatively connected to the processor; the memory stores computer-executable instructions; the processor executes the computer-executable instructions stored in the memory to implement the trajectory prediction method as described in the above-mentioned first aspect of the embodiment.

[0009] Another aspect of the present disclosure provides a chip, including a processing circuit, wherein the processing circuit is configured to execute the trajectory prediction method as described in the above-mentioned first aspect of the embodiment.

[0010] Another aspect of the present disclosure provides a computer-readable storage medium, wherein the computer-readable storage medium stores computer-executable instructions, and when the computer-executable instructions are executed by a processor, they are used to implement the trajectory prediction method as described in the above-mentioned embodiment of the first aspect.

[0011] Another aspect of the present disclosure provides a computer program product on which a computer program is stored. When the program is executed by a processor, the trajectory prediction method as described in the above-mentioned first aspect is implemented.

[0012] The trajectory prediction method proposed in the present disclosure obtains the historical lane change probability distribution of the target vehicle at the previous moment of the current moment; wherein the historical lane change probability distribution is used to indicate the probability of the target vehicle performing different lane change behaviors at the previous moment; predicts the initial lane change probability distribution of the target vehicle at the current moment; generates the target lane change probability distribution of the target vehicle at the current moment based on the historical lane change probability distribution and the initial lane change probability distribution; and predicts the driving trajectory of the target vehicle according to the target lane change probability distribution, thereby, by obtaining the historical lane change probability distribution of the target vehicle at the previous moment of the current moment, the past lane change behaviors of the target vehicle can be understood. It provides an important reference for predicting the lane changing behavior at the current moment; it predicts the initial lane changing probability distribution of the target vehicle at the current moment, and realizes the preliminary estimation of the lane changing behavior of the target vehicle in combination with the current traffic environment and vehicle status. Then, by combining the historical lane changing probability distribution and the initial lane changing probability distribution, the target lane changing probability distribution is generated. The generation of the target lane changing probability distribution not only takes into account the past lane changing behavior of the target vehicle, but also incorporates the real-time information of the current environment, thereby improving the accuracy of the target lane changing probability distribution. Therefore, the driving trajectory of the target vehicle is predicted based on the target lane changing probability distribution, thereby improving the accuracy of the driving trajectory prediction.

[0013] Additional aspects and advantages of the present disclosure will be given in part in the following description and in part will be obvious from the following description or learned through practice of the present disclosure. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] The above and / or additional aspects and advantages of the present disclosure will become apparent and easily understood from the following description of the embodiments in conjunction with the accompanying drawings, in which:

[0015] Figure 1 A schematic diagram of a flow chart of a trajectory prediction method provided by an embodiment of the present disclosure;

[0016] Figure 2 A schematic diagram of a flow chart of another trajectory prediction method provided by an embodiment of the present disclosure;

[0017] Figure 3 A schematic diagram of a flow chart of another trajectory prediction method provided by an embodiment of the present disclosure;

[0018] Figure 4 A schematic diagram of the principle of a trajectory prediction method provided by an embodiment of the present disclosure;

[0019] Figure 5 A schematic diagram of the structure of a trajectory prediction device provided by an embodiment of the present disclosure;

[0020] Figure 6 is a block diagram of a vehicle according to an exemplary embodiment;

[0021] Figure 7 It is a schematic diagram of the structure of the chip proposed in the embodiment of the present disclosure. DETAILED DESCRIPTION

[0022] Embodiments of the present disclosure are described in detail below, examples of which are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are intended to be used to explain the present disclosure, and should not be construed as limiting the present disclosure.

[0023] The rate matching method, device, vehicle, chip and storage medium of the embodiments of the present disclosure are described below with reference to the accompanying drawings.

[0024] Figure 1 A schematic diagram of a flow chart of a trajectory prediction method provided by an embodiment of the present disclosure.

[0025] This embodiment takes the trajectory prediction method configured as a trajectory prediction device as an example, and the trajectory prediction device can be set on an autonomous driving vehicle.

[0026] like Figure 1 As shown, the trajectory prediction method may include the following steps:

[0027] Step 101, obtaining the historical lane change probability distribution of the target vehicle at the previous moment before the current moment.

[0028] Among them, the historical lane change probability distribution is used to indicate the probability of the target vehicle performing different lane change behaviors at the previous moment.

[0029] As an example, the historical lane change probability distribution of the target vehicle at the previous moment is extracted from the historical data, wherein the historical lane change probability distribution is used to indicate the probability of the target vehicle performing different lane change behaviors at the previous moment, and the lane change behaviors include: lane keeping, left lane change and right lane change, and the target vehicles are other vehicles around the autonomous driving vehicle.

[0030] For example, the historical lane change probability distribution P = [0.6, 0.3, 0.1], which means that the probability of the target vehicle maintaining the lane at the previous moment is 60%, the probability of changing lanes to the left is 30%, and the probability of changing lanes to the right is 10%.

[0031] Step 102, predicting the initial lane change probability distribution of the target vehicle at the current moment.

[0032] The initial lane change probability distribution is used to indicate the initial probability of the target vehicle performing different lane change behaviors at the current moment.

[0033] As a possible implementation method, based on the vehicle state and environmental information of the target vehicle at the current moment, the lane change behavior of the target vehicle is preliminarily predicted, and based on the predicted lane change probability distribution, an initial lane change probability distribution is generated. The initial lane change probability distribution is used to indicate the initial probability of the target vehicle performing different lane change behaviors at the current moment.

[0034] Step 103: Generate a target lane change probability distribution of the target vehicle at the current moment based on the historical lane change probability distribution and the initial lane change probability distribution.

[0035] Among them, the target lane change probability distribution is used to indicate the target probability of the target vehicle performing different lane change behaviors at the current moment.

[0036] In order to improve the prediction accuracy of the lane change behavior of the target vehicle at the current moment, as a possible implementation method, the target lane change probability distribution is generated by integrating the target vehicle's past lane change behavior and the real-time information of the current environment. Among them, the target lane change probability distribution is used to indicate the target probability of the target vehicle performing different lane change behaviors at the current moment. For example, the target lane change probability distribution P = [0.7, 0.2, 0.1] means that the probability of the target vehicle keeping the lane at the current moment is 70%, the probability of changing lanes to the left is 20%, and the probability of changing lanes to the right is 10%.

