Target trajectory prediction method, device, electronic device and storage medium

Abstract

The present application relates to a target trajectory prediction method, apparatus, electronic device, and storage medium. By collecting the real trajectory data of a target at a preset frequency, obtaining n consecutive sampling values recorded from the first historical moment to the current moment, a first real trajectory data group is obtained. A first predicted trajectory data group for predicting the trajectory of the target from the first historical moment to the current moment is obtained, and the first predicted trajectory data group includes n consecutive predicted values. Comparing the first predicted trajectory data group with the first real trajectory data group to obtain a first error result between the first predicted trajectory data group and the first real trajectory data group. Compensating the second predicted trajectory data group according to the first error result to obtain the predicted trajectory of the target from the second historical moment to the next moment, and the second predicted trajectory data group includes n consecutive predicted values, which solves the problem of low accuracy of the predicted trajectory of the target and improves the accuracy of the predicted trajectory of the target.

Description

Technical Field

[0001] This application relates to the field of autonomous driving technology, and particularly to a method, device, electronic device, and storage medium for predicting an object trajectory. Background Art

[0002] Vehicle trajectory prediction is one of the key technologies in autonomous driving. The existing vehicle trajectory prediction methods mainly include the following categories: one is the method based on vehicle dynamics and kinematic models; the second is to first predict the driving intention of the vehicle and then generate a predicted trajectory. For the method of generating a trajectory according to the driving intention, after generating candidate predicted trajectories through algorithms such as RRT, the parameters of the trajectory evaluation function are learned by the maximum likelihood method, so as to select the trajectory with the minimum cost. There are also models such as Gaussian processes, and the parameters of the model are learned through training data to generate predicted trajectories. Among them, the methods for predicting driving intention include using cost functions, machine learning, hidden Markov models, Bayesian networks, etc.; the third is the method based on deep learning to directly generate trajectories, including interactive algorithms based on LSTM, interactive algorithms based on Graph / Attention, etc.

[0003] However, the above three types of vehicle trajectory prediction methods all rely on the accuracy of offline data and the coverage rate of training scenarios. For example, for the method based on vehicle dynamics and kinematic models, offline vehicle parameters and vehicle state parameters at the current time are required. The predicted trajectory deduced from the parameter evolution may exceed the real trajectory of a human driver in complex working conditions; for example, for the trajectory prediction method based on driver intention, when calculating the cost function cost, it is impossible to effectively measure the real intention of the driver in complex scenarios. For example, when the target vehicle is on the rightmost ramp where it can only merge into the main road, and the target vehicle is close to the right lane line of its own vehicle. It is difficult for the method based on driver intention to effectively evaluate the real intention of the driver, thereby affecting the trajectory predicted by the algorithm; for example, for the method of predicting the trajectory based on a deep learning model, the offline training data needs to cover various scenarios. When the training scenario data does not include special complex scenarios, it may cause the model prediction to not converge and the real trajectory of the vehicle driven by the driver cannot be obtained. Therefore, in the actual application process of the above three types of vehicle prediction trajectory methods, due to factors such as the accuracy of offline data and the coverage rate of training scenarios, there is a significant gap between the predicted trajectory and the actual trajectory.

[0004] Aiming at the problem of low accuracy of the predicted target trajectory in the related technology, no effective solution has been proposed yet. Summary of the Invention

[0005] In this embodiment, a method, device, electronic device, and storage medium for predicting an object trajectory are provided to solve the problem of low accuracy of the predicted target trajectory in the related technology.

[0006] In a first aspect, in the present embodiment, a target trajectory prediction method is provided. The real trajectory data of the target is collected at a preset frequency, and n consecutive sampling values recorded from the first historical moment to the current moment are obtained to obtain a first real trajectory data group, and a first predicted trajectory data group for predicting the trajectory of the target from the first historical moment to the current moment is obtained, where the first predicted trajectory data group includes n consecutive predicted values; the first predicted trajectory data group and the first real trajectory data group are compared to obtain a first error result between the first predicted trajectory data group and the first real trajectory data group; the second predicted trajectory data group is compensated according to the first error result to obtain the predicted trajectory of the target from the second historical moment to the next moment, where the second predicted trajectory data group includes n consecutive predicted values.

[0007] In some embodiments, n error values are obtained according to the sampling values in the first real trajectory data group and the corresponding predicted values in the first predicted trajectory data group; each error value is judged for validity, and the error values judged to be invalid are corrected.

