Obstacle position prediction method and device and electronic equipment

Abstract

The invention provides an obstacle position prediction method and device, electronic equipment, a computer readable storage medium and a computer program product. The method comprises the following steps: fitting a trajectory curve according to actual positions corresponding to an obstacle at a plurality of historical moments, and predicting according to the trajectory curve to obtain a first predicted position of the obstacle at a future target moment; predicting a second prediction position of the obstacle at the target moment according to the kinematics information of the obstacle at the current moment; and performing weighted summation processing on the first prediction position and the second prediction position of the obstacle at the target moment to obtain a third prediction position of the obstacle at the target moment. According to the invention, accurate obstacle position prediction can be realized with low cost.

Description

Technical Field

[0001] The present application relates to vehicle technology, and in particular to an obstacle position prediction method, device, electronic device, computer-readable storage medium, and computer program product. Background Art

[0002] Autonomous driving technology is an advanced transportation technology that allows vehicles to complete driving tasks without human intervention. This technology integrates the achievements of many fields such as automatic control, architecture, artificial intelligence, and visual computing. It is the product of the high development of computer science, pattern recognition, and intelligent control technology.

[0003] Autonomous driving technology usually includes three key parts: perception, prediction, and planning control. For the prediction part, obstacle position prediction plays a vital supporting role. In the solutions provided by related technologies, obstacle trajectories are usually predicted based on big data models. However, this solution is heavily dependent on the quality of the training data set, and the scene classification method, the size of the scene library, and the comprehensiveness of the types will also directly affect the obstacle trajectory prediction effect. Summary of the invention

[0004] The present application provides an obstacle position prediction method, device, electronic device, computer-readable storage medium and computer program product, which can achieve accurate obstacle position prediction at a relatively low cost.

[0005] The technical solution of this application is implemented as follows:

[0006] The present application provides an obstacle position prediction method, comprising:

[0007] Fitting a trajectory curve according to the actual positions of the obstacle at multiple historical moments, and predicting a first predicted position of the obstacle at a future target moment according to the trajectory curve;

[0008] Predicting a second predicted position of the obstacle at the target time according to the kinematic information of the obstacle at the current time;

[0009] A weighted sum process is performed on the first predicted position and the second predicted position of the obstacle at the target time to obtain a third predicted position of the obstacle at the target time.

[0010] The present application provides an obstacle position prediction device, comprising:

[0011] A first prediction module is used to fit a trajectory curve according to the actual positions of the obstacle at multiple historical moments, and predict a first predicted position of the obstacle at a future target moment according to the trajectory curve;

[0012] A second prediction module, used for predicting a second predicted position of the obstacle at the target time according to the kinematic information of the obstacle at the current time;

[0013] The weighted summation module is used to perform weighted summation processing on the first predicted position and the second predicted position of the obstacle at the target time to obtain a third predicted position of the obstacle at the target time.

[0014] The present application provides an electronic device, including:

[0015] A memory for storing executable instructions;

[0016] The processor is used to implement the obstacle position prediction method provided in the present application when executing the executable instructions stored in the memory.

[0017] The present application provides a computer-readable storage medium storing executable instructions for causing a processor to execute and implement the obstacle position prediction method provided in the present application.

[0018] The present application provides a computer program product, which includes executable instructions for causing a processor to execute and implement the obstacle position prediction method provided in the present application.

[0019] This application has the following beneficial effects:

[0020] This application fits a trajectory curve according to the actual positions of the obstacle at multiple historical moments, and predicts the first predicted position of the obstacle at the future target moment according to the trajectory curve; predicts the second predicted position of the obstacle at the target moment according to the kinematic information of the obstacle at the current moment; and performs weighted summation processing on the first predicted position and the second predicted position of the obstacle at the target moment to obtain the third predicted position of the obstacle at the target moment. This application comprehensively considers the trajectory curve and kinematic information, which can make the final third predicted position of the obstacle at the target moment more accurate, which helps to improve the obstacle avoidance effect during the autonomous driving process. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For those skilled in the art, other drawings can be obtained based on these drawings without creative work.

[0022] Figure 1 It is a schematic diagram of the architecture of the obstacle position prediction system provided in the embodiment of the present application;

[0023] Figure 2 is a structural schematic diagram of a vehicle-mounted device provided in an embodiment of the present application;

[0024] Figure 3A It is a first flow chart of the obstacle position prediction method provided in an embodiment of the present application;

[0025] Figure 3B 2 is a schematic diagram of a second process of the obstacle position prediction method provided in an embodiment of the present application;

[0026] Figure 3C is a third flow chart of the obstacle position prediction method provided in an embodiment of the present application;

[0027] Figure 4 4 is a schematic diagram of a fourth process of the obstacle position prediction method provided in an embodiment of the present application;

[0028] Figure 5 is a schematic diagram of establishing a geodetic coordinate system provided in an embodiment of the present application;

[0029] Figure 6 is a schematic diagram of recording obstacle information provided by an embodiment of the present application;

[0030] Figure 7 is a schematic diagram of predicting a position by interpolation fitting provided in an embodiment of the present application;

[0031] Figure 8 is a schematic diagram of predicting position through a kinematic model provided in an embodiment of the present application;

[0032] Fig. 9 It is a schematic diagram of performing weighted sum processing on the first predicted position and the second predicted position to obtain the third predicted position provided by an embodiment of the present application;

[0033] Fig.10 is a first schematic diagram of an obstacle avoidance decision provided in an embodiment of the present application;

[0034] Fig.11 This is a second schematic diagram of the obstacle avoidance decision provided in an embodiment of the present application. DETAILED DESCRIPTION

[0035] In order to make the purpose, technical solutions and advantages of the present application clearer, the present application will be further described in detail below in conjunction with the accompanying drawings. The described embodiments should not be regarded as limiting the present application. All other embodiments obtained by ordinary technicians in the field without making creative work are within the scope of protection of this application.