[0037] As an example, the historical lane change probability distribution and the initial lane change probability distribution are combined to generate a target lane change probability distribution of the target vehicle.

[0038] Step 104 , predicting the driving trajectory of the target vehicle according to the target lane change probability distribution.

[0039] As an example, the driving trajectory of the target vehicle is predicted based on the lane changing behavior with the highest probability in the target lane changing probability distribution. For example, if the lane changing behavior with the highest probability in the target lane changing probability distribution is lane keeping, it is predicted that the target vehicle will continue to drive along the current lane.

[0040] As another example, a corresponding trajectory is generated according to the probability of each lane change behavior in the target lane change probability distribution, and the driving trajectory of the target vehicle is generated by combining the driving trajectories corresponding to each lane change behavior. For example, the driving trajectories of each lane change behavior are weighted to obtain the driving trajectory of the target vehicle.

[0041] In summary, by obtaining the historical lane change probability distribution of the target vehicle at the previous moment before the current moment, we can understand the target vehicle's past lane change behavior, which provides an important reference for predicting the lane change behavior at the current moment; predicting the initial lane change probability distribution of the target vehicle at the current moment realizes the preliminary estimation of the target vehicle's lane change behavior in combination with the current traffic environment and vehicle status, and then, by combining the historical lane change probability distribution and the initial lane change probability distribution, the target lane change probability distribution is generated, so that the generation of the target lane change probability distribution not only takes into account the target vehicle's past lane change behavior, but also incorporates the real-time information of the current environment, thereby improving the accuracy of the target lane change probability distribution, thereby predicting the target vehicle's driving trajectory based on the target lane change probability distribution, thereby improving the accuracy of the driving trajectory prediction.

[0042] In order to clearly illustrate how the target lane change probability distribution of the target vehicle is generated based on the historical lane change probability distribution and the initial lane change probability distribution in the above embodiment, the present disclosure proposes another trajectory prediction method.

[0043] Figure 2 A flowchart of another trajectory prediction method provided by an embodiment of the present disclosure.

[0044] like Figure 2 As shown, the trajectory prediction method may include the following steps:

[0045] Step 201, obtaining the historical lane change probability distribution of the target vehicle at the previous moment before the current moment.

[0046] Among them, the historical lane change probability distribution is used to indicate the probability of the target vehicle performing different lane change behaviors at the previous moment.

[0047] Step 202, predicting the initial lane change probability distribution of the target vehicle at the current moment.

[0048] The initial lane change probability distribution is used to indicate the initial probability of the target vehicle performing different lane change behaviors at the current moment.

[0049] Step 203, obtaining the state transition probability matrix of the target vehicle.

[0050] Among them, the element in the i-th row and j-th column of the state transition probability matrix is ​​used to indicate the probability of the target vehicle switching from the i-th lane change behavior at the previous moment to the j-th lane change behavior at the current moment.

[0051] As an example, test data of a target vehicle is collected, wherein the test data includes lane change behavior records of the target vehicle at different time points, and a state transition probability matrix is ​​generated based on the test data. The state transition probability matrix is ​​a two-dimensional matrix, and the element in the i-th row and j-th column of the state transition probability matrix is ​​used to indicate the probability of the target vehicle switching from the i-th lane change behavior at the previous moment to the j-th lane change behavior at the current moment.

[0052] For example, the state transition probability matrix of the target vehicle is , the state transition probability matrix can be represented by Table 1.

[0053] Table 1 State transition probability matrix

[0054]

[0055] That is to say, in Table 1, the lane change behavior at the previous moment was "lane keeping", the probability of continuing "lane keeping" at the current moment is 0.7, the probability of changing to "left lane change" is 0.3, and the probability of changing to "right lane change" is 0; the lane change behavior at the previous moment was "left lane change", the probability of continuing to "left lane change" at the current moment is 0.7, the probability of changing to "right lane change" is 0, and the probability of changing to "lane keeping" is 0.3; the lane change behavior at the previous moment was "right lane change", the probability of continuing to "right lane change" at the current moment is 0.7, the probability of changing to "left lane change" is 0, and the probability of changing to "lane keeping" is 0.3.

[0056] Step 204 , using the state transition probability matrix, transforms the historical lane change probability distribution to obtain the transition probability distribution of the target vehicle at the current moment.

[0057] It should be understood that since the state transition probability matrix includes the transition rules between the lane changing behavior of the target vehicle at the previous moment and the lane changing behavior at the current moment, in order to improve the prediction accuracy of the lane changing behavior of the target vehicle at the current moment, as a possible implementation method, the state transition probability matrix is ​​used to convert the historical lane changing probability distribution into the conversion probability distribution at the current moment.

[0058] As an example, the transition probability distribution of the target vehicle at the current moment is determined based on the product of the state transition probability matrix and the historical lane change probability distribution.

[0059] For example, the historical lane change probability distribution P = [0.6, 0.3, 0.1], the state transition probability matrix T = , the product of the state transition probability matrix T and the historical lane change probability distribution P = , then the transition probability distribution is .

[0060] Step 205 : generating a target lane change probability distribution based on the initial lane change probability distribution and the conversion probability distribution.

[0061] In order to further improve the prediction accuracy of the lane changing behavior of the target vehicle at the current moment, as a possible implementation method, a Bayesian filter is used to update the probability distribution by combining the transition probability distribution (as the prior probability in the Bayesian formula) and the initial lane change probability distribution (as the observation probability in the Bayesian update formula) to obtain the updated probability distribution. Based on the updated probability distribution, the target lane change probability distribution is generated.

[0062] As an example, based on the initial lane change probability distribution, the conversion probability distribution is updated to obtain an updated conversion probability distribution; the updated conversion probability distribution is filtered to obtain a filtered probability distribution; from the filtered probability distribution, the probability to be adjusted that matches the lane change intention of the target vehicle at the current moment is determined; according to the distance between the target vehicle and the target boundary line of the current driving lane, the probability adjustment factor of the probability to be adjusted is determined; wherein the target boundary line matches the lane change intention; according to the probability adjustment factor, the probability to be adjusted in the filtered probability distribution is adjusted to obtain the target lane change probability distribution.