[0008] In some embodiments, a first threshold is set, the n error values are respectively compared with the first threshold, and the error values greater than the first threshold are determined to be invalid error values. In some embodiments, two error values adjacent to the invalid error value are obtained, and the average value of the two adjacent error values is used to replace the invalid error value.

[0009] In some embodiments, the n error values are added to the n consecutive predicted values in the second predicted trajectory data group correspondingly.

[0010] In some embodiments, a second error result is obtained, where the second error result is obtained by comparing the second predicted trajectory data group and a second real trajectory data group, the second real trajectory data group includes n consecutive sampling values of the target from the second historical moment to the next moment, and the second error result includes n error values; the average values of the first error result and the second error result are calculated respectively, the average values of the first error result and the second error result are compared, and whether to use the first error result to compensate the second predicted trajectory data group is determined according to the obtained comparison result.

[0011] In some embodiments, the generation time interval between the first predicted trajectory data group and the second predicted trajectory data group is less than or equal to the time interval from the current moment to the next moment.

[0012] Second aspect, in this embodiment, a target trajectory prediction device is provided, including:

[0013] An acquisition module, configured to collect real trajectory data of a target at a preset frequency, obtain n consecutive sampling values recorded from a first historical moment to the current moment, to obtain a first real trajectory data group, and obtain a first predicted trajectory data group for predicting the trajectory of the target from the first historical moment to the current moment, where the first predicted trajectory data group includes n consecutive predicted values;

[0014] A comparison module, configured to compare the first predicted trajectory data group with the first real trajectory data group to obtain a first error result between the first predicted trajectory data group and the first real trajectory data group;

[0015] A compensation module, configured to compensate a second predicted trajectory data group according to the first error result to obtain a predicted trajectory of the target from a second historical moment to the next moment, where the second predicted trajectory data group includes n consecutive predicted values.

[0016] Third aspect, in this embodiment, an electronic device is provided, including a memory, a processor, and a computer program stored on the memory and executable on the processor, where when the processor executes the computer program, the target trajectory prediction method described in the first aspect above is implemented.

[0017] Fourth aspect, in this embodiment, a storage medium is provided, on which a computer program is stored, and when the program is executed by a processor, the target trajectory prediction method described in the first aspect above is implemented.

[0018] Compared with the related art, in the target trajectory prediction method, device, electronic device, and storage medium provided in this embodiment, real trajectory data of a target is collected at a preset frequency, n consecutive sampling values recorded from a first historical moment to the current moment are obtained to obtain a first real trajectory data group, and a first predicted trajectory data group for predicting the trajectory of the target from the first historical moment to the current moment is obtained, where the first predicted trajectory data group includes n consecutive predicted values; the first predicted trajectory data group is compared with the first real trajectory data group to obtain a first error result between the first predicted trajectory data group and the first real trajectory data group; the second predicted trajectory data group is compensated according to the first error result to obtain a predicted trajectory of the target from a second historical moment to the next moment, where the second predicted trajectory data group includes n consecutive predicted values, which solves the problem of low accuracy of the predicted trajectory of the target in the related art and improves the accuracy of the predicted trajectory of the target.

[0019] Details of one or more embodiments of the present application are set forth in the following drawings and description to make other features, objects, and advantages of the present application more concise and understandable. Description of the Drawings

[0020] The drawings described herein are used to provide a further understanding of the present application and form a part of the present application. The illustrative embodiments and descriptions thereof of the present application are used to explain the present application and do not constitute an improper limitation of the present application. In the drawings:

[0021] Figure 1 is a block diagram of the terminal hardware structure of the target trajectory prediction method according to an embodiment of the present application;

[0022] Figure 2 is a flowchart of the target trajectory prediction method according to an embodiment of the present application;

[0023] Figure 3 is a schematic diagram of the vehicle trajectory data acquisition method according to an embodiment of the present application;

[0024] Figure 4 is a flowchart of the vehicle trajectory prediction method according to an embodiment of the present application;

[0025] Figure 5 is a block diagram of the structure of the target trajectory prediction device according to an embodiment of the present application. Detailed Embodiments

[0026] To understand the purpose, technical solution, and advantages of the present application more clearly, the present application will be described and explained below with reference to the drawings and embodiments.