[0036] In the following description, reference is made to "some embodiments", which describes a subset of all possible embodiments, but it is understood that "some embodiments" may be the same subset or different subsets of all possible embodiments, and may be combined with each other without conflict. In the following description, reference is made to the term "multiple" to mean at least two.

[0037] In the following description, the terms "first\second\third" involved are merely used to distinguish similar objects and do not represent a specific ordering of the objects. It can be understood that "first\second\third" can be interchanged with a specific order or sequence where permitted, so that the embodiments of the present application described herein can be implemented in an order other than that illustrated or described herein.

[0038] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as those commonly understood by those skilled in the art to which this application belongs. The terms used herein are only for the purpose of describing the embodiments of this application and are not intended to limit this application.

[0039] The embodiments of the present application provide an obstacle position prediction method, device, electronic device, computer-readable storage medium, and computer program product, which can achieve accurate obstacle position prediction at a relatively low cost. The following describes an exemplary application of the electronic device provided by the embodiments of the present application. The electronic device provided by the embodiments of the present application can be implemented as various types of terminal devices or as a server.

[0040] See also Figure 1 , Figure 1 It is a schematic diagram of the architecture of the obstacle position prediction system 100 provided in an embodiment of the present application. The vehicle-mounted device 400 is connected to the server 200 via the network 300, wherein the network 300 may be a wide area network or a local area network, or a combination of the two.

[0041] In some embodiments, taking the electronic device as a terminal device as an example, the obstacle position prediction method provided in the embodiment of the present application can be implemented by the terminal device. For example, the vehicle-mounted device 400 deployed in the vehicle obtains the obstacle ( Figure 1The vehicle-mounted device 400 includes a sensor (such as a human being) that is used as an example of an obstacle at multiple historical moments and the kinematic information of the obstacle at the current moment. The perception module may include a single sensor or multiple sensors (such as a camera + radar); the vehicle-mounted device 400 fits a trajectory curve according to the actual positions of the obstacle at multiple historical moments, and predicts the first predicted position of the obstacle at the future target moment according to the trajectory curve; the vehicle-mounted device 400 predicts the second predicted position of the obstacle at the target moment according to the kinematic information of the obstacle at the current moment; the vehicle-mounted device 400 performs a weighted summation process on the first predicted position and the second predicted position of the obstacle at the target moment to obtain the third predicted position of the obstacle at the target moment. Then, the vehicle-mounted device 400 may plan a control strategy (such as whether to brake the vehicle) according to the third predicted position of the obstacle at the target moment, and control the vehicle according to the control strategy. In the above method, the vehicle-mounted device 400 realizes obstacle position prediction locally without transmitting data to a remote server for processing, which can greatly reduce the time and delay of data transmission, thereby improving the real-time processing and improving the response speed; at the same time, it is not restricted by network conditions, which is especially important for scenarios where the network environment is unstable or the network cannot be connected, and can ensure the continuity and availability of obstacle position prediction.

[0042] In some embodiments, taking the electronic device as a server as an example, the obstacle position prediction method provided in the embodiment of the present application can be implemented by the server. For example, the vehicle-mounted device 400 sends the actual positions corresponding to the obstacle at multiple historical moments and the kinematic information of the obstacle at the current moment to the server 200; the server 200 fits the trajectory curve according to the actual positions corresponding to the obstacle at multiple historical moments, and predicts the first predicted position of the obstacle at the future target moment according to the trajectory curve; the server 200 predicts the second predicted position of the obstacle at the target moment according to the kinematic information of the obstacle at the current moment; the server 200 performs weighted summation processing on the first predicted position and the second predicted position of the obstacle at the target moment to obtain the third predicted position of the obstacle at the target moment. Then, the server 200 can plan and obtain the control strategy according to the third predicted position of the obstacle at the target moment, and send the control strategy to the vehicle-mounted device 400 for execution; or, the server 200 can send the third predicted position of the obstacle at the target moment to the vehicle-mounted device 400, so that the vehicle-mounted device 400 can plan and obtain the control strategy and execute the control strategy. In the above manner, the server 200 usually has powerful computing power and storage resources, and can efficiently predict the position of obstacles; the vehicle-mounted device 400 does not need to undertake complex computing tasks, which can reduce the resource consumption of the vehicle-mounted device 400.

[0043] In some embodiments, the electronic device can implement the obstacle position prediction method provided in the embodiments of the present application by running a computer program. For example, the computer program can be a native program or software module in the operating system; it can be a native application (APP, Application), that is, a program that needs to be installed in the operating system to run; it can also be a small program, that is, a program that can be run only by downloading it to a browser environment; it can also be a small program that can be embedded in any APP, and the small program can be controlled to run or closed by the user. In short, the above-mentioned computer program can be any form of application, module or plug-in.

