Method and device for predicting speed of preceding vehicle in following vehicle team and vehicle end control equipment

Abstract

The invention relates to a method and device for predicting the speed of front vehicles in a following motorcade and vehicle end control equipment, and relates to the technical field of vehicles, and the method comprises the steps: obtaining a real-time speed set of each communicated front vehicle in a period of time under the condition that a vehicle establishes communication with at least two front vehicles following in front of the vehicle in the driving direction of the vehicle; inputting the vehicle speed into a vehicle speed prediction model to obtain an output first adjacent vehicle speed prediction result; the adjacent vehicle speed first prediction result comprises a plurality of preliminary prediction vehicle speeds; the adjacent vehicle is a front vehicle closest to the driving direction of the vehicle; obtaining a preset compensation value and processing the adjacent vehicle speed first prediction result based on the preset compensation value to obtain an adjacent vehicle speed second prediction result; the preset compensation value is determined based on a historical acceleration error value, a historical impact degree error value and a historical feature point error value of the historical following motorcade; and the real-time vehicle speed set is processed through the vehicle speed prediction model, and the processing result is corrected based on the preset compensation value, so that accurate long-time prediction of the future vehicle speed of the vehicle in front of the target lane is realized.

Description

Technical Field

[0001] The present application relates to the field of vehicle technology, and in particular to a method, device and vehicle-side control equipment for predicting the speed of a leading vehicle in a following convoy. Background Art

[0002] With the development of technology, people have put forward higher and higher requirements for the economy, comfort and safety of automobiles. However, on actual roads, especially on congested urban roads, due to the large differences in driving ability of each person and the limitations of the human driver's vision range, drivers often cannot accurately predict the movement of the vehicle in front, which can easily lead to difficulties in following the vehicle. For example, the irregular driving habits of the leading vehicle can easily cause the following vehicle to accelerate or decelerate suddenly, resulting in vehicle shock waves. It can be seen that it is very detrimental to driving safety and will greatly reduce the fuel economy and driving comfort of the car. Therefore, it becomes very necessary to predict the speed of the vehicle in front. Summary of the invention

[0003] Based on this, it is necessary to provide a method, device, vehicle-side control equipment and computer-readable storage medium for predicting the speed of a leading vehicle in a following fleet, which can accurately predict the speed of the leading vehicle in order to solve the above technical problems.

[0004] In a first aspect, the present application provides a method for predicting the speed of a leading vehicle in a following convoy, comprising:

[0005] When the vehicle establishes communication with at least two preceding vehicles in the direction of travel of the vehicle, the real-time vehicle speed of each of the at least two preceding vehicles within a period of time is obtained to obtain a real-time vehicle speed set;

[0006] The real-time vehicle speed set is input into a pre-trained vehicle speed prediction model to obtain a first prediction result of the neighboring vehicle speed output by the vehicle speed prediction model; wherein the first prediction result of the neighboring vehicle speed includes a plurality of preliminary predicted vehicle speeds of the neighboring vehicle in a future time period; the neighboring vehicle is the preceding vehicle that is closest to the own vehicle in the driving direction of the own vehicle;

[0007] Obtain a preset compensation value, and process the first prediction result of the neighboring vehicle speed based on the preset compensation value to obtain a second prediction result of the neighboring vehicle speed; wherein the preset compensation value is determined based on a historical acceleration error value, a historical impact error value, and a historical feature point error value corresponding to a historical following fleet.

[0008] In one embodiment, it also includes:

[0009] Obtaining a historical prediction result of the speed of an adjacent vehicle corresponding to a historical speed set; the historical speed set includes the historical speeds of each of the at least two preceding vehicles communicating with the vehicle within a period of historical time; the historical prediction result of the speed of an adjacent vehicle is obtained by using the output of the pre-trained speed prediction model with the historical speed set as a model input;

[0010] Obtaining historical actual vehicle speeds of neighboring vehicles associated with a plurality of historical initial predicted vehicle speeds in the historical prediction results of the neighboring vehicle speeds, and calculating an error value between each of the historical actual vehicle speeds of the neighboring vehicles and the corresponding historical initial predicted vehicle speeds;

[0011] Calling the objective function to process each of the historical acceleration error values, each of the historical impact error values, and the historical feature point error values ​​to obtain a minimum matching error in the matching error;

[0012] The matching error minimum value, and the maximum and minimum values ​​of the error values ​​are processed by an optimal solution algorithm to determine the preset compensation value.

[0013] In one embodiment, it also includes:

[0014] Performing differentiation processing on the historical actual vehicle speeds of each neighboring vehicle to obtain corresponding first acceleration characteristics;

[0015] Performing differentiation processing on the historical initial predicted vehicle speed corresponding to the historical actual vehicle speed of each neighboring vehicle to obtain corresponding second acceleration characteristics;

[0016] Based on each of the first acceleration characteristics and the second acceleration characteristics, a historical acceleration error value corresponding to each of the neighboring vehicles' historical actual vehicle speeds is determined.