[0027] Unless otherwise defined, technical terms or scientific terms involved in this application shall have the general meanings understood by those with ordinary skills in the technical field to which this application belongs. In this application, words such as "a", "an", "one kind", "the", "these", etc. do not indicate a limitation in quantity, and they can be singular or plural. The terms "include", "comprise", "have" and any variations thereof involved in this application are intended to cover non-exclusive inclusion; for example, a process, method, system, product or device including a series of steps or modules (units) is not limited to the listed steps or modules (units), but may include unlisted steps or modules (units), or may include other steps or modules (units) inherent in these processes, methods, products or devices. The "connection", "coupling", "connected" and other similar words involved in this application are not limited to physical or mechanical connections, but may include electrical connections, whether direct or indirect. The "plurality" involved in this application means two or more. "And / or" describes the association relationship of associated objects and indicates that three relationships can exist. For example, "A and / or B" can represent: A exists alone, A and B exist simultaneously, and B exists alone. Usually, the character " / " indicates that the objects before and after are in an "or" relationship. The terms "first", "second", "third", etc. involved in this application only distinguish similar objects and do not represent a specific order for the objects.

[0028] The method embodiments provided in this embodiment can be executed on a terminal, a computer, or a similar computing device. For example, when running on a terminal, Figure 1 is a hardware structural block diagram of a terminal for the target trajectory prediction method according to an embodiment of this application. As Figure 1 shown, the terminal may include one or more ( Figure 1 only one is shown in the figure) processors and a memory for storing data. Among them, the processor may include, but is not limited to, a processing device such as a microprocessor MCU or a programmable logic device FPGA. The above terminal may also include a transmission device for communication functions and input / output devices. Those of ordinary skill in the art can understand that Figure 1 the structure shown is only schematic and does not limit the structure of the above terminal. For example, the terminal may also include more or fewer components than Figure 1 shown in the figure, or have a different configuration from Figure 1 shown.

[0029] The memory can be used to store computer programs, for example, software programs and modules of application software, such as the computer program corresponding to the target trajectory prediction method in this embodiment. The processor executes various functional applications and data processing by running the computer program stored in the memory, that is, the above-mentioned method is implemented. The memory may include high-speed random access memory, and may also include non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memories. In some instances, the memory may further include a memory remotely disposed relative to the processor, and these remote memories can be connected to the terminal through a network. Examples of the above-mentioned network include but are not limited to the Internet, enterprise intranet, local area network, mobile communication network, and combinations thereof.

[0030] The transmission device is used to receive or send data via a network. The above-mentioned network includes the wireless network provided by the communication provider of the terminal. In one instance, the transmission device includes a network adapter (Network Interface Controller, abbreviated as NIC), which can be connected to other network devices through a base station and thus communicate with the Internet. In one instance, the transmission device can be a radio frequency (RF) module, which is used to communicate with the Internet wirelessly.

[0031] In one embodiment, a target trajectory prediction method is provided. Taking the terminal in which this method is applied as an example for illustration, Figure 1 is the flowchart of the target trajectory prediction method. As shown in Figure 2 the following steps are included in this process: Figure 2 As shown in

[0032] Step S101: Collect the real trajectory data of the target at a preset frequency, obtain n consecutive sampling values recorded from the first historical moment to the current moment, obtain the first real trajectory data group, and obtain the first predicted trajectory data group for predicting the trajectory of the target from the first historical moment to the current moment, where the first predicted trajectory data group contains n consecutive predicted values.

[0033] During the movement of the target, on the one hand, the real trajectory data of the target is collected at a preset frequency through a sensor, and the data storage method of last-in-first-out is adopted. Every first period, n consecutive sampling values are recorded as the first real trajectory data group. On the other hand, a predicted trajectory data group is generated every second period, and each predicted trajectory data group contains n consecutive predicted values.

[0034] Among them, the target can be a means of transportation such as a car or a motorcycle, or an intelligent device such as a robot or an unmanned aerial vehicle. The n consecutive sampled values and the n consecutive predicted values correspond one by one. Both the sampled values and the predicted values include the spatial coordinates of the target, and the spatial coordinates include the horizontal position and / or the vertical position.

[0035] Step S102: Compare the first predicted trajectory data group with the first true trajectory data group to obtain a first error result between the first predicted trajectory data group and the first true trajectory data group.

[0036] Subtract the n consecutive sampled values and the n consecutive predicted values correspondingly to obtain n error values. The n error values can be directly used as the first error result, or after processing the n error values, the processed n error values can be used as the first error result.