[0044] Taking the electronic device provided in the embodiment of the present application as an example, which is a vehicle-mounted device, it can be understood that in the case where the electronic device is a server, Figure 2 Parts of the structure shown in FIG. 1 (such as a user interface, a presentation module, and an input processing module) may be omitted. Figure 2 , Figure 2 is a structural diagram of a vehicle-mounted device 400 provided in an embodiment of the present application, Figure 2 The vehicle-mounted device 400 shown includes: at least one processor 410, a memory 450, at least one network interface 420 and a user interface 430. The various components in the vehicle-mounted device 400 are coupled together via a bus system 440. It is understood that the bus system 440 is used to achieve connection and communication between these components. In addition to the data bus, the bus system 440 also includes a power bus, a control bus and a status signal bus. However, for the sake of clarity, the bus system 440 is not described in detail. Figure 2 Various buses are labeled as bus system 440 .

[0045] Processor 410 can be an integrated circuit chip with signal processing capabilities, such as a general-purpose processor, a digital signal processor (DSP), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc., where the general-purpose processor can be a microprocessor or any conventional processor, etc.

[0046] The user interface 430 includes one or more output devices 431 that enable presentation of media content, including one or more speakers and / or one or more visual display screens. The user interface 430 also includes one or more input devices 432, including user interface components that facilitate user input, such as microphones, touch screen displays, cameras, other input buttons and controls.

[0047] The memory 450 may be removable, non-removable, or a combination thereof. Exemplary hardware devices include solid-state memory, hard disk drives, optical disk drives, etc. The memory 450 may optionally include one or more storage devices that are physically remote from the processor 410.

[0048] The memory 450 includes a volatile memory or a non-volatile memory, and may also include both volatile and non-volatile memories. The non-volatile memory may be a read-only memory (ROM), and the volatile memory may be a random access memory (RAM). The memory 450 described in the embodiments of the present application is intended to include any suitable type of memory.

[0049] In some embodiments, memory 450 can store data to support various operations, examples of which include programs, modules, and data structures, or a subset or superset thereof, as exemplarily described below.

[0050] Operating system 451, including system programs for processing various basic system services and performing hardware-related tasks, such as a framework layer, a core library layer, a driver layer, etc., for implementing various basic services and processing hardware-based tasks;

[0051] A network communication module 452, used to reach other electronic devices via one or more (wired or wireless) network interfaces 420, exemplary network interfaces 420 include: Bluetooth, Wireless Compatibility Certification (WiFi), and Universal Serial Bus (USB), etc.;

[0052] a presentation module 453 for enabling presentation of information via one or more output devices 431 (e.g., display screen, speaker, etc.) associated with the user interface 430 (e.g., a user interface for operating peripherals and displaying content and information);

[0053] The input processing module 454 is used to detect one or more user inputs or interactions from one of the one or more input devices 432 and translate the detected inputs or interactions.

[0054] In some embodiments, the obstacle position prediction device provided in the embodiments of the present application can be implemented in software. Figure 2The obstacle position prediction device 455 stored in the memory 450 is shown, which can be software in the form of a program and a plug-in, etc., including the following software modules: a first prediction module 4551, a second prediction module 4552 and a weighted summation module 4553. These modules are logical, so they can be arbitrarily combined or further split according to the functions implemented. The functions of each module will be described below.

[0055] The obstacle position prediction method provided in the embodiment of the present application will be described in conjunction with the exemplary application and implementation of the electronic device provided in the embodiment of the present application.

[0056] See also Figure 3A , Figure 3A is a flow chart of the obstacle position prediction method provided in the embodiment of the present application, which will be combined with Figure 3A The steps shown are explained.

[0057] In step 101, a trajectory curve is fitted according to the actual positions of the obstacle at multiple historical moments, and a first predicted position of the obstacle at a future target moment is predicted based on the trajectory curve.

[0058] Here, the actual positions of the obstacles at multiple historical moments are obtained. Taking the vehicle scene as an example, the actual positions of the obstacles at multiple historical moments can be obtained through the vehicle's perception module, where the perception module can include a single sensor or multiple sensors (such as camera + radar).

[0059] Then, the trajectory curve is fitted according to the actual positions of the obstacles at multiple historical moments, that is, according to a series of discrete data points (the actual positions of the obstacles at multiple historical moments), a continuous curve that can best approximate these points is found through mathematical methods to describe or predict the potential relationship or trend between the data points. There is no limitation on the method of fitting the trajectory curve, for example, polynomial fitting can be used. The actual positions of the obstacles at multiple historical moments are described in the same coordinate system, so that the fitting effect can be guaranteed.

[0060] The fitted prediction curve describes the relationship between time and obstacle position. Therefore, the predicted position of the obstacle at the future target time can be predicted based on the trajectory curve (for ease of distinction, the predicted position obtained here is named the first predicted position).

[0061] It is worth noting that the historical moments involved in the embodiments of the present application refer to moments that have already occurred, and the most recent historical moment among multiple historical moments may be the current moment.

[0062] In step 102, a second predicted position of the obstacle at a target time is predicted based on the kinematic information of the obstacle at the current time.

[0063] Here, the kinematic information of the obstacle at the current moment is obtained. Taking the vehicle scene as an example, the kinematic information of the obstacle at the current moment can also be obtained through the perception module of the vehicle. It is worth noting that the kinematic information refers to the physical quantity that describes the motion state and motion characteristics of the obstacle, such as speed and acceleration.