[0017] In one embodiment, after obtaining each of the first acceleration characteristics and each of the second acceleration characteristics, the method further includes:

[0018] Performing differentiation processing on each of the first acceleration characteristics and each of the second acceleration characteristics to obtain corresponding each of the first impact degree characteristics and each of the second impact degree characteristics;

[0019] Based on each of the first impact degree characteristics and each of the second impact degree characteristics, a historical impact degree error value corresponding to the historical actual vehicle speed of each of the neighboring vehicles is determined.

[0020] In one embodiment, the step of obtaining the historical actual vehicle speed of the neighboring vehicle associated with the plurality of historical initial predicted vehicle speeds in the historical prediction result of the neighboring vehicle speed comprises: obtaining the historical actual vehicle speed of the neighboring vehicle associated with K historical initial predicted vehicle speeds in the historical prediction result of the neighboring vehicle speed; wherein K ≥ 3 and K is an integer;

[0021] After obtaining the first acceleration characteristics and the second acceleration characteristics, the method further includes:

[0022] Obtaining the product of the first acceleration feature at a previous moment and the first acceleration feature at a next moment, respectively corresponding to the second to K-1th first acceleration features, and obtaining a first number of positive numbers, and a second number of zero and negative numbers in the product;

[0023] A historical turning point feature is determined based on the first number and the second number, and the negation of the historical turning point feature is taken to obtain a corresponding historical feature point error value.

[0024] In one embodiment, the processing of the first prediction result of the neighboring vehicle speed based on the preset compensation value to obtain the second prediction result of the neighboring vehicle speed includes:

[0025] Each of the preliminary predicted vehicle speeds in the first prediction result of the neighboring vehicle speed is added to the preset compensation value to obtain a second prediction result of the neighboring vehicle speed.

[0026] In one embodiment, it also includes:

[0027] The GA-BP neural network model is trained based on the historical vehicle speed data of the historical car-following fleet to obtain a pre-trained vehicle speed prediction model.

[0028] In a second aspect, the present application further provides a device for predicting the speed of a leading vehicle in a following vehicle group, the device comprising:

[0029] A data acquisition module, for acquiring the real-time vehicle speed of each of the at least two preceding vehicles within a period of time when the vehicle establishes communication with the at least two preceding vehicles in the preceding direction, and obtaining a real-time vehicle speed set;

[0030] A data processing module is used to input the real-time vehicle speed set into a pre-trained vehicle speed prediction model to obtain a first prediction result of the neighboring vehicle speed output by the vehicle speed prediction model; wherein the first prediction result of the neighboring vehicle speed includes a plurality of preliminary predicted vehicle speeds of the neighboring vehicle in a future period; the neighboring vehicle is the preceding vehicle that is closest to the own vehicle in the driving direction of the own vehicle;

[0031] The result correction module is used to obtain a preset compensation value, process the first prediction result of the neighboring vehicle speed based on the preset compensation value, and obtain a second prediction result of the neighboring vehicle speed; wherein the preset compensation value is determined based on the historical acceleration error value, historical impact error value and historical feature point error value corresponding to the historical following fleet.

[0032] In a third aspect, the present application further provides a vehicle-side control device, including a memory and a processor, wherein the memory stores a computer program, and when the processor executes the computer program, the following steps are implemented:

[0033] When the vehicle establishes communication with at least two preceding vehicles in the direction of travel of the vehicle, the real-time vehicle speed of each of the at least two preceding vehicles within a period of time is obtained to obtain a real-time vehicle speed set;

[0034] The real-time vehicle speed set is input into a pre-trained vehicle speed prediction model to obtain a first prediction result of the neighboring vehicle speed output by the vehicle speed prediction model; wherein the first prediction result of the neighboring vehicle speed includes a plurality of preliminary predicted vehicle speeds of the neighboring vehicle in a future time period; the neighboring vehicle is the preceding vehicle that is closest to the own vehicle in the driving direction of the own vehicle;

[0035] Obtain a preset compensation value, and process the first prediction result of the neighboring vehicle speed based on the preset compensation value to obtain a second prediction result of the neighboring vehicle speed; wherein the preset compensation value is determined based on a historical acceleration error value, a historical impact error value, and a historical feature point error value corresponding to a historical following fleet.

[0036] In a fourth aspect, the present application further provides a computer-readable storage medium having a computer program stored thereon, wherein when the computer program is executed by a processor, the following steps are implemented:

[0037] When the vehicle establishes communication with at least two preceding vehicles in the direction of travel of the vehicle, the real-time vehicle speed of each of the at least two preceding vehicles within a period of time is obtained to obtain a real-time vehicle speed set;

[0038] The real-time vehicle speed set is input into a pre-trained vehicle speed prediction model to obtain a first prediction result of the neighboring vehicle speed output by the vehicle speed prediction model; wherein the first prediction result of the neighboring vehicle speed includes a plurality of preliminary predicted vehicle speeds of the neighboring vehicle in a future time period; the neighboring vehicle is the preceding vehicle that is closest to the own vehicle in the driving direction of the own vehicle;

[0039] Obtain a preset compensation value, and process the first prediction result of the neighboring vehicle speed based on the preset compensation value to obtain a second prediction result of the neighboring vehicle speed; wherein the preset compensation value is determined based on a historical acceleration error value, a historical impact error value, and a historical feature point error value corresponding to a historical following fleet.