[0063] That is to say, the initial lane change probability distribution is used as the observation probability in the Bayesian update formula, and the transition probability distribution is updated based on the Bayesian update formula to obtain the updated transition probability distribution; for example, the Bayesian update formula ,in, represents the prior probability, represents the observation probability, represents normalized constants, such as the transition probability distribution ( Q = [0.5002, 0.1998, 0.3] ), the initial lane change probability distribution ( R = [0.5, 0.3, 0.2]), and the calculation of the unnormalized , and , , , , Z = 0.2501 + 0.05994 + 0.06 = 0.37004, normalized =0.2501 / 0.37004=0.6759, normalized =0.05994 / 0.37004=0.1620, normalized =0.06 / 0.37004=0.1621, the updated transition probability distribution = [0.6759, 0.1620, 0.1621].

[0064] In order to improve the accuracy of the lane change probability distribution, as a possible implementation method, a sliding window average filtering is performed on the updated conversion probability distribution to obtain a filtered probability distribution.

[0065] As an example, the lane change probability distributions at multiple consecutive moments are weighted averaged to obtain a filtered probability distribution. For example, the updated conversion probability distribution, the historical lane change probability distribution at the previous moment, and the historical lane change probability distributions at the previous two moments are weighted averaged to obtain a filtered probability distribution, where the updated conversion probability distribution has the largest weight.

[0066] Furthermore, in order to further improve the accuracy of the lane change probability distribution, as a possible implementation method, based on the lane change intention of the target vehicle at the current moment, the filtered probability distribution is adjusted to obtain the target lane change probability distribution.

[0067] As an example, based on the driving status information of the target vehicle at the current moment, the lane changing intention of the target vehicle at the current moment is determined, and from the filtered probability distribution, the probability to be adjusted that matches the lane changing intention of the target vehicle at the current moment is determined, and based on the distance between the target vehicle and the target boundary line of the current driving lane, the probability adjustment factor of the probability to be adjusted is determined; wherein, the target boundary line matches the lane changing intention, and according to the probability adjustment factor, the probability to be adjusted in the filtered probability distribution is adjusted to obtain the target lane changing probability distribution.

[0068] That is to say, according to the lane change intention model, based on the target vehicle's driving state information at the current moment (such as speed, acceleration, position information, etc.), the lane change intention of the target vehicle at the current moment (such as left lane change, right lane change) is determined. When the lane change intention is left lane change, the left lane change probability in the filtered probability distribution is used as the probability to be adjusted, and the distance between the target vehicle and the left boundary line of the current driving lane is obtained. The probability adjustment factor is determined according to the distance between the target vehicle and the left boundary line of the current driving lane, wherein the probability adjustment factor is expressed as the following formula: ,in, represents the distance between the target vehicle and the left boundary line of the current driving lane, and d is a set constant; the left lane change probability in the filtered probability distribution is multiplied by the probability adjustment factor to adjust the left lane change probability in the filtered probability distribution to obtain the target lane change probability distribution.

[0069] Among them, when the lane change intention is to change right, the right lane change probability in the filtered probability distribution is used as the probability to be adjusted, the distance between the target vehicle and the right boundary line of the current driving lane is obtained, and the probability adjustment factor is determined according to the distance between the target vehicle and the right boundary line of the current driving lane; the right lane change probability in the filtered probability distribution is multiplied by the probability adjustment factor to adjust the right lane change probability in the filtered probability distribution to obtain the target lane change probability distribution.

[0070] It should be noted that, according to the probability adjustment factor, the probability to be adjusted in the filtered probability distribution is adjusted, which reduces the lane changing probability of target vehicles that are far away from the lane line, and increases the lane changing probability of target vehicles that are close to the lane line or have crossed the line.

[0071] In addition, it should be noted that when the target vehicle has no intention to change lanes, there is no need to adjust the filtered probability distribution, and the filtered probability distribution is directly used as the target lane change probability distribution.

[0072] Step 206, predicting the driving trajectory of the target vehicle according to the target lane change probability distribution.

[0073] It should be noted that the execution process of steps 201 to 202 and step 206 can be implemented in any way in the embodiments of the present disclosure, and the embodiments of the present disclosure do not limit this and will not be described in detail.

[0074] In summary, by obtaining the state transition probability matrix of the target vehicle; using the state transition probability matrix, the historical lane change probability distribution is transformed to obtain the conversion probability distribution of the target vehicle at the current moment; based on the initial lane change probability distribution and the conversion probability distribution, the target lane change probability distribution is generated. Therefore, combined with the historical lane change probability and the possibility of current state transfer, the conversion probability distribution of the target vehicle at the current moment can be accurately obtained, and then the initial lane change probability distribution and the conversion probability distribution are combined. It can more comprehensively understand the lane change behavior of the target vehicle at the current moment, dynamically adjust the prediction results to adapt to the ever-changing traffic environment, reduce the risk of misjudgment and missed judgment, thereby optimizing path planning and improving driving safety and efficiency.

[0075] In order to clearly illustrate how the initial lane change probability distribution of the target vehicle at the current moment is predicted in the above embodiment, the present disclosure proposes a flow chart of another trajectory prediction method.

[0076] Figure 3 A flowchart of another trajectory prediction method provided by an embodiment of the present disclosure.

[0077] like Figure 3 As shown, the trajectory prediction method may include the following steps:

[0078] Step 301, obtaining the historical lane change probability distribution of the target vehicle at the previous moment before the current moment.

[0079] Among them, the historical lane change probability distribution is used to indicate the probability of the target vehicle performing different lane change behaviors at the previous moment.

[0080] Step 302, obtaining driving status information of the target vehicle at the current moment and multiple historical moments before the current moment.

[0081] As an example, the driving status information of the target vehicle at the current moment and multiple historical moments between the current moment is obtained from the stored data of the target vehicle, wherein the driving status information includes: the position, speed, acceleration, direction, lane position, etc. of the target vehicle.