[0064] Since the kinematic information of the obstacle at the current moment is known, the predicted position of the obstacle at the target moment can be predicted by the kinematic model (for the sake of distinction, the predicted position obtained here is named the second predicted position).

[0065] In step 103, a weighted sum process is performed on the first predicted position and the second predicted position of the obstacle at the target time to obtain a third predicted position of the obstacle at the target time.

[0066] Here, the first predicted position and the second predicted position of the obstacle at the target time are weighted and summed to obtain the third predicted position of the obstacle at the target time. It is worth noting that the sum of the weight corresponding to the first predicted position and the weight corresponding to the second predicted position is 1, and the weight corresponding to the first predicted position and the weight corresponding to the second predicted position are not specifically limited, and can be pre-set, for example, both are set to 0.5, or can be calculated with reference to historical data.

[0067] It is worth noting that the embodiments of the present application can be applied to any subject that needs to predict the position of obstacles, such as vehicles, robots, drones, etc.; the obstacles involved in the embodiments of the present application generally refer to any objects that may hinder the movement of the above-mentioned subjects.

[0068] In some embodiments, an obstacle position prediction method is applied to a vehicle; after performing weighted sum processing on a first predicted position and a second predicted position of the obstacle at a target time to obtain a third predicted position of the obstacle at the target time, the obstacle position prediction method further includes: controlling the vehicle to avoid the obstacle based on the third predicted position of the obstacle at the target time.

[0069] Here, taking the vehicle scenario as an example, after obtaining the third predicted position of the obstacle at the target time, the vehicle can be controlled to avoid the obstacle. For example, a control strategy is planned based on the actual position of the vehicle at the current time and the third predicted position of the obstacle at the target time, and the vehicle is controlled to avoid the obstacle according to the control strategy, wherein the obstacle avoidance strategy needs to be referred to during planning, and the referenced obstacle avoidance strategy can be pre-set. In the above manner, accurate obstacle avoidance can be achieved, improving the safety of autonomous driving.

[0070] like Figure 3A As shown, the embodiment of the present application fits a trajectory curve according to the actual positions corresponding to the obstacle at multiple historical moments, and predicts the first predicted position of the obstacle at the future target moment according to the trajectory curve; predicts the second predicted position of the obstacle at the target moment according to the kinematic information of the obstacle at the current moment; and performs weighted sum processing on the first predicted position and the second predicted position of the obstacle at the target moment to obtain the third predicted position of the obstacle at the target moment. The embodiment of the present application comprehensively considers the trajectory curve and kinematic information, so that the third predicted position of the obstacle at the target moment obtained can be more accurate, which helps to improve the obstacle avoidance effect during the autonomous driving process.

[0071] In some embodiments, see Figure 3B , Figure 3B is a flow chart of the obstacle position prediction method provided in the embodiment of the present application, based on Figure 3A , after step 102, step 201 may also be performed.

[0072] In step 201, weight correction processing is performed based on the actual position of the obstacle at the historical moment, the first predicted position and the second predicted position to obtain the historical weight corresponding to the historical moment; wherein the historical weight corresponding to the historical moment is used to perform weighted sum processing on the first predicted position and the second predicted position of the obstacle at the target moment.

[0073] For example, for any one of the multiple historical moments that appear in step 101 (referred to as historical moment T0 for ease of distinction), weight correction processing is performed based on the actual position of the obstacle at historical moment T0, the first predicted position, and the second predicted position to obtain the historical weight corresponding to historical moment T0, so that after weighted summing processing is performed on the first predicted position and the second predicted position of the obstacle at historical moment T0 based on the historical weight corresponding to historical moment T0, the actual position of the obstacle at historical moment T0 can be obtained.

[0074] The historical weights corresponding to the historical moment T0 include the weights corresponding to the first predicted position and the weights corresponding to the second predicted position. Based on this, the first predicted position and the second predicted position of the obstacle at the target moment are weighted and summed to obtain the third predicted position of the obstacle at the target moment.

[0075] In some embodiments, the obstacle position prediction method is applied to a vehicle; the obstacle position prediction method also includes: determining the activity level of the obstacle based on the historical weight corresponding to the historical moment; determining the obstacle avoidance strategy for the obstacle based on the activity level of the obstacle; and controlling the vehicle to avoid the obstacle based on the obstacle avoidance strategy for the obstacle and the third predicted position of the obstacle at the target moment.

[0076] Here, the activity level of the obstacle is determined based on the historical weights corresponding to the historical moments, wherein the activity level of the obstacle is positively correlated with the weight corresponding to the second predicted position in the historical weights corresponding to the historical moments, that is, the larger the weight corresponding to the second predicted position, the greater the impact of the movement of the obstacle on the actual position of the obstacle, and therefore the higher the activity level of the obstacle (the more active it is).

[0077] Then, the obstacle avoidance strategy for the obstacle is determined according to the activity level of the obstacle. For each activity level, a corresponding obstacle avoidance strategy can be pre-set. Taking the obstacle avoidance strategy including the advance braking distance as an example, the advance braking distance is positively correlated with the activity level of the obstacle; taking the obstacle avoidance strategy including the obstacle bypassing distance as an example, the obstacle bypassing distance is positively correlated with the activity level of the obstacle.