[0040] The method, device, vehicle-side control device and computer-readable storage medium for predicting the speed of a leading vehicle in a following convoy provide a method for predicting the speed of a leading vehicle in a following convoy, comprising: when the own vehicle establishes communication with at least two leading vehicles following in front of it in its driving direction, obtaining the real-time speed of each of the at least two leading vehicles within a period of time to obtain a real-time speed set; inputting the real-time speed set into a pre-trained speed prediction model to obtain a first prediction result of the speed of the neighboring vehicle output by the speed prediction model; wherein the first prediction result of the speed of the neighboring vehicle includes multiple preliminary predicted speeds of the neighboring vehicle in a future time period; the neighboring vehicle is the leading vehicle in front of the own vehicle in the driving direction and closest to the own vehicle; obtaining a preset compensation value, processing the first prediction result of the speed of the neighboring vehicle based on the preset compensation value, and obtaining a second prediction result of the speed of the neighboring vehicle; wherein the preset compensation value is determined based on the historical acceleration error value, the historical impact error value and the historical feature point error value corresponding to the historical following convoy. By using a pre-trained speed prediction model to process the real-time speed set of the following fleet in front of the own vehicle, a first prediction result of the neighboring vehicle speed of the preceding vehicle (neighboring vehicle) traveling adjacent to the preceding vehicle is obtained. Then, based on a preset compensation value determined by the historical acceleration error value, the historical impact error value and the historical feature point error value corresponding to the historical following fleet, multiple initial predicted speeds in the first prediction result of the neighboring vehicle speed are corrected respectively to obtain a more accurate second prediction result of the neighboring vehicle speed for the preceding neighboring vehicle. In this way, a long-term and accurate prediction of the future speed of the preceding vehicle in the target lane is achieved, which is conducive to adjusting the speed of the own vehicle in advance according to the prediction of the future speed of the preceding neighboring vehicle, and provides a basis for improving the energy efficiency, safety and comfort performance of vehicles in congested roads. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] In order to more clearly illustrate the technical solutions in the embodiments of the present application or related technologies, the drawings required for use in the embodiments of the present application or related technical descriptions will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without paying creative work.

[0042] Figure 1 A schematic diagram of a flow chart of a method for predicting the speed of a leading vehicle in a following vehicle group in one embodiment;

[0043] Figure 2 A schematic diagram of a process for determining a preset compensation value in one embodiment;

[0044] Figure 3 A schematic diagram of a flow chart of a method for predicting the speed of a leading vehicle in a following vehicle group in another embodiment;

[0045] Figure 4 A schematic diagram of a training model fitness decline curve in one embodiment;

[0046] Figure 5 A schematic diagram of a long-term prediction effect diagram of a training model test set in one embodiment;

[0047] Figure 6 1 is a schematic diagram of the structure of a device for predicting the speed of a leading vehicle in a following vehicle group in one embodiment;

[0048] Figure 7 This is a diagram of the internal structure of a vehicle-side control device in one embodiment. DETAILED DESCRIPTION

[0049] In order to make the purpose, technical solution and advantages of the present application more clearly understood, the present application is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0050] In recent years, with the development of intelligent network technology, many scholars have proposed corresponding methods for predicting the speed of the preceding vehicle, but most of them focus on how to predict the short-term speed of the vehicle. For example, a real-time prediction method for short-term vehicle speed conditions based on the interaction between the preceding vehicle (a preceding vehicle traveling adjacent to the vehicle in front of the vehicle) and the vehicle itself, by obtaining the historical speed and distance information of the vehicle and the preceding vehicle, and building a model based on an artificial neural network, to achieve short-term prediction of vehicle speed. It can be seen that most of the existing publicly available speed prediction methods focus on short-term prediction of vehicle speed, but the problem of long-term prediction has not been well solved, and the prediction accuracy also needs to be improved.

[0051] In an exemplary embodiment, Figure 1 As shown, a method for predicting the speed of a leading vehicle in a car-following convoy is provided. The method can be applied to a vehicle, a server corresponding to the vehicle, or a system including a vehicle and a corresponding server, and is implemented through the interaction between the vehicle and the server. This application takes the method applied to a vehicle as an example, and includes the following steps 101 to 103. Among them:

[0052] Step 101, when the vehicle establishes communication with at least two preceding vehicles traveling in front of the vehicle in its traveling direction, the real-time vehicle speed of each of the at least two preceding vehicles within a period of time is obtained to obtain a real-time vehicle speed set.

[0053] It should be noted that the ego vehicle can be the current vehicle or the vehicle itself, or it can be a data collection vehicle in a car-following vehicle team. Car-following means that several vehicles are traveling one after another, and the vehicle team formed by multiple cars following each other does not include other vehicles. It should also be noted that the vehicle team formed by the car-following vehicles and the ego vehicle are all located in the same lane.

[0054] Among them, in the vehicle team formed by the following vehicles, along the driving direction of the vehicle team, the leading vehicle can be the head vehicle of the team, and there can be at least one vehicle between the vehicle and the head vehicle of the team; that is, along the driving direction of the vehicle team, there are at least two following vehicles in front of the vehicle. Each vehicle in the vehicle team can be configured to have a communication function with the vehicle itself, so that the vehicle itself can obtain some driving information of other following vehicles in the vehicle team based on the communication function. The driving information here includes, for example, the distance between other following vehicles and the vehicle itself, the driving speed of other following vehicles, whether other following vehicles are in a braking state, etc. This application does not limit the specific content of the driving information, and the specific information contained in the driving information can be set according to needs.