[0082] Step 303: Input multiple driving state information into the lane change prediction model to obtain the predicted lane change probability distribution of the target vehicle at the current moment output by the lane change prediction model.

[0083] As a possible implementation method, a lane change prediction model is used to predict the lane change behavior of the target vehicle at the current moment.

[0084] As an example, multiple driving status information is input into the lane change prediction model to obtain the predicted lane change probability distribution of the target vehicle at the current moment output by the lane change prediction model. For example, the predicted lane change probability distribution is [0.6, 0.3, 0.1], indicating that the probability of the target vehicle keeping the lane at the current moment is 60%, the probability of changing lanes to the left is 30%, and the probability of changing lanes to the right is 10%.

[0085] Step 304 , correcting the predicted lane change probability distribution to obtain an initial lane change probability distribution.

[0086] In order to improve the rationality of the initial probability of the target vehicle performing different lane changing behaviors at the current moment, as a possible implementation method, the predicted lane changing probability distribution output by the lane changing prediction model is corrected, and the initial probability of the target vehicle performing different lane changing behaviors at the current moment is determined based on the corrected predicted lane changing probability distribution.

[0087] As a first possible implementation method, a heading angle deviation between a heading angle of a target vehicle at a current moment and a heading angle of a target lane is predicted; wherein the target lane is the lane to which the target vehicle changes to when performing a corresponding lane change behavior at the current moment; in response to a heading angle deviation being greater than a set deviation threshold, a first probability to be corrected in the predicted lane change probability distribution is suppressed to obtain an initial lane change probability distribution; wherein the first probability to be corrected is used to indicate the probability of a lane change behavior corresponding to the target vehicle changing lanes to the target lane.

[0088] That is to say, the deviation between the heading angle of the target vehicle at the current moment and the heading angle of the target lane to which it plans to change lanes is calculated. If the calculated heading angle deviation exceeds the preset deviation threshold, it means that there is a large inconsistency between the current direction of the target vehicle and the direction of the lane to which it plans to change lanes. In this case, in order to more accurately reflect the actual possibility of the target vehicle executing the lane change behavior in the current situation, the first probability to be corrected in the predicted lane change probability distribution output by the lane change prediction model is adjusted. That is, the probability value of the lane change behavior corresponding to the target vehicle changing lanes to the target lane in the predicted lane change probability distribution is suppressed. The suppression operation reduces the possibility that the lane change prediction model believes that the target vehicle will execute the lane change behavior under the current conditions, thereby obtaining an adjusted initial lane change probability distribution, so that the prediction of the lane change prediction model is more in line with the actual situation and the accuracy of the driving trajectory prediction is improved.

[0089] As a second possible implementation method, the lane changing time of the target vehicle's most recent lane change is obtained; in response to the time difference between the lane changing time and the current time being less than a set duration, the second probability to be corrected in the predicted lane changing probability distribution is suppressed to obtain an initial lane changing probability distribution; wherein the second probability to be corrected is used to indicate the probability of the target vehicle changing lanes to a lane that matches the lane changing behavior of the most recent lane change.

[0090] That is to say, firstly, the specific time of the target vehicle's most recent lane change, i.e., the lane change time, is recorded, and then the time difference between this lane change time and the current time is calculated. If the time difference is less than the set duration, it is considered that the target vehicle has already performed a lane change behavior in a short period of time, and therefore the probability of the target vehicle performing the same lane change behavior again (i.e., changing lanes to a lane that matches the most recent lane change behavior) is relatively low. Therefore, the second probability to be corrected in the predicted lane change probability distribution output by the lane change prediction model is suppressed, i.e., the probability value indicating that the target vehicle changes lanes to a lane that matches the most recent lane change behavior in the predicted lane change probability distribution is suppressed; by suppressing the second probability to be corrected, the possibility that the lane change prediction model believes that the target vehicle will repeat similar lane change behaviors in a short period of time is reduced, thereby obtaining an initial lane change probability distribution that is more in line with the actual situation.

[0091] As a third possible implementation method, first driving information of the target vehicle at the current moment and second driving information at the previous moment are obtained; based on the first driving information and the second driving information, the heading angle difference between the target vehicle at the current moment and the previous moment is obtained, and it is determined whether the target vehicle has an intention to change lanes at the current moment; if the target vehicle has no intention to change lanes and / or the heading angle difference is less than a set difference threshold, the third probability to be corrected in the predicted lane change probability distribution is suppressed to obtain an initial lane change probability distribution; wherein the third probability to be corrected is used to indicate the probability of a lane change behavior of changing lanes in the lane in which the target vehicle was traveling at the previous moment; if the target vehicle has an intention to change lanes and / or the heading angle difference is greater than the set difference threshold, the fourth probability to be corrected in the predicted lane change probability distribution is increased to obtain an initial lane change probability distribution; wherein the fourth probability to be corrected is used to indicate the probability of a lane change behavior of changing lanes in the lane in which the target vehicle was traveling at the previous moment.

[0092] That is to say, by analyzing the driving information of the target vehicle at the current moment and the previous moment, the heading angle difference between the target vehicle at the current moment and the previous moment is calculated, and based on the driving information at the current moment and the previous moment, it is judged whether the target vehicle has the intention to change lanes. If it is judged that the target vehicle has no intention to change lanes at the current moment, or even if there is an intention to change lanes but the heading angle difference is small (less than the set difference threshold), then the probability value indicating the lane change in the lane in which the target vehicle was traveling at the previous moment (i.e., the third probability to be corrected) in the predicted lane change probability distribution is reduced (suppressed), thereby obtaining the initial lane change probability distribution. On the contrary, if the target vehicle has the intention to change lanes, or the heading angle difference is significant (greater than the set difference threshold), then the system will correspondingly increase (increase) the probability value indicating the lane change in the lane in which the target vehicle was traveling at the previous moment (i.e., the fourth probability to be corrected) in the predicted lane change probability distribution to obtain the initial lane change probability distribution.

[0093] It should be noted that the above is only an example of executing one of the possible implementation methods. In actual application, two or three of the three possible implementation methods can be executed simultaneously, and the present disclosure does not limit this.