[0078] After determining the obstacle avoidance strategy for the obstacle, a control strategy is planned based on the obstacle avoidance strategy for the obstacle and the third predicted position of the obstacle at the target time, and the vehicle is controlled based on the control strategy to achieve obstacle avoidance processing for the obstacle.

[0079] In the above manner, a suitable obstacle avoidance strategy is determined according to the activity level of the obstacle to implement obstacle avoidance processing, which can avoid resource waste caused by excessive obstacle avoidance and safety problems caused by insufficient obstacle avoidance.

[0080] In some embodiments, the obstacle avoidance strategy for obstacles includes an advance braking distance; the above-mentioned obstacle avoidance strategy for obstacles and the third predicted position of the obstacle at the target time can be achieved in the following way: when the distance between the actual position of the vehicle at the current time and the third predicted position of the obstacle at the target time is less than the advance braking distance, the vehicle is braked.

[0081] Here, the planned control strategy may be: when the distance between the actual position of the vehicle at the current moment and the third predicted position of the obstacle at the target moment is less than the advance braking distance (the advance braking distance included in the determined obstacle avoidance strategy), the vehicle is braked. In this way, collision with obstacles is avoided by timely braking, thereby improving the safety of autonomous driving.

[0082] like Figure 3BAs shown, the embodiment of the present application corrects the weights with reference to the actual position of the obstacle at the historical moment, the first predicted position and the second predicted position, so that the weights used in the weighted summation of the first predicted position and the second predicted position of the obstacle at the target moment are more reasonable, thereby making the final third predicted position of the obstacle at the target moment more accurate.

[0083] In some embodiments, see Figure 3C , Figure 3C is a flow chart of the obstacle position prediction method provided in the embodiment of the present application, based on Figure 3B , after step 201, step 301 may also be executed.

[0084] In step 301, a weighted moving average process is performed on the historical weight sequence to obtain an average historical weight; wherein the historical weight sequence includes historical weights corresponding to multiple historical moments; wherein the average historical weight is used to perform a weighted sum process on the first predicted position and the second predicted position of the obstacle at the target moment.

[0085] Here, a historical weight sequence can be constructed based on the historical weights corresponding to multiple historical moments, and a weighted moving average process can be performed on the historical weight sequence to obtain an average historical weight. The principle of weighted moving average processing is that the more recent data is assigned a higher weight, so that the recent changes in the data can be better reflected. That is, the closer a historical moment is to the current moment, the greater the weight assigned to the historical weight corresponding to the historical moment in the weighted moving average processing process.

[0086] The average historical weight includes the weight corresponding to the first predicted position and the weight corresponding to the second predicted position. Based on this, the first predicted position and the second predicted position of the obstacle at the target time are weighted and summed to obtain the third predicted position of the obstacle at the target time.

[0087] In some embodiments, the length of the historical weight sequence is fixed. For example, as time goes by, historical weights corresponding to new historical moments are continuously added to the historical weight sequence, while historical weights corresponding to the oldest historical moments are continuously deleted to ensure that the length of the historical weight sequence is fixed. In this way, the historical weight sequence is updated in real time and the historical weights with the weakest reference are deleted in time, which can improve the overall referenceability of the historical weight sequence.

[0088] In some embodiments, the obstacle position prediction method is applied to a vehicle; the obstacle position prediction method also includes: determining the activity level of the obstacle based on at least one of the historical weight sequence and the average historical weight; determining an obstacle avoidance strategy for the obstacle based on the activity level of the obstacle; and controlling the vehicle to avoid the obstacle based on the obstacle avoidance strategy for the obstacle and the third predicted position of the obstacle at the target time.

[0089] Here, the activity level of the obstacle is determined based on at least one of the historical weight sequence and the average historical weight, wherein the activity level of the obstacle is positively correlated with the weight corresponding to the second predicted position in the historical weight sequence, and the activity level of the obstacle is positively correlated with the weight corresponding to the second predicted position in the average historical weight. Then, an obstacle avoidance strategy for the obstacle is determined based on the activity level of the obstacle, and the vehicle is controlled to avoid the obstacle based on the obstacle avoidance strategy for the obstacle and the third predicted position of the obstacle at the target time.

[0090] In the above method, the historical weight sequence includes historical weights corresponding to multiple historical moments, and the average historical weight is obtained by weighted moving average processing of the historical weight sequence. Therefore, the information referred to when determining the activity level of the obstacle is more comprehensive, which can improve the accuracy of the determined activity level of the obstacle; determining a suitable obstacle avoidance strategy according to the activity level of the obstacle and then implementing obstacle avoidance processing can avoid resource waste caused by excessive obstacle avoidance and safety problems caused by insufficient obstacle avoidance.

[0091] In some embodiments, the obstacle avoidance strategy for obstacles includes an advance braking distance; the above-mentioned obstacle avoidance strategy for obstacles and the third predicted position of the obstacle at the target time can be achieved in the following way: when the distance between the actual position of the vehicle at the current time and the third predicted position of the obstacle at the target time is less than the advance braking distance, the vehicle is braked.

[0092] Here, the planned control strategy may be: when the distance between the actual position of the vehicle at the current moment and the third predicted position of the obstacle at the target moment is less than the advance braking distance (the advance braking distance included in the determined obstacle avoidance strategy), the vehicle is braked. In this way, collision with obstacles is avoided by timely braking, thereby improving the safety of autonomous driving.