[0037] Step S103: Compensate the second predicted trajectory data group according to the first error result to obtain the predicted trajectory of the target from the second historical moment to the next moment. Among them, the second predicted trajectory data group includes n consecutive predicted values.

[0038] Add the n error values or the processed n error values to the n predicted values in the second predicted trajectory data group correspondingly to achieve the compensation of the second predicted trajectory data group.

[0039] The trajectory prediction method in the related art directly gives the predicted trajectory based on offline data. However, when predicting the trajectory in real time, the error between the predicted trajectory and the true trajectory is ignored. In this embodiment, through the above steps S101 to S103, the error result between the true trajectory data group recorded from the first historical moment to the current moment and the corresponding predicted trajectory data group is calculated, and the calculated error result is compensated into the subsequent predicted trajectory data group. Based on the compensated predicted trajectory data group, the prediction of the target trajectory in the future time is realized, achieving the purpose of real-time online correction of the target predicted trajectory, solving the problem of low accuracy of the target predicted trajectory in the related art, and improving the accuracy of the target predicted trajectory. At the same time, the algorithm complexity of the above target trajectory prediction method is low, the CPU / memory resources are less occupied during the operation, the application scenario is wide, the matching degree is high, and it is applicable to the trajectory correction of various prediction models or algorithms.

[0040] In some of the embodiments, comparing the first predicted trajectory data group with the first true trajectory data group to obtain a first error result between the first predicted trajectory data group and the first true trajectory data group includes: obtaining n error values according to the sampled values in the first true trajectory data group and the corresponding predicted values in the first predicted trajectory data group; making a validity judgment on each error value, and correcting the error values judged to be invalid.

[0041] The terminal sensor samples the real trajectory data of the target at a preset frequency. During the sampling process, the error of the sensor will affect the accuracy of the sampling value, thereby affecting the magnitude of the error value. Therefore, it is necessary to judge the validity of n error values and correct the invalid error values according to the judgment results. In some embodiments, judging the validity of each error value includes: setting a first threshold, comparing the n error values with the first threshold respectively, and determining the error values greater than the first threshold as invalid error values.

[0042] Calculate the standard deviation of the above n error values, and set the first threshold to k times the within-group standard deviation, where the value of k is set according to the target trajectory prediction requirement. Compare the above error values with the set first threshold respectively, determine the error values greater than the first threshold as invalid error values that need to be corrected, and determine the error values less than or equal to the first threshold as valid error values that do not need to be corrected.

[0043] In some embodiments, correcting the error values determined to be invalid includes: obtaining two error values adjacent to the invalid error value, and replacing the invalid error value with the average value of the two adjacent error values.

[0044] For the invalid error values, they are corrected by the upsampling and downsampling filling method, that is, selecting the average value of the two adjacent error values of the error value to replace the error value.

[0045] In some embodiments, the first error result includes n error values. Compensating the second predicted trajectory data group according to the first error result includes: adding the n error values to the n consecutive predicted values in the second predicted trajectory data group correspondingly.

[0046] The second predicted trajectory data group is the predicted trajectory of the target from the second historical moment to the next moment calculated by the algorithm. Add the above error values to the corresponding consecutive predicted values in the second predicted trajectory data group respectively, reduce the error between the trajectory predicted by the real-time algorithm and the real trajectory to correct the future predicted trajectory, and improve the accuracy of the trajectory prediction.

[0047] In some embodiments, the error result includes n error values. Compensating the second predicted trajectory data group according to the error result further includes:

[0048] Obtain a second error result, where the second error result is obtained by comparing a second predicted trajectory data set and a second true trajectory data set. The second true trajectory data set includes n consecutive sampling values of the target from the second historical moment to the next moment. The second error result contains n error values; calculate the average values of the first error result and the second error result respectively, compare the average values of the first error result and the second error result, and determine whether to use the first error result to compensate the second predicted trajectory data set according to the obtained comparison result.

[0049] Calculate the mean values of the first error result and the second error result respectively, and compare the relationship between the two mean values. The method for comparing the mean values is not limited, including but not limited to determining whether the above mean values are within a unified threshold range. If the above mean values are within the preset threshold range, compensation is performed; determine the size relationship between the difference between the above mean values and a certain fixed threshold point. If the difference between the above mean values is less than the fixed threshold point, compensation is performed.

[0050] In some embodiments, the generation time interval between the first predicted trajectory data set and the second predicted trajectory data set is less than or equal to the time interval from the current moment to the next moment.