[0055] In addition, on the basis of establishing communication connections between the vehicle and other following vehicles in the vehicle team, it is also possible to further set up communication connections between other following vehicles in the vehicle team. This is only an optional implementation method provided by this application, but this application does not make specific limitations on this.

[0056] It should also be added that, in a vehicle team including the own vehicle, if the head vehicle in the team drives away from the lane where the own vehicle is located, the front vehicle before the original head vehicle in the direction of travel of the own vehicle will be updated to the new head vehicle in the team, so that the number of vehicles in the vehicle team including the own vehicle remains unchanged; correspondingly, along the direction of travel of the own vehicle, if any other vehicle in front of the own vehicle drives away from the lane where the own vehicle is located, the front vehicle before the original head vehicle in the team will be updated to the new head vehicle in the team, and the original head vehicle in the team will be updated to the front vehicle between the new head vehicle and the own vehicle, so that the number of vehicles in the vehicle team including the own vehicle remains unchanged.

[0057] In the case where the self-vehicle has established a communication connection with multiple vehicles in the team that are following the self-vehicle in the driving direction, step 101 can be executed to obtain the real-time speeds of the corresponding multiple front vehicles within a period of time based on the self-vehicle, so as to obtain a real-time speed set including the speeds of the multiple front vehicles at multiple times. The real-time speed set here is, for example, a real-time speed matrix, which includes multiple rows of data and multiple columns of data, wherein each row of data in the real-time speed matrix is ​​the real-time speed of a corresponding front vehicle at each time, and each column of data in the real-time speed matrix includes the real-time speed of each front vehicle at the same time; wherein, optionally, the multiple real-time speeds in a row are arranged from left to right in the order of corresponding time, and the multiple real-time speeds in a column are arranged from top to bottom in accordance with the distance between the front vehicle and the self-vehicle, that is, the first row of the real-time speed matrix is ​​the real-time speed of the front vehicle at multiple times, and the last row of the real-time speed matrix is ​​the real-time speed of the front vehicle traveling adjacent to the self-vehicle at multiple times.

[0058] It should also be added that, for the collection of the vehicle speed of each vehicle in the team, the corresponding time is the same. As for the interval between two adjacent times, this application does not make specific restrictions; for example, the interval between any two adjacent times can be set to be the same.

[0059] It should also be added that the present application does not make any specific limitation on the communication connection method between the vehicle and the preceding vehicle. For example, V2V (Vehicle to Vehicle, fleet communication technology) communication can be used to realize the communication connection between the vehicle and the preceding vehicle. Of course, other communication methods can also be selected.

[0060] Step 102, input the real-time vehicle speed set into the pre-trained vehicle speed prediction model to obtain a first prediction result of the neighboring vehicle speed output by the vehicle speed prediction model; wherein the first prediction result of the neighboring vehicle speed includes multiple preliminary predicted vehicle speeds of the neighboring vehicle in the future time period; the neighboring vehicle is the preceding vehicle that is closest to the own vehicle in the driving direction of the own vehicle.

[0061] Among them, the above-mentioned neighboring vehicle is the front vehicle that is closest to the own vehicle in the driving direction. It can also be interpreted that the neighboring vehicle is the front vehicle that is adjacent to the own vehicle in the driving direction.

[0062] In the case where the real-time vehicle speed set is obtained based on the execution of step 101, step 102 may be further executed to process the collected real-time vehicle speed set using the pre-trained vehicle speed prediction model to obtain the first prediction result for the future vehicle speed of the neighboring vehicle output by the vehicle speed prediction model (the first prediction result for the neighboring vehicle speed). Using the pre-trained vehicle speed prediction model to process the real-time vehicle speed set of the following vehicle fleet in front of the vehicle is conducive to ensuring the real-time and high efficiency of data processing, and the pre-trained vehicle speed prediction model can also be continuously trained as needed to make the first prediction result for the vehicle speed of the neighboring vehicle (the front vehicle) output by the vehicle speed prediction model more accurate.

[0063] Among them, in the embodiment provided by the present application, the first prediction result of the speed of the neighboring vehicle includes the preliminary predicted speed of the neighboring vehicle at multiple moments in a future period of time, and does not include the predicted speed of the other following vehicles in the following vehicle, but this is only an optional implementation method provided by the present application; the present application is not limited to this, and if necessary, it is also possible to set the result output after processing by the pre-trained speed prediction model to include multiple preliminary predicted speeds of multiple vehicles in the following vehicle team in the future period of time. The present application does not limit the number of preliminary predicted speeds included in the first prediction result of the speed of the neighboring vehicle, and the number of data output by the speed prediction model can be set according to needs. In addition, the interval time of the data output by the speed prediction model can also be set according to needs.

[0064] Step 103, obtaining a preset compensation value, processing the first prediction result of the neighboring vehicle speed based on the preset compensation value, and obtaining a second prediction result of the neighboring vehicle speed; wherein the preset compensation value is determined based on the historical acceleration error value, the historical impact error value, and the historical feature point error value corresponding to the historical following fleet.