[0093] like Figure 3CAs shown, the embodiment of the present application performs weighted moving average processing on the historical weight sequence to obtain the average historical weight, and uses the average historical weight to perform weighted sum processing on the first predicted position and the second predicted position of the obstacle at the target time. In this way, more information can be referenced to determine the weights used for the weighted summation processing, thereby improving the accuracy of the third predicted position of the obstacle at the target time.

[0094] Below, an exemplary application of the embodiment of the present application in a practical application scenario will be described. As an example, the embodiment of the present application provides the following Figure 4 A flowchart of the obstacle position prediction method shown in FIG. Figure 4 , explained in step form.

[0095] Step 1) After the obstacle position prediction function is turned on, a geodetic coordinate system is established with the geometric center of the vehicle (i.e., the vehicle) at the starting time T0 as the origin, such as Figure 5 shown.

[0096] Step 2) During the process of the vehicle traveling from time T0 to time T1, the upstream perception module of the vehicle records the actual position and kinematic information of the obstacle according to the operation cycle ( Figure 4 The actual position and kinematic information of obstacles are collectively referred to as obstacle information). Figure 6 The actual positions of obstacles at different times are shown, namely O1, O2, etc. Among them, the duration from time T0 to time T1 is regarded as a prediction period, which can be set to 50ms to 100ms, for example. Among them, the number of obstacle points n collected from time T0 to time T1 can be 3 to 5, which of course does not constitute a limitation on the embodiments of the present application.

[0097] Step 3) The actual position of the vehicle, the kinematic information of the vehicle, the actual position of the obstacle, and the kinematic information of the obstacle are uniformly converted into the geodetic coordinate system established in step 1).

[0098] Step 4) At time T1, the trajectory curve is fitted by polynomial interpolation, and the position On+1 of the obstacle at time TΔ in the future is predicted based on the trajectory curve, where the time difference between time TΔ and time T1 is Δt, as Figure 7 As shown. Taking n=3 as an example, the following polynomial formula is shown:

[0099] x(t 0 )=a 0 +t 0 a 1 +t 0 2 a 2 +t 0 3 a3

[0100] y(t 0 )=b 0 +t 0 b 1 +t 0 2 b 2 +t 0 3 b 3

[0101] x(t 1 )=a 0 +t 1 a 1 +t 1 2 a 2 +t 1 3 a 3

[0102] y(t 1 )=b 0 +t 1 b 1 +t 1 2 b 2 +t 1 3 b 3

[0103] x(t 2 )=a 0 +t 2 a 1 +t 2 2 a 2 +t 2 3 a 3

[0104] y(t 2 )=b 0 +t 2 b 1 +t 2 2 b 2 +t 2 3 b 3

[0105] The above t 0 The time is the time when the obstacle is at position O1, x(t 0 ) represents the horizontal coordinate in position O1, y(t 0 ) represents the ordinate in position O1, and so on.0 、x(t 0 ), y(t 0 ), t 1 、x(t 1 ), y(t 1 ), t 2 、x(t 2 ) and y(t 2 ) are all known quantities, so the coefficient a in the polynomial formula can be solved 0 、a 1 、a 2 、a 3 、b 0 、b 1 、b 2 、b 3 Of course, the above polynomial formula is only an example and does not constitute a limitation on the embodiments of the present application.

[0106] Then, substitute the time TΔ into the above polynomial formula to obtain the first predicted position (x a ,y a ).

[0107] Step 5) Based on the kinematic information of the obstacle at the actual position On, a kinematic model is used to predict the second predicted position (x b ,y b ).like Figure 8 As shown in FIG. 1 , the kinematic information of the obstacle at the actual position On includes the vehicle speed v and the acceleration a. The vehicle speed v and the acceleration a are projected to the x and y directions respectively, and the second predicted position (x b ,y b ), the formula is as follows:

[0108]

[0109] Among them, x 0 Represents the horizontal coordinate of the actual position of the obstacle at time T1, y 0 Represents the vertical coordinate of the actual position of the obstacle at time T1.

[0110] Step 6) The first predicted position (x a ,y a ) and the second predicted position (x b ,y b ) is weighted and summed to obtain the third predicted position (x c ,y c ),like Fig. 9 The formula for weighted summation is as follows:

[0111] x c =x a (1-w)+x b w

[0112] y c =y a (1-w)+y b w

[0113] Among them, (1-w) represents the weight corresponding to the first prediction position, and w represents the weight corresponding to the second prediction position.

[0114] Step 7) When reaching the time TΔ, the weight w is modified. Specifically, the actual position of the obstacle at the time TΔ is collected. And the weight w (corresponding to the historical weight above) is corrected by calculating the Euclidean distance. The formula is as follows:

[0115]

[0116] Step 8) In the process of repeatedly executing steps 4) to 7), the historical weight sequence w can be obtained after multiple prediction cycles. 0 ,…w n , where 0 to n represent the time sequence, w 0 represents the oldest historical weight, w n represents the latest historical weight, w 0 ,…w n The number of is consistent with the number of obstacle points n, and w 0 ,…w n The initial value of is set to 0.5. In addition, each time the historical weight w is obtained through step 7), it is added to the historical weight sequence as the latest historical weight, and the oldest historical weight in the historical weight sequence is deleted to ensure that the length of the historical weight sequence is fixed.