[0094] Step 305 , based on the historical lane change probability distribution and the initial lane change probability distribution, generate a target lane change probability distribution of the target vehicle at the current moment.

[0095] Among them, the target lane change probability distribution is used to indicate the target probability of the target vehicle performing different lane change behaviors at the current moment.

[0096] Step 306: predict the driving trajectory of the target vehicle according to the target lane change probability distribution.

[0097] It should be noted that the execution process of step 301 and step 305 to step 306 can be implemented in any way in the embodiments of the present disclosure, and the embodiments of the present disclosure do not limit this and will not be described in detail.

[0098] In summary, by obtaining the driving status information of the target vehicle at the current moment and multiple historical moments before the current moment; inputting multiple driving status information into the lane change prediction model to obtain the predicted lane change probability distribution of the target vehicle at the current moment output by the lane change prediction model; correcting the predicted lane change probability distribution to obtain the initial lane change probability distribution, thereby, by inputting the driving status information of the target vehicle at the current moment and multiple historical moments into the lane change prediction model to obtain the predicted lane change probability distribution, and correcting the predicted lane change probability distribution, the accuracy of the initial lane change probability distribution is improved, so that based on the more accurate initial lane change probability distribution, the automatic driving system can plan the path in advance, avoid potential risk areas, and ensure driving safety and efficiency.

[0099] In any embodiment of the present disclosure, if Figure 4 As shown, the trajectory prediction method of the embodiment of the present disclosure can also be implemented based on the following steps:

[0100] 1. Obtain the probability of the target vehicle output by the lane change prediction model (i.e., obtain the predicted lane change probability distribution of the target vehicle output by the lane change prediction model at the current moment).

[0101] 2. Rule correction

[0102] (1) If the heading angle deviation of the target vehicle from the target lane is too large, the lane change probability is suppressed;

[0103] (2) If the target vehicle has just completed a lane change (within 0.5 s), it is almost impossible to change lanes again immediately, which suppresses the probability of lane change;

[0104] (3) When the target vehicle has almost no angle change and lane change intention, the lane change probability is suppressed;

[0105] (4) When the target vehicle has an extremely obvious angle change and lane change intention, the lane change probability is increased.

[0106] Based on the above rules, the predicted lane change probability distribution is corrected to obtain the initial lane change probability distribution.

[0107] 3. Bayesian filter

[0108] A Bayesian filter is used to update the probability distribution by combining the transition probability distribution (as the prior probability in the Bayesian formula) and the initial lane change probability distribution (as the observation probability in the Bayesian update formula) to obtain the updated transition probability distribution.

[0109] 4. Average filter

[0110] The updated conversion probability distribution is subjected to sliding window average filtering to obtain a filtered probability distribution; for example, the probabilities of the past three consecutive frames are weighted averaged, that is, the updated conversion probability distribution, the historical lane change probability distribution at the previous moment, and the historical lane change probability distribution at the previous two moments are weighted averaged to obtain a filtered probability distribution, wherein the updated conversion probability distribution has the largest weight.

[0111] 5. Decreasing filtering

[0112] (1) Decreasing formula: ; Indicates the distance between the target vehicle and the target boundary line of the current driving lane, and d is a set constant;

[0113] (2) The target vehicle's lane change probability decays exponentially according to the distance between the vehicle body and the lane line;

[0114] (3) The goal is to minimize the lane-changing probability of target vehicles that are far from the lane line and to increase the lane-changing probability of target vehicles that are close to the lane line or have crossed the lane line.

[0115] In order to implement the above embodiment, the embodiment of the present disclosure also proposes a trajectory prediction device.

[0116] Figure 5 A schematic diagram of the structure of a trajectory prediction device provided in an embodiment of the present disclosure.

[0117] like Figure 5 As shown, the trajectory prediction device 500 includes: an acquisition module 510 , a first prediction module 520 , a generation module 530 and a second prediction module 540 .

[0118] Among them, the acquisition module 510 is used to obtain the historical lane change probability distribution of the target vehicle at the previous moment of the current moment; wherein the historical lane change probability distribution is used to indicate the probability of the target vehicle performing different lane change behaviors at the previous moment; the first prediction module 520 is used to predict the initial lane change probability distribution of the target vehicle at the current moment; the generation module 530 is used to generate the target lane change probability distribution of the target vehicle at the current moment based on the historical lane change probability distribution and the initial lane change probability distribution; the second prediction module 540 is used to predict the driving trajectory of the target vehicle according to the target lane change probability distribution.

[0119] As a possible implementation method, a generation module 530 is used to obtain a state transition probability matrix of the target vehicle; wherein the element in the i-th row and j-th column of the state transition probability matrix is ​​used to indicate the probability of the target vehicle switching from the i-th lane change behavior at the previous moment to the j-th lane change behavior at the current moment, and i and j are positive integers; the state transition probability matrix is ​​used to transform the historical lane change probability distribution to obtain the conversion probability distribution of the target vehicle at the current moment; based on the initial lane change probability distribution and the conversion probability distribution, a target lane change probability distribution is generated; wherein the target lane change probability distribution is used to indicate the target probability of the target vehicle performing different lane change behaviors at the current moment.

[0120] As a possible implementation method, a generation module 530 is used to update the conversion probability distribution based on the initial lane change probability distribution to obtain an updated conversion probability distribution; filter the updated conversion probability distribution to obtain a filtered probability distribution; determine, from the filtered probability distribution, a probability to be adjusted that matches the lane change intention of the target vehicle at the current moment; determine a probability adjustment factor of the probability to be adjusted based on the distance between the target vehicle and a target boundary line of the current driving lane; wherein the target boundary line matches the lane change intention; and adjust the probability to be adjusted in the filtered probability distribution based on the probability adjustment factor to obtain a target lane change probability distribution.

[0121] As a possible implementation method, the first prediction module 520 is used to obtain the driving status information of the target vehicle at the current moment and multiple historical moments before the current moment; input the multiple driving status information into the lane change prediction model to obtain the predicted lane change probability distribution of the target vehicle at the current moment output by the lane change prediction model; and correct the predicted lane change probability distribution to obtain the initial lane change probability distribution.