[0065] In order to further improve the prediction accuracy of the future speed of the neighboring vehicle (the front vehicle) traveling in front of the vehicle, after executing step 102 to obtain the first prediction result of the neighboring vehicle speed, step 103 can be further executed to correct the first prediction result of the neighboring vehicle speed by using the preset compensation value determined based on the driving data of the historical following team and the corresponding historical acceleration error value, historical impact error value and historical feature point error value, so that the accuracy of the second prediction result of the neighboring vehicle speed obtained after correction is higher. Among them, the preset compensation value is specifically used to correct each preliminary predicted speed in the first prediction result of the neighboring vehicle speed, and the target predicted speed obtained from the corrected preliminary predicted speed constitutes the second prediction result of the neighboring vehicle speed.

[0066] It can be seen that the present application processes the real-time speed set of the following car team in front of the vehicle by using the pre-trained speed prediction model to obtain the first prediction result of the speed of the neighboring car for the front car (neighboring car) traveling adjacent to the front of the vehicle, and then corrects the multiple initial prediction speeds in the first prediction result of the neighboring car speed based on the preset compensation value determined by the historical acceleration error value, the historical impact error value and the historical feature point error value corresponding to the historical following car team, so as to obtain a more accurate second prediction result of the neighboring car speed for the front neighboring car, so as to achieve the long-term accurate prediction of the future speed of the front car in the target lane, which is conducive to adjusting the speed of the vehicle in advance according to the prediction of the future speed of the front neighboring car, and provides a basis for improving the energy efficiency, safety and comfort performance of vehicles in congested sections. That is, it should be emphasized that the present application does not predict the future speed of the previous neighboring car of the vehicle based on the real-time speed information of the previous neighboring car of the vehicle, but predicts the future speed of the previous neighboring car of the vehicle based on the real-time speed information of multiple front vehicles that followed the vehicle before.

[0067] Please combine Figure 1 Reference Figure 2 In an exemplary embodiment, the present application further provides a method for determining a preset compensation value, which may include steps 201 to 204, wherein:

[0068] Step 201, obtaining a historical prediction result of the speed of a neighboring vehicle corresponding to a historical speed set; the historical speed set includes the historical speeds of at least two leading vehicles communicating with the vehicle within a historical period of time; the historical prediction result of the speed of the neighboring vehicle is obtained by using the output of a pre-trained speed prediction model with the historical speed set as a model input;

[0069] Step 202, obtaining the historical actual speed of the neighboring vehicle associated with the plurality of historical initial predicted speeds in the historical prediction result of the neighboring vehicle speed, and calculating the error value between the historical actual speed of each neighboring vehicle and the corresponding historical initial predicted speed;

[0070] Step 203, calling the objective function to process each historical acceleration error value, each historical impact error value and historical feature point error value to obtain the minimum matching error in the matching error;

[0071] Step 204 : Process the minimum value of the matching error, and the maximum and minimum values ​​of the error values ​​through an optimal solution algorithm to determine a preset compensation value.

[0072] Specifically, the above-mentioned preset compensation value is determined based on historical vehicle information, that is, before predicting the future speed of the neighboring vehicle traveling in front of the own vehicle, in order to realize the prediction function and to ensure the accuracy of the prediction result, some preliminary preparation work is required; these preliminary work, for example, can obtain the historical prediction result of the neighboring vehicle speed corresponding to the historical speed set by executing step 201; this step can be specifically executed as follows, obtaining multiple historical speeds of each leading vehicle in the following vehicle team communicating with the own vehicle, that is, obtaining the historical speed of each leading vehicle in all leading vehicles communicating with the own vehicle within at least a period of historical time, thereby obtaining at least one historical speed set; and then processing the historical speed set based on the pre-trained speed prediction model to obtain the output neighboring vehicle speed historical prediction result; wherein, the neighboring vehicle speed historical prediction result includes multiple future prediction speeds for the neighboring vehicle. The historical initial predicted speed of the speed; then, by executing step 202, the historical actual speed of the neighboring vehicle corresponding to all the historical initial predicted speeds in the historical prediction results of the neighboring vehicle speed is obtained. Through the historical actual speeds of each neighboring vehicle and the corresponding historical initial predicted speeds, the accuracy of the historical prediction results of the neighboring vehicle speed can be learned, and the difference (error value) between the historical prediction results of the neighboring vehicle speed and the actual situation can be learned; then, step 203 can be executed to call the objective function to process the historical acceleration error values, the historical impact error values ​​and the historical feature point error values ​​for the historical following vehicle team, so as to obtain the minimum matching error in the corresponding matching error; finally, step 204 is executed to process the minimum matching error obtained by executing step 203 and the maximum and minimum values ​​of the error values ​​obtained by executing step 202 based on the optimal solution algorithm, so as to determine the preset compensation value.

[0073] It should be added that the present application does not specifically limit the historical time periods and corresponding roads corresponding to the historical vehicle speeds collected by the present application. For example, different preset compensation values ​​may be respectively used for different time periods, different roads, and / or different road sections. The above steps 201 to 204 are only one method of determining the preset compensation value provided by the present application, but the present application is not limited thereto.