[0051] Set the generation time interval between the first predicted trajectory data set and the second predicted trajectory data set to ensure that the future predicted trajectory can be corrected in real time according to the first error result.

[0052] In one embodiment, taking the prediction of vehicle trajectory as an example, Figure 3 A schematic diagram of a method for obtaining vehicle trajectory data is given. As Figure 3 shown, during the vehicle movement process, on the one hand, the true trajectory data of the vehicle is collected by the sensor at a preset frequency. Assuming that the collection starts from the t moment, the data storage method of first in last out is adopted, and n consecutive sampling values are recorded every time interval τ as the true trajectory data set. For example, when the sensor starts to collect the true trajectory data of the vehicle from the t time, when the time reaches t + l, n points are filled, where l = k × τ. The first true trajectory data set collected within the time period from t to (t + l) is numbered as 1-(1), 1-(2), 1-(3), ……, 1-(n - 1), and into 1-(n); after the time interval τ, the sensor starts to collect the second true trajectory data set from the t + τ time. When the time reaches t + τ + l, n points are filled. The second true trajectory data set collected within the time period from t + τ to (t + τ + l) is numbered as 2-(1), 2-(2), 2-(3), ……, 2-(n - 1), and into 2-(n); after the time interval τ, the sensor starts to collect the next set of true trajectory data from the t + 2τ time, and so on. After the first true trajectory data set is filled with n points, the comparison between the predicted trajectory and the true trajectory can start.

[0053] On the other hand, starting from time t, a set of initial predicted trajectory data groups is generated every time interval τ. Each set of predicted trajectory data groups contains n consecutive predicted values. At time t, the first predicted trajectory data group is obtained, which is used to predict the vehicle trajectory in the future time period from t to (t + l). The predicted trajectory data numbers are recorded as Tra1-(1), Tra1-(2), Tra1-(3), ……, Tra1-(n - 1), Tra1-(n); at time t + τ, the second predicted trajectory data group is obtained, which is used to predict the vehicle trajectory in the future time period from t + τ to (t + τ + l). The predicted trajectory data numbers are recorded as Tra2-(1), Tra2-(2), Tra2-(3), ……, Tra2-(n - 1), Tra2-(n); at time t + 3τ, the next set of predicted trajectory data groups is obtained, and so on. Among them, the predicted duration corresponding to the predicted trajectory data group is l, and l = k×τ. When the time reaches t + l, the sensor exactly acquires the first set of real trajectory data in the time period from t to (t + l), and then the first predicted trajectory data group predicted at time t can be compared with the first set of real trajectory data acquired in the time period from t to (t + l).

[0054] In one embodiment, the algorithm for implementing the target trajectory prediction method in this embodiment is deployed to the domain controller in the vehicle, and the target trajectory prediction method is used on urban roads and elevated highways. Among them, the domain controller includes a domain master processor, an operating system, application software, and algorithms. Figure 4 The flowchart of a vehicle trajectory prediction method is given, as Figure 4 shown. The vehicle trajectory prediction method includes the following steps:

[0055] Step S201, obtain the predicted trajectory data group and the real trajectory data group. Among them, the domain controller outputs a set of predicted trajectories of the vehicle in the next 4 s every 25 ms. This set of predicted trajectories consists of 20 points (predicted values), and the interval between every two points is 200 ms. Every 1 s, 20 consecutive predicted data are recorded, that is, a predicted trajectory data group is obtained; a real trajectory is sampled every 200 ms. After 20 sampling values are filled, the real trajectory is stored in a first-in, first-out manner, that is, a real trajectory data group is obtained.

[0056] Step S202, in the case of obtaining each real trajectory data group, that is, every 1 s, compare the first predicted trajectory data group with the first real trajectory data group to obtain the first error result.

[0057] Step S203, calculate the standard deviation of the n error values in the first error result, and set the first threshold to k times the standard deviation. The standard deviation formula is as follows:

[0058]

[0059] Among them, δ(X) represents the standard deviation of the error value, X represents the group number of the error results, x i is the error value, μ is the average value of n error values, i is a natural number from 1 to n, n represents the serial number of the error value, and in this embodiment, n is 20.

[0060] Step S204: Compare the magnitudes of the n error values with a first threshold value. The error values greater than the first threshold value are invalid error values, and the invalid error values are replaced with the average value of adjacent error values.