[0122] As a possible implementation method, the first prediction module 520 is used to predict the heading angle deviation between the heading angle of the target vehicle at the current moment and the heading angle of the target lane; wherein the target lane is the lane to which the target vehicle changes to when performing the corresponding lane change behavior at the current moment; in response to the heading angle deviation being greater than the set deviation threshold, the first probability to be corrected in the predicted lane change probability distribution is suppressed to obtain an initial lane change probability distribution; wherein the first probability to be corrected is used to indicate the probability of the lane change behavior corresponding to the target vehicle changing lanes to the target lane.

[0123] As a possible implementation method, the first prediction module 520 is used to obtain the lane changing time of the target vehicle's most recent lane change; in response to the time difference between the lane changing time and the current time being less than a set duration, the second probability to be corrected in the predicted lane changing probability distribution is suppressed to obtain an initial lane changing probability distribution; wherein the second probability to be corrected is used to indicate the probability of the target vehicle changing lanes to a lane that matches the lane changing behavior of the most recent lane change.

[0124] As a possible implementation method, the first prediction module 520 is used to obtain the first driving information of the target vehicle at the current moment and the second driving information at the previous moment; based on the first driving information and the second driving information, obtain the heading angle difference between the target vehicle at the current moment and the previous moment, and judge whether the target vehicle has the intention to change lanes at the current moment; if the target vehicle has no intention to change lanes and / or the heading angle difference is less than the set difference threshold, suppress the third probability to be corrected in the predicted lane change probability distribution to obtain the initial lane change probability distribution; wherein the third probability to be corrected is used to indicate the probability of lane changing behavior of changing lanes in the lane in which the target vehicle was traveling at the previous moment.

[0125] As a possible implementation method, the first prediction module 520 is used to increase the fourth probability to be corrected in the predicted lane change probability distribution to obtain an initial lane change probability distribution if the target vehicle has the intention to change lanes and / or the heading angle difference is greater than a set difference threshold; wherein the fourth probability to be corrected is used to indicate the probability of changing lanes for the lane in which the target vehicle was traveling at the previous moment.

[0126] The trajectory prediction device proposed in the present disclosure obtains the historical lane change probability distribution of the target vehicle at the previous moment of the current moment; wherein the historical lane change probability distribution is used to indicate the probability of the target vehicle performing different lane change behaviors at the previous moment; predicts the initial lane change probability distribution of the target vehicle at the current moment; wherein the initial lane change probability distribution is used to indicate the initial probability of the target vehicle performing different lane change behaviors at the current moment; based on the historical lane change probability distribution and the initial lane change probability distribution, generates the target lane change probability distribution of the target vehicle at the current moment; according to the target lane change probability distribution, predicts the driving trajectory of the target vehicle, thereby obtaining the historical lane change probability distribution of the target vehicle at the previous moment of the current moment Distribution can be used to understand the target vehicle's past lane-changing behavior, which provides an important reference for predicting the lane-changing behavior at the current moment; the initial lane-changing probability distribution of the target vehicle at the current moment is predicted, which realizes the preliminary estimation of the target vehicle's lane-changing behavior in combination with the current traffic environment and vehicle status. Then, by combining the historical lane-changing probability distribution and the initial lane-changing probability distribution, the target lane-changing probability distribution is generated. The generation of the target lane-changing probability distribution not only takes into account the target vehicle's past lane-changing behavior, but also incorporates the real-time information of the current environment, thereby improving the accuracy of the target lane-changing probability distribution. Therefore, the target vehicle's driving trajectory is predicted based on the target lane-changing probability distribution, thereby improving the accuracy of the driving trajectory prediction.

[0127] In order to implement the above-mentioned embodiments, the present disclosure also proposes a vehicle, comprising a processor and a memory communicatively connected to the processor; the memory stores computer-executable instructions; the processor executes the computer-executable instructions stored in the memory to implement the trajectory prediction method as described in the above-mentioned method embodiments.

[0128] In order to implement the above embodiment, the present disclosure further proposes a chip, wherein the chip includes a processing circuit, and the processing circuit is configured to execute the trajectory prediction method as described in the above method embodiment.

[0129] In order to implement the above embodiment, the present disclosure further proposes a computer-readable storage medium, in which computer-executable instructions are stored. When the computer-executable instructions are executed by a processor, they are used to implement the trajectory prediction method as described in the above method embodiment.

[0130] In order to implement the above embodiments, the present disclosure further proposes a computer program product, on which a computer program is stored. When the computer program is executed by a processor, the trajectory prediction method as described in the above method embodiments is implemented.

[0131] Figure 66 is a block diagram of a vehicle 600 according to an exemplary embodiment. For example, the vehicle 600 may be a hybrid vehicle, a non-hybrid vehicle, an electric vehicle, a fuel cell vehicle, or other types of vehicles. The vehicle 600 may be an autonomous vehicle, a semi-autonomous vehicle, or a non-autonomous vehicle.

[0132] Reference Figure 6 , the vehicle 600 may include various subsystems, for example, an infotainment system 610, a perception system 620, a decision control system 630, a drive system 640, and a computing platform 650. The vehicle 600 may also include more or fewer subsystems, and each subsystem may include multiple components. In addition, each subsystem and each component of the vehicle 600 may be interconnected by wire or wireless means.

[0133] In some embodiments, the infotainment system 610 may include a communication system, an entertainment system, and a navigation system, among others.

[0134] The perception system 620 may include several sensors for sensing information about the environment around the vehicle 800. For example, the perception system 620 may include a global positioning system (the global positioning system may be a GPS system, or a Beidou system or other positioning systems), an inertial measurement unit (IMU), a laser radar, a millimeter wave radar, an ultrasonic radar, and a camera.

[0135] The decision control system 630 may include a computing system, a vehicle controller, a steering system, a throttle, and a braking system.

[0136] The drive system 640 may include components that provide powered motion for the vehicle 600. In one embodiment, the drive system 640 may include an engine, an energy source, a transmission system, and wheels. The engine may be one or a combination of an internal combustion engine, an electric motor, and an air compression engine. The engine is capable of converting energy provided by the energy source into mechanical energy.