[0074] Although the predicted result (the first predicted result of the neighboring vehicle's speed) can already reflect the traffic flow characteristics well, due to the different models and drivers of the vehicle in front of the main vehicle (neighboring vehicle), there is a large uncertainty in the reaction time, and therefore it is easy to have time and space misalignment errors; based on this, the present application corrects the initial predicted speed in the first predicted result of the neighboring vehicle's speed through a preset compensation value determined in advance, which is beneficial to improving the prediction accuracy of the neighboring vehicle's future speed; thereby enabling the own vehicle to learn of the future speed of the previous neighboring vehicle in advance, and thus adjust the driving speed of the own vehicle in advance, which is beneficial to reducing the occurrence of traffic accidents between following vehicles, and thus is beneficial to improving the driving safety of the vehicle.

[0075] Furthermore, the present application predicts the future speed of the neighboring vehicle in front of the own vehicle based on the driving information of multiple preceding vehicles traveling in front of the own vehicle, rather than predicting the future speed of the neighboring vehicle based solely on the driving information of the neighboring vehicle in front of the own vehicle, and in the prediction process, it is further combined with a preset compensation value obtained based on historical driving information. Compared with predicting the future speed of the neighboring vehicle based solely on the driving information of the neighboring vehicle in front of the own vehicle, the information factor for predicting the future speed of the neighboring vehicle is added, which is beneficial to improving the accuracy of predicting the future speed of the neighboring vehicle, and can also avoid the vehicle shock wave phenomenon caused by situations such as sudden acceleration or deceleration of the following vehicle due to the irregular driving habits of the lead vehicle, thereby helping to improve the driving safety of the own vehicle.

[0076] Please refer to Figure 1 and Figure 2 In an exemplary embodiment, the present application further provides a method for determining a historical acceleration error value, which may specifically include steps 205 to 207 after obtaining the historical actual speed of the neighboring vehicle associated with the plurality of historical initial predicted speeds in the historical prediction result of the neighboring vehicle speed by executing step 202, wherein:

[0077] Step 205, performing differentiation processing on the historical actual vehicle speed of each neighboring vehicle to obtain corresponding first acceleration characteristics;

[0078] Step 206, performing differentiation processing on the historical initial predicted vehicle speed corresponding to the historical actual vehicle speed of each neighboring vehicle to obtain the corresponding second acceleration characteristics;

[0079] Step 207: Determine the historical acceleration error value corresponding to the historical actual vehicle speed of each neighboring vehicle based on the first acceleration characteristics and the second acceleration characteristics.

[0080] Specifically, for example, the corresponding first acceleration characteristics can be obtained by differentiating the historical actual vehicle speeds of each neighboring vehicle; the corresponding second acceleration characteristics can be obtained by differentiating the historical initial predicted speeds corresponding to the historical actual vehicle speeds of each neighboring vehicle, and then the historical acceleration error values ​​corresponding to the historical actual vehicle speeds of each neighboring vehicle are determined by calculating the difference between the corresponding matching first acceleration characteristics and the second acceleration characteristics.

[0081] Here is a calculation formula for the acceleration characteristic A(i), for example: Among them It represents the actual historical speed of the neighboring vehicle at a certain moment (i), or the historical initial predicted speed at a certain moment (i); A(i) is the first acceleration feature corresponding to the actual historical speed of the neighboring vehicle at a certain moment (i), or the second acceleration feature corresponding to the historical initial predicted speed at a certain moment (i).

[0082] Furthermore, a calculation formula for the historical acceleration error value is provided, for example: A (i) = A real (i)-A predict (i) where A real (i) represents the first acceleration feature corresponding to the actual historical speed of the neighboring vehicle at a certain time (i), A predict (i) represents the second acceleration characteristic corresponding to the historical initial predicted vehicle speed at a certain moment (i).

[0083] Please refer to Figure 1 and Figure 2 In an exemplary embodiment, after obtaining each first acceleration feature and each second acceleration feature, the method further includes:

[0084] Performing differential processing on each first acceleration feature and each second acceleration feature to obtain corresponding each first impact degree feature and each second impact degree feature;

[0085] Based on each of the first impact degree characteristics and each of the second impact degree characteristics, a historical impact degree error value corresponding to the historical actual vehicle speed of each neighboring vehicle is determined.

[0086] Specifically, for example, the first acceleration characteristics corresponding to the historical actual vehicle speeds of each neighboring vehicle can be differentiated to obtain the corresponding first impact degree characteristics; the second acceleration characteristics corresponding to each historical initial predicted vehicle speed can be differentiated to obtain the corresponding second impact degree characteristics, and then the historical impact degree error value corresponding to the historical actual vehicle speed of each neighboring vehicle can be determined by calculating the difference between the first impact degree characteristics and the second impact degree characteristics of each corresponding match.

[0087] Here is a calculation formula for the impact characteristic G(i) as follows: G(i) is the first impact degree characteristic corresponding to the historical actual speed of the neighboring vehicle at a certain moment (i), or the second impact degree characteristic corresponding to the historical initial predicted speed at a certain moment (i).

[0088] Furthermore, a historical impact error value δ is provided: G The calculation formula of (i) is as follows: G (i) = G real (i)-G predict (i) where G real (i) represents the first impact characteristic corresponding to the actual historical speed of the neighboring vehicle at a certain time (i), G predict (i) represents the second impact degree characteristic corresponding to the historical initial predicted vehicle speed at a certain moment (i).