[0117] Perform weighted moving average processing on the obtained historical weight sequence to obtain the average historical weight. The formula is as follows:

[0118] w t =a n w n +a n-1 w n-1 +…+a 0 w 0

[0119] Among them, a n +a n-1 +…+a 0 =1, and in descending order.

[0120] The obtained average historical weight is used to implement the weighted summation process in the above step 6).

[0121] Step 9) Determine the activity level of the obstacle. The larger the weight corresponding to the second predicted position, the greater the prediction weight of the kinematic model, and the higher the activity level of the obstacle. Therefore, the number N of historical weights with a value greater than 0.5 in the historical weight sequence and the average historical weight can be given to the downstream obstacle avoidance decision module, so that the obstacle avoidance decision module determines the activity level of the obstacle and determines the obstacle avoidance strategy for the obstacle according to the activity level of the obstacle.

[0122] The decision-making method of the obstacle avoidance strategy is shown in Table 1.

[0123] Table 1 Obstacle avoidance strategy decision table

[0124]

[0125]

[0126] It is worth noting that the embodiments of the present application do not limit the specific form of the obstacle avoidance strategy, including but not limited to braking, detouring, parking, etc. Taking the obstacle avoidance strategy as an example, in the obstacle avoidance decision process, for the same third predicted position, different advance braking distances can be adopted according to different activity levels. The higher the activity level, the greater the advance braking distance. For example, under the premise that the third predicted position is the same, if the obstacle is a basketball (such as Fig.10 As shown in the figure, since the basketball does not have autonomous movement ability, its activity level is usually low, so the determined advance braking distance is small, that is, the vehicle will brake only when it is close to the basketball, which can ensure safety and avoid hindering the normal driving of the vehicle as much as possible; if the obstacle is a puppy (such as Fig.11 As shown in the figure, since the puppy has the ability to move independently and its activity level is usually high, the determined advance braking distance is large, that is, the vehicle brakes when it is far away from the puppy, which can improve safety as much as possible and avoid the occurrence of safety accidents.

[0127] The embodiments of the present application have at least the following technical effects:

[0128] 1) Comprehensively considering the trajectory curve and kinematic information to predict the obstacle position can improve the position prediction accuracy while reducing the implementation cost.

[0129] 2) The obstacle position prediction accuracy is further improved by continuously iterating the historical weight sequence online.

[0130] 3) The activity level of the obstacle is determined based on the historical weight sequence and the average historical weight, so as to facilitate the determination of the appropriate obstacle avoidance strategy according to the activity level of the obstacle during the autonomous driving process, taking into account both driving efficiency and safety.

[0131] The following is a description of an exemplary structure of the obstacle position prediction device 455 provided in the embodiment of the present application implemented as a software module. In some embodiments, Figure 2 As shown, the software modules stored in the obstacle position prediction device 455 of the memory 450 may include: a first prediction module 4551, which is used to fit a trajectory curve according to the actual positions of the obstacle corresponding to multiple historical moments, and predict the first predicted position of the obstacle at the future target moment according to the trajectory curve; a second prediction module 4552, which is used to predict the second predicted position of the obstacle at the target moment according to the kinematic information of the obstacle at the current moment; and a weighted summation module 4553, which is used to perform weighted summation processing on the first predicted position and the second predicted position of the obstacle at the target moment to obtain the third predicted position of the obstacle at the target moment.

[0132] In some embodiments, the obstacle position prediction device 455 also includes a weight correction module, which is used to: perform weight correction processing based on the actual position of the obstacle at the historical moment, the first predicted position and the second predicted position to obtain the historical weight corresponding to the historical moment; wherein the historical weight corresponding to the historical moment is used to perform weighted summation processing on the first predicted position and the second predicted position of the obstacle at the target moment.

[0133] In some embodiments, the weight correction module is also used to: perform weighted moving average processing on the historical weight sequence to obtain an average historical weight; wherein the historical weight sequence includes historical weights corresponding to multiple historical moments; wherein the average historical weight is used to perform weighted sum processing on the first predicted position and the second predicted position of the obstacle at the target moment.

[0134] In some embodiments, the obstacle position prediction device 455 is applied to a vehicle; the obstacle position prediction device 455 also includes an activity level determination module, which is used to: determine the activity level of the obstacle based on at least one of the historical weight sequence and the average historical weight; the obstacle position prediction device 455 also includes an obstacle avoidance strategy determination module, which is used to: determine the obstacle avoidance strategy for the obstacle based on the activity level of the obstacle; the obstacle position prediction device 455 also includes an obstacle avoidance module, which is used to: control the vehicle to avoid the obstacle based on the obstacle avoidance strategy for the obstacle and the third predicted position of the obstacle at the target time.

[0135] In some embodiments, the obstacle avoidance strategy for obstacles includes an advance braking distance; the obstacle avoidance module is also used to: brake the vehicle when the distance between the actual position of the vehicle at the current moment and the third predicted position of the obstacle at the target moment is less than the advance braking distance.

[0136] In some embodiments, the obstacle position prediction device 455 is applied to a vehicle; the obstacle position prediction device 455 also includes an obstacle avoidance module, which is used to control the vehicle to avoid the obstacle according to the third predicted position of the obstacle at the target time.