[0137] Some or all functions of the vehicle 600 are controlled by a computing platform 650. The computing platform 650 may include at least one processor 651 and a memory 652, and the processor 651 may execute instructions 653 stored in the memory 652.

[0138] The processor 651 may be any conventional processor, such as a commercially available CPU. The processor may also include a graphics processor (Graphic Process Unit, GPU), a field programmable gate array (Field Programmable Gate Array, FPGA), a system on chip (System on Chip, SOC), an application specific integrated circuit (Application Specific Integrated Circuit, ASIC) or a combination thereof.

[0139] The memory 652 may be implemented by any type of volatile or nonvolatile memory device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic memory, flash memory, magnetic disk or optical disk.

[0140] In addition to the instructions 653 , the memory 652 may also store data, such as road maps, route information, and data such as the location, direction, and speed of the vehicle. The data stored in the memory 652 may be used by the computing platform 650 .

[0141] In the embodiment of the present disclosure, the processor 651 may execute instruction 653 to complete all or part of the steps of the above method embodiment.

[0142] Figure 7 is a schematic diagram of the structure of the chip proposed in the embodiment of the present disclosure. Figure 7 The structure of the chip 700 is shown, but is not limited to this.

[0143] The chip 700 includes a processing circuit 701 , and the processing circuit 701 is configured to execute any of the above methods.

[0144] In some embodiments, the chip 700 further includes one or more interface circuits 702. Optionally, the interface circuit 702 is connected to the memory 703, and the interface circuit 702 can be used to receive signals from the memory 703 or other devices, and the interface circuit 702 can be used to send signals to the memory 703 or other devices. For example, the interface circuit 702 can read instructions stored in the memory 703 and send the instructions to the processing circuit 701.

[0145] In some embodiments, the interface circuit 702 performs at least one of the communication steps such as sending and / or receiving in the above method, and the processing circuit 701 performs other steps.

[0146] In some embodiments, terms such as interface circuit, interface, transceiver pin, and transceiver may be used interchangeably.

[0147] In some embodiments, the chip 700 further includes one or more memories 703 for storing instructions. Optionally, all or part of the memory 703 may be outside the chip 700.

[0148] In the description of this specification, the description with reference to the terms "one embodiment", "some embodiments", "example", "specific example", or "some examples" etc. means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present disclosure. In this specification, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described may be combined in any one or more embodiments or examples in a suitable manner. In addition, those skilled in the art may combine and combine the different embodiments or examples described in this specification and the features of the different embodiments or examples, without contradiction.

[0149] In addition, the terms "first" and "second" are used for descriptive purposes only and should not be understood as indicating or implying relative importance or implicitly indicating the number of the indicated technical features. Thus, a feature defined as "first" or "second" may explicitly or implicitly include at least one of the features. In the description of the present disclosure, "plurality" means at least two, such as two, three, etc., unless otherwise clearly and specifically defined.

[0150] Any process or method description in a flowchart or otherwise described herein may be understood to represent a module, segment or portion of code that includes one or more executable instructions for implementing the steps of a custom logical function or process, and the scope of the preferred embodiments of the present disclosure includes additional implementations in which functions may not be performed in the order shown or discussed, including performing functions in a substantially simultaneous manner or in reverse order depending on the functions involved, which should be understood by technicians in the technical field to which the embodiments of the present disclosure belong.

[0151] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as an ordered list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by an instruction execution system, device or apparatus (such as a computer-based system, a system including a processor, or other system that can fetch instructions from an instruction execution system, device or apparatus and execute instructions), or in combination with these instruction execution systems, devices or apparatuses. For the purposes of this specification, "computer-readable medium" can be any device that can contain, store, communicate, propagate or transmit a program for use by an instruction execution system, device or apparatus, or in combination with these instruction execution systems, devices or apparatuses. More specific examples (non-exhaustive list) of computer-readable media include the following: an electrical connection with one or more wires (electronic device), a portable computer disk box (magnetic device), a random access memory (RAM), a read-only memory (ROM), an erasable and programmable read-only memory (EPROM or flash memory), an optical fiber device, and a portable compact disk read-only memory (CDROM). In addition, the computer-readable medium may even be paper or other suitable medium on which the program is printed, since the program may be obtained electronically, for example, by optically scanning the paper or other medium and then editing, interpreting or processing in other suitable ways if necessary, and then stored in a computer memory.

[0152] It should be understood that the various parts of the present disclosure can be implemented in hardware, software, firmware or a combination thereof. In the above-mentioned embodiments, multiple steps or methods can be implemented in software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented in hardware, as in another embodiment, it can be implemented by any one of the following technologies known in the art or their combination: a discrete logic circuit having a logic gate circuit for implementing a logic function for a data signal, a dedicated integrated circuit having a suitable combination of logic gate circuits, a programmable gate array (PGA), a field programmable gate array (FPGA), etc.

[0153] A person skilled in the art may understand that all or part of the steps in the method for implementing the above-mentioned embodiment may be completed by instructing related hardware through a program, and the program may be stored in a computer-readable storage medium, which, when executed, includes one or a combination of the steps of the method embodiment.

[0154] In addition, each functional unit in each embodiment of the present disclosure may be integrated into a processing module, or each unit may exist physically separately, or two or more units may be integrated into one module. The above-mentioned integrated module may be implemented in the form of hardware or in the form of a software functional module. If the integrated module is implemented in the form of a software functional module and sold or used as an independent product, it may also be stored in a computer-readable storage medium.

[0155] The storage medium mentioned above may be a read-only memory, a disk or an optical disk, etc. Although the embodiments of the present disclosure have been shown and described above, it can be understood that the above embodiments are exemplary and cannot be understood as limitations of the present disclosure. A person of ordinary skill in the art may change, modify, replace and modify the above embodiments within the scope of the present disclosure.