[0089] Please refer to Figure 1 and Figure 2In an exemplary embodiment, the step of obtaining the historical actual speed of the neighboring vehicle associated with multiple historical initial predicted speeds in the historical prediction results of the neighboring vehicle speed executed in the above step 202 can be executed as follows: obtaining the historical actual speed of the neighboring vehicle associated with K historical initial predicted speeds in the historical prediction results of the neighboring vehicle speed; wherein K ≥ 3 and K is an integer.

[0090] After obtaining each first acceleration feature and each second acceleration feature, the method further includes:

[0091] Obtain the product of the first acceleration feature at a previous moment and the first acceleration feature at a next moment, respectively corresponding to the second to K-1th first acceleration features, and obtain a first number of positive numbers, and a second number of zeros and negative numbers in the product;

[0092] The historical turning point feature is determined based on the first quantity and the second quantity, and the historical turning point feature is negated to obtain the corresponding historical feature point error value.

[0093] Specifically, for example, the first acceleration features of the previous moment and the first acceleration features of the next moment that are adjacent to each first acceleration feature in time sequence can be extracted from the 2nd to the K-1th first acceleration features, and then for the 2nd to the K-1th first acceleration features, the products of the first acceleration features of the previous moment and the first acceleration features of the next moment extracted are calculated respectively, and the number of positive numbers, zeros and negative numbers in these products is recorded, and the total number of positive numbers is taken as the first number, and the total number of negative numbers and zeros is taken as the second number; wherein the first number and the second number obtained here are also the historical turning point features corresponding to the actual historical speeds of the neighboring vehicles, and then the corresponding historical feature point error value can be obtained by taking the opposite of the historical turning point features.

[0094] Here is a calculation formula for the historical turning point feature P(i), for example: Among them, the turning point characteristic condition 1 is: when the product of the accelerations at the i-th moment and the i+1 moment is greater than zero, the total number of moments satisfying condition 1 is recorded as the first number m1; the turning point characteristic condition 2 is: when the product of the accelerations at the i-th moment and the i+1 moment is less than or equal to zero, the total number of moments satisfying condition 2 is recorded as the second number m2. Furthermore, a historical characteristic point error value δ is provided. P The calculation formula of (i) is, for example, δ P (i) = -P(i).

[0095] The objective function provided in this application can be, for example, Among them, i=0 refers to the moment corresponding to the first historical initial predicted vehicle speed among multiple historical initial predicted vehicle speeds, or the moment corresponding to the historical actual vehicle speed of the neighboring vehicle corresponding to multiple historical initial predicted vehicle speeds, and T is the entire data period corresponding to multiple historical initial predicted vehicle speeds.

[0096] In an exemplary embodiment, the above-mentioned step 103 of processing the first prediction result of the neighboring vehicle speed based on the preset compensation value to obtain the second prediction result of the neighboring vehicle speed can be executed as follows: each preliminary predicted speed in the first prediction result of the neighboring vehicle speed is added to the preset compensation value to obtain the second prediction result of the neighboring vehicle speed.

[0097] That is, the multiple future predicted speeds of the neighboring vehicle included in the second prediction result of the neighboring vehicle's speed are obtained by adding the various preliminary predicted speeds in the first prediction result and the preset compensation value, so as to realize the correction of the first prediction result of the neighboring vehicle's speed based on the preset compensation value, thereby improving the prediction accuracy of the neighboring vehicle's future speed.

[0098] In an exemplary embodiment, the method further includes: training a GA-BP neural network model based on historical vehicle speed data of the historical following vehicle fleet to obtain a pre-trained vehicle speed prediction model. GA is the abbreviation of Genetic Algorithm, which is translated into genetic algorithm in Chinese; BP is Back Propagation, which is translated into neural network in Chinese.

[0099] That is, the pre-trained vehicle speed prediction model used in the present application can be obtained based on the training of the GA-BP neural network model, and the sample data used for training includes the historical vehicle speed data of the historical following vehicle fleet.

[0100] In response to the problem of how to predict the future speed of the neighboring vehicle in the front vehicle of the following convoy, the present application also provides an optional embodiment, specifically providing a V2V-based long-term vehicle speed prediction method such as Figure 3 As shown in the figure, the driving data set collected on urban roads is first acquired and processed through V2V, and then a BP neural network long-term vehicle speed prediction model based on GA optimization is established and trained. After that, online prediction is performed, and finally feature matching parameters are used to eliminate the temporal and spatial misalignment error.

[0101] The above-mentioned long-term vehicle speed prediction method based on the V2V environment specifically includes the following steps S21-S24, wherein: Step S21, obtain and process the vehicle driving data set obtained by V2V, and analyze the relationship between the speed characteristics of the front and rear vehicles in the following car team; Step S22, establish a GA-BP neural network long-term vehicle speed prediction model and train the neural network model; Step S23, use the trained BP neural network model to perform online prediction on the real-time collected test data set. Step S24, evaluate the preliminary misalignment error, and use feature matching parameters to eliminate the misalignment error to finally achieve a more accurate prediction of the future long-term vehicle speed sequence.