[0137] The embodiment of the present application provides a computer program product or a computer program, which includes executable instructions stored in a computer-readable storage medium. The processor of the electronic device reads the executable instructions from the computer-readable storage medium, and the processor executes the executable instructions, so that the electronic device implements the obstacle position prediction method described in the embodiment of the present application.

[0138] An embodiment of the present application provides a computer-readable storage medium storing executable instructions, wherein executable instructions are stored. When the executable instructions are executed by a processor, the processor will implement the obstacle position prediction method provided by the embodiment of the present application.

[0139] In some embodiments, the computer-readable storage medium may be a memory such as FRAM, ROM, PROM, EPROM, EEPROM, flash memory, magnetic surface storage, optical disk, or CD-ROM; or it may be various devices including one or any combination of the above memories.

[0140] In some embodiments, executable instructions may be in the form of a program, software, software module, script or code, written in any form of programming language (including compiled or interpreted languages, or declarative or procedural languages), and may be deployed in any form, including as a stand-alone program or as a module, component, subroutine or other unit suitable for use in a computing environment.

[0141] As an example, executable instructions may, but need not, correspond to a file in a file system, may be stored as part of a file that stores other programs or data, such as in one or more scripts in a HyperText Markup Language (HTML) document, in a single file dedicated to the program in question, or in multiple coordinated files (e.g., files storing one or more modules, subroutines, or code portions).

[0142] As an example, the executable instructions may be deployed to be executed on one electronic device, or on multiple electronic devices located at one site, or on multiple electronic devices distributed at multiple sites and interconnected by a communication network.

[0143] The above are only embodiments of the present application and are not intended to limit the protection scope of the present application. Any modifications, equivalent substitutions and improvements made within the spirit and scope of the present application are included in the protection scope of the present application.

Claims

1. A method for predicting obstacle positions, characterized in that: include: Fitting a trajectory curve according to the actual positions of the obstacle at multiple historical moments, and predicting a first predicted position of the obstacle at a future target moment according to the trajectory curve; Predicting a second predicted position of the obstacle at the target time according to the kinematic information of the obstacle at the current time; A weighted sum process is performed on the first predicted position and the second predicted position of the obstacle at the target time to obtain a third predicted position of the obstacle at the target time.

2. The method according to claim 1, characterized in that Before performing weighted sum processing on the first predicted position and the second predicted position of the obstacle at the target time, the method further includes: Perform weight correction processing according to the actual position of the obstacle at the historical moment, the first predicted position, and the second predicted position to obtain a historical weight corresponding to the historical moment; The historical weight corresponding to the historical moment is used to perform weighted sum processing on the first predicted position and the second predicted position of the obstacle at the target moment.

3. The method according to claim 2, characterized in that After performing weight correction processing according to the actual position of the obstacle at the historical moment, the first predicted position and the second predicted position to obtain the historical weight corresponding to the historical moment, the method further includes: Performing weighted moving average processing on the historical weight sequence to obtain an average historical weight; wherein the historical weight sequence includes historical weights corresponding to multiple historical moments respectively; The average historical weight is used to perform weighted sum processing on the first predicted position and the second predicted position of the obstacle at the target moment.

4. The method according to claim 3, characterized in that The method is applied to a vehicle; the method further comprises: determining an activity level of the obstacle according to at least one of the historical weight sequence and the average historical weight; determining an obstacle avoidance strategy for the obstacle according to the activity level of the obstacle; The vehicle is controlled to perform obstacle avoidance processing on the obstacle according to the obstacle avoidance strategy for the obstacle and the third predicted position of the obstacle at the target time.

5. The method according to claim 4, characterized in that The obstacle avoidance strategy for the obstacle includes an advance braking distance; and controlling the vehicle to perform obstacle avoidance processing on the obstacle according to the obstacle avoidance strategy for the obstacle and the third predicted position of the obstacle at the target time includes: When the distance between the actual position of the vehicle at the current moment and the third predicted position of the obstacle at the target moment is less than the advance braking distance, braking processing is performed on the vehicle.

6. The method according to claim 1, characterized in that The method is applied to a vehicle; after performing weighted sum processing on the first predicted position and the second predicted position of the obstacle at the target time to obtain the third predicted position of the obstacle at the target time, the method further includes: The vehicle is controlled to perform obstacle avoidance processing on the obstacle according to the third predicted position of the obstacle at the target time.

7. An obstacle position prediction device, characterized in that: include: A first prediction module is used to fit a trajectory curve according to the actual positions of the obstacle at multiple historical moments, and predict a first predicted position of the obstacle at a future target moment according to the trajectory curve; A second prediction module, used for predicting a second predicted position of the obstacle at the target time according to the kinematic information of the obstacle at the current time; The weighted summation module is used to perform weighted summation processing on the first predicted position and the second predicted position of the obstacle at the target time to obtain a third predicted position of the obstacle at the target time.

8. An electronic device, characterized in that: include: A memory for storing executable instructions; A processor, configured to implement the method according to any one of claims 1 to 6 when executing the executable instructions stored in the memory.

9. A computer-readable storage medium, characterized in that: Executable instructions are stored, and when executed by a processor, the method described in any one of claims 1 to 6 is implemented.

10. A computer program product, characterized in that The method comprises executable instructions for implementing the method according to any one of claims 1 to 6 when executed by a processor.

Images