[0061] Step S205: Obtain a second error result, calculate the average values of the first error result and the second error result respectively, compare these two average values, and determine the validity of the first error result according to the comparison result.

[0062] Step S206: If the first error result is determined to be valid, then use the first error result to compensate the second predicted trajectory data.

[0063] In one embodiment, a target trajectory prediction device is further provided. This device is used to implement the above-mentioned embodiments and preferred implementation manners, and those that have been described will not be elaborated again. The following terms such as "module", "unit", "sub-unit", etc. can be a combination of software and / or hardware that can achieve a predetermined function. Although the device described in the following embodiments is preferably implemented in software, implementation in hardware, or a combination of software and hardware is also possible and contemplated.

[0064] Figure 5 is the structural block diagram of the target trajectory prediction device in this embodiment, as Figure 5 shown, this device includes:

[0065] An acquisition module, configured to collect the real trajectory data of the target at a preset frequency, obtain n consecutive sampling values recorded from the first historical moment to the current moment to obtain a first real trajectory data group, and obtain a first predicted trajectory data group for predicting the trajectory of the target from the first historical moment to the current moment, where the first predicted trajectory data group includes n consecutive predicted values;

[0066] A comparison module, coupled to the acquisition module, configured to compare the first predicted trajectory data group with the first real trajectory data group to obtain a first error result between the first predicted trajectory data group and the first real trajectory data group;

[0067] A compensation module, coupled to the acquisition module, configured to compensate the second predicted trajectory data group according to the first error result to obtain a predicted trajectory of the target from the second historical moment to the next moment, where the second predicted trajectory data group includes n consecutive predicted values.

[0068] It should be noted that each of the above modules can be a functional module or a program module, and can be implemented either by software or by hardware. For the modules implemented by hardware, each of the above modules can be located in the same processor; or each of the above modules can also be located in different processors in any combined form.

[0069] In this embodiment, an electronic device is further provided, including a memory and a processor. A computer program is stored in the memory, and the processor is configured to run the computer program to execute the steps in any one of the above method embodiments.

[0070] Optionally, the above electronic device may further include a transmission device and an input / output device, wherein the transmission device is connected to the above processor, and the input / output device is connected to the above processor.

[0071] Optionally, in this embodiment, the above processor may be configured to execute the following steps through a computer program:

[0072] S101, collect the real trajectory data of the target at a preset frequency, obtain n consecutive sampling values recorded from the first historical moment to the current moment, obtain the first real trajectory data group, and obtain the first predicted trajectory data group for predicting the trajectory of the target from the first historical moment to the current moment, wherein the first predicted trajectory data group includes n consecutive predicted values;

[0073] S102, compare the first predicted trajectory data group with the first real trajectory data group to obtain the first error result between the first predicted trajectory data group and the first real trajectory data group

[0074] S103, compensate the second predicted trajectory data group according to the first error result to obtain the predicted trajectory of the target from the second historical moment to the next moment, wherein the second predicted trajectory data group includes n consecutive predicted values.

[0075] It should be noted that specific examples in this embodiment can refer to the examples described in the above embodiments and optional implementation manners, and will not be repeated in this embodiment.

[0076] In addition, in combination with the target trajectory prediction method provided in the above embodiments, a storage medium can also be provided to implement in this embodiment. A computer program is stored on the storage medium; when the computer program is executed by a processor, any one of the above target trajectory prediction methods in the embodiments is implemented.

[0077] It should be understood that the specific embodiments described herein are for explaining this application rather than limiting it. All other embodiments obtained by those of ordinary skill in the art without creative efforts according to the embodiments provided in this application fall within the protection scope of this application.

[0078] Obviously, the accompanying drawings are only some examples or embodiments of this application. For those of ordinary skill in the art, this application can also be applied to other similar situations based on these drawings without creative efforts. Additionally, it can be understood that although the work done during the development here may be complex and time-consuming, for those of ordinary skill in the art, certain design, manufacturing, or production changes based on the technical content disclosed in this application are only routine technical means and should not be regarded as insufficient disclosure of this application.

[0079] The term "embodiment" in this application means that the specific features, structures, or characteristics described in combination with the embodiment can be included in at least one embodiment of this application. The phrase appears in various positions in the specification and does not necessarily mean the same embodiment, nor does it mean being independent or alternative to other embodiments and mutually exclusive. Those of ordinary skill in the art can clearly or implicitly understand that the embodiments described in this application can be combined with other embodiments without conflict.