Claims

1. A trajectory prediction method, characterized in that: include: Obtaining a historical lane change probability distribution of the target vehicle at a previous moment at the current moment; wherein the historical lane change probability distribution is used to indicate the probability of the target vehicle performing different lane change behaviors at the previous moment; Predicting an initial lane change probability distribution of the target vehicle at the current moment; Based on the historical lane change probability distribution and the initial lane change probability distribution, generating a target lane change probability distribution of the target vehicle at the current moment; Predicting the driving trajectory of the target vehicle according to the target lane change probability distribution; The generating, based on the historical lane change probability distribution and the initial lane change probability distribution, a target lane change probability distribution of the target vehicle at the current moment, comprises: Obtaining a state transition probability matrix of the target vehicle; wherein the element in the i-th row and j-th column of the state transition probability matrix is ​​used to indicate the probability of the target vehicle switching from the i-th lane change behavior at the previous moment to the j-th lane change behavior at the current moment, and i and j are positive integers; Using the state transition probability matrix, transforming the historical lane change probability distribution to obtain the transition probability distribution of the target vehicle at the current moment; The target lane change probability distribution is generated based on the initial lane change probability distribution and the conversion probability distribution; wherein the target lane change probability distribution is used to indicate the target probability of the target vehicle performing different lane change behaviors at the current moment.

2. The method according to claim 1, characterized in that The generating the target lane change probability distribution based on the initial lane change probability distribution and the conversion probability distribution includes: Based on the initial lane change probability distribution, updating the transition probability distribution to obtain an updated transition probability distribution; Performing filtering on the updated conversion probability distribution to obtain a filtered probability distribution; Determining, from the filtered probability distribution, a probability to be adjusted that matches the lane change intention of the target vehicle at the current moment; Determining a probability adjustment factor of the probability to be adjusted according to a distance between the target vehicle and a target boundary line of a current driving lane; wherein the target boundary line matches the lane change intention; The probability to be adjusted in the filtered probability distribution is adjusted according to the probability adjustment factor to obtain the target lane change probability distribution.

3. The method according to claim 1, characterized in that The predicting the initial lane change probability distribution of the target vehicle at the current moment includes: Acquire driving status information of the target vehicle at the current moment and multiple historical moments before the current moment; Inputting the plurality of driving state information into a lane change prediction model to obtain a predicted lane change probability distribution of the target vehicle at the current moment output by the lane change prediction model; The predicted lane change probability distribution is corrected to obtain the initial lane change probability distribution.

4. The method according to claim 3, characterized in that The correcting the predicted lane change probability distribution to obtain the initial lane change probability distribution includes: Predicting a heading angle deviation between the heading angle of the target vehicle at the current moment and the heading angle of a target lane; wherein the target lane is the lane to which the target vehicle changes lanes when executing a corresponding lane changing behavior at the current moment; In response to the heading angle deviation being greater than a set deviation threshold, the first probability to be corrected in the predicted lane change probability distribution is suppressed to obtain the initial lane change probability distribution; wherein the first probability to be corrected is used to indicate the probability of the lane change behavior corresponding to the target vehicle changing lanes to the target lane.

5. The method according to claim 3, characterized in that: The correcting the predicted lane change probability distribution to obtain the initial lane change probability distribution includes: Obtaining the time of the target vehicle's most recent lane change; In response to the time difference between the lane change moment and the current moment being less than a set duration, the second probability to be corrected in the predicted lane change probability distribution is suppressed to obtain the initial lane change probability distribution; wherein the second probability to be corrected is used to indicate the probability of the target vehicle changing lanes to a lane that matches the lane change behavior of the most recent lane change.

6. The method according to claim 3, characterized in that The correcting the predicted lane change probability distribution to obtain the initial lane change probability distribution includes: Acquire the first driving information of the target vehicle at the current moment and the second driving information of the previous moment; According to the first driving information and the second driving information, obtaining a heading angle difference between the target vehicle at the current moment and the previous moment, and determining whether the target vehicle has a lane change intention at the current moment; If the target vehicle has no lane change intention and / or the heading angle difference is less than a set difference threshold, suppressing the third probability to be corrected in the predicted lane change probability distribution to obtain the initial lane change probability distribution; The third probability to be corrected is used to indicate the probability of a lane changing behavior of changing the lane in which the target vehicle was traveling at the previous moment.

7. The method according to claim 6, characterized in that The method further comprises: If the target vehicle has an intention to change lanes and / or the heading angle difference is greater than the set difference threshold, increasing the fourth probability to be corrected in the predicted lane change probability distribution to obtain the initial lane change probability distribution; The fourth probability to be corrected is used to indicate the probability of a lane changing behavior of changing the lane in which the target vehicle was traveling at the previous moment.

8. A trajectory prediction device, characterized in that: include: An acquisition module, used to acquire a historical lane change probability distribution of a target vehicle at a previous moment at a current moment; wherein the historical lane change probability distribution is used to indicate the probability of the target vehicle performing different lane change behaviors at the previous moment; A first prediction module, used to predict the initial lane change probability distribution of the target vehicle at the current moment; A generating module, configured to generate a target lane change probability distribution of the target vehicle at the current moment based on the historical lane change probability distribution and the initial lane change probability distribution; A second prediction module, used to predict the driving trajectory of the target vehicle according to the target lane change probability distribution; The generating, based on the historical lane change probability distribution and the initial lane change probability distribution, a target lane change probability distribution of the target vehicle at the current moment, comprises: Obtaining a state transition probability matrix of the target vehicle; wherein the element in the i-th row and j-th column of the state transition probability matrix is ​​used to indicate the probability of the target vehicle switching from the i-th lane change behavior at the previous moment to the j-th lane change behavior at the current moment, and i and j are positive integers; Using the state transition probability matrix, transforming the historical lane change probability distribution to obtain the transition probability distribution of the target vehicle at the current moment; The target lane change probability distribution is generated based on the initial lane change probability distribution and the conversion probability distribution; wherein the target lane change probability distribution is used to indicate the target probability of the target vehicle performing different lane change behaviors at the current moment.

9. A vehicle, characterized in that: include: A processor, and a memory communicatively connected to the processor; The memory stores computer-executable instructions; The processor executes the computer-executable instructions stored in the memory to implement the method according to any one of claims 1 to 7.

10. A chip, characterized in that: The chip comprises a processing circuit configured to perform the method according to any one of claims 1-7.

11. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer-executable instructions, which are used to implement the method according to any one of claims 1 to 7 when executed by a processor.

Images