[0080] The above-described embodiments only represent several implementation manners of this application, and their descriptions are relatively specific and detailed, but they should not be construed as limiting the scope of patent protection. It should be noted that for those of ordinary skill in the art, without departing from the concept of this application, several modifications and improvements can still be made, and these all fall within the protection scope of this application. Therefore, the protection scope of this application shall be subject to the appended claims.

Claims

1. A target trajectory prediction method, characterized in that, Including: Collecting the real trajectory data of the target at a preset frequency, obtaining n consecutive sampling values recorded from the first historical moment to the current moment to get a first real trajectory data group, and obtaining a first predicted trajectory data group for predicting the trajectory of the target from the first historical moment to the current moment, where the first predicted trajectory data group contains n consecutive predicted values; Comparing the first predicted trajectory data group with the first real trajectory data group to obtain a first error result between the first predicted trajectory data group and the first real trajectory data group; Compensating the second predicted trajectory data group according to the first error result to obtain the predicted trajectory of the target from the second historical moment to the next moment, where the second predicted trajectory data group contains n consecutive predicted values; the generation time interval between the first predicted trajectory data group and the second predicted trajectory data group is less than or equal to the time interval from the current moment to the next moment; The comparing the first predicted trajectory data group with the first real trajectory data group to obtain a first error result between the first predicted trajectory data group and the first real trajectory data group includes: Subtracting the n consecutive predicted values in the first predicted trajectory data group from the n consecutive sampling values correspondingly to obtain n error values; Judging the validity of each error value and correcting the error values judged to be invalid; Compensating the second predicted trajectory data group according to the first error result includes: Adding the n error values to the n consecutive predicted values in the second predicted trajectory data group correspondingly.

2. The target trajectory prediction method according to claim 1, wherein Judging the validity of each error value includes: Setting a first threshold, comparing the n error values with the first threshold respectively, and determining the error values greater than the first threshold as invalid error values.

3. The target trajectory prediction method according to claim 1, wherein Correcting the error values judged to be invalid includes: Obtaining two error values adjacent to the invalid error value, and replacing the invalid error value with the average value of the two adjacent error values.

4. The target trajectory prediction method according to claim 1, characterized in that, The first error result contains n error values. Compensating the second predicted trajectory data group according to the first error result includes: Obtaining a second error result, where the second error result is obtained by comparing the second predicted trajectory data group with a second real trajectory data group, the second real trajectory data group includes n consecutive sampling values of the target from the second historical moment to the next moment, and the second error result contains n error values; Calculating the average values of the first error result and the second error result respectively, comparing the average values of the first error result and the second error result, and determining whether to use the first error result to compensate the second predicted trajectory data group according to the obtained comparison result.

5. A target trajectory prediction device, characterized in that, Including: An acquisition module, configured to collect real trajectory data of a target at a preset frequency, obtain n consecutive sampling values recorded from a first historical moment to the current moment, to obtain a first real trajectory data set, and obtain a first predicted trajectory data set for predicting the trajectory of the target from the first historical moment to the current moment, where the first predicted trajectory data set includes n consecutive predicted values; A comparison module, configured to compare the first predicted trajectory data set with the first real trajectory data set to obtain a first error result between the first predicted trajectory data set and the first real trajectory data set; A compensation module, configured to compensate a second predicted trajectory data set according to the first error result to obtain a predicted trajectory of the target from a second historical moment to the next moment, where the second predicted trajectory data set includes n consecutive predicted values; a generation time interval between the first predicted trajectory data set and the second predicted trajectory data set is less than or equal to a time interval from the current moment to the next moment; The comparing the first predicted trajectory data set with the first real trajectory data set to obtain a first error result between the first predicted trajectory data set and the first real trajectory data set includes: Subtracting the n consecutive predicted values in the first predicted trajectory data set from the n consecutive sampling values correspondingly to obtain n error values; Judging the validity of each error value, and correcting the error values judged to be invalid; Compensating the second predicted trajectory data set according to the first error result includes: Adding the n error values to the n consecutive predicted values in the second predicted trajectory data set correspondingly.

6. An electronic device, comprising a memory and a processor, characterized in that, A computer program is stored in the memory, and the processor is configured to run the computer program to execute the target trajectory prediction method according to any one of claims 1 to 4.

7. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, the steps of the target trajectory prediction method according to any one of claims 1 to 4 are implemented.

Images