[0102] Further, based on the above steps S21 to S24, the present application also provides an optional implementation as follows:

[0103] Step S21, obtaining the speed information of the following vehicles on the urban roads. The information of the fleets with a number of N in each time period every day is intercepted as the training target.

[0104] Step S22, process the historical data set. Here, take the morning rush hour data set as an example, define the data collection vehicle (self-vehicle) number as Veh_host within the distance range supported by V2V communication, and the vehicle in front of the data collection vehicle (neighboring vehicle) number as Veh_N. If the vehicle in front of the main vehicle initially defined leaves the current lane, the vehicle in front of the main vehicle will be updated to control the total number of vehicles in front of the self-vehicle to remain N, that is, the forward update of the head vehicle in the team will actually be involved at this time; the head vehicle in the team is numbered Veh_1, and similarly, if the vehicle in front of the team initially defined leaves, the vehicle in front of the team will be updated to control the total number of vehicles in front of the self-vehicle to remain N. It can be seen that the total number of vehicles in the lane where the data collection vehicle Veh_host is located is N. Since each vehicle is driven by a human driver, there is a reaction time. The average reaction time of the drivers of the N vehicles in front of the collection vehicle is recorded as t i Therefore, the total time delay from the start of Veh_1 to the start of the vehicle in front of the host vehicle Veh_N can be expressed as: Here is the T delay That is, the time between the start of the first car in the team and the start of the car in front of the main car (the adjacent car).

[0105] Here, a specific embodiment is provided, in which the time when the data collection vehicle and the first vehicle in the team establish V2V communication is defined as the starting time t0, and the time after the N vehicles in front of the data collection vehicle T is recorded from this time delay The speed sequence of seconds, the speed information of the N vehicles ahead is expressed by a matrix (historical speed set) as follows:

[0106] You can select the first m·T in the data set delay, the eigenvalue vector of the sample with m∈(0,1) is used as the training vector S i : The remaining time is (1-m)·T delay ,m∈(0,1) sample data is used as the test set. When the data collection vehicle establishes V2V communication with the first vehicle in the team, the data collection vehicle also establishes V2V communication with other vehicles traveling along the same road between it and the first vehicle in the team. The first m data refers to the first m columns of data extracted in chronological order, not randomly extracted m columns of data.

[0107] Step S23, training the GA-BP neural network model. Let the population size be GA_NUM, let the input data be X, the true output be y, and the output predicted by the model be ω1 is the weight matrix connecting the input layer and the hidden layer, e1 is the bias connecting the input layer and the hidden layer, ω2 is the weight matrix connecting the output layer and the hidden layer, and e2 is the bias connecting the output layer and the hidden layer. The optimization strategy adopted is: first optimize the network parameter initialization forward propagation to find the optimal initialization parameters, and finally optimize it by back propagation. This optimization strategy network prediction speed is very fast, so it can be applied to real-time prediction. The specific content is shown in the following steps:

[0108] S1. Set the population size and randomly generate the initial population. The dimension of the individuals in the population represents the number of parameters that need to be optimized in the BP neural network. Set the population size to N p , the target number of iterations is N m Second, randomly generate the initial value of the population;

[0109] S2. Set the crossover probability to P c ;

[0110] S3. The training set S i Input into the BP neural network with randomly generated parameters, and get the output value of each individual according to the forward calculation of different BP neural networks

[0111] S4. Let the error be equal to the fitness, calculate the fitness value of each individual, and sort the fitness values ​​obtained for each individual in descending order.

[0112] S5. Calculate the vector distance and fitness value difference between the optimal solutions of adjacent populations in all populations. When the vector distance and fitness value difference meet the corresponding conditions, migration communication between sub-populations is carried out, so that the local worst solution and the local optimal solution between adjacent populations can communicate and learn.

[0113] S6. Use the roulette wheel method to select new individuals from the parent chromosomes. The greater the fitness, the greater the probability of being selected and thus serving as the next generation population.

[0114] S7. Perform crossover and mutation on the selected individuals according to the given crossover and mutation probabilities.

[0115] S8. When the target number of iterations N is not reached m When , it returns to S4. Finally, when the number of iterations reaches the maximum, the local optimal solutions of the remaining two groups are compared, and the one with the lowest fitness value is taken as the global optimal solution.

[0116] S9. Obtain the global optimal solution, including the optimal solution of the number of neurons, the number of network layers, the initial weights, and the initial threshold, and use the obtained optimal solution parameters as the setting parameters of the BP neural network, and use the training set as the input of the model.

[0117] S10. Calculate the output of the hidden layer.

[0118] S11. Perform output calculation of the output layer.

[0119] S12. Error calculation.

[0120] S13. Record the number of samples that have been learned, and update the weights and thresholds.

[0121] S14. Determine whether the iteration is completed. If not, return to S13 to continue iterating until the number of iterations is completed and the chromosome with the highest fitness is selected as the final result, thereby obtaining a trained genetic algorithm-optimized BP neural network long-term vehicle speed prediction model.

[0122] Among them, the fitness value in S4 can be expressed as: The calculation of the output layer in S11 can be: Where A1 = tansig(ω1*X+e1).

[0123] The parameters of the neural network prediction model example of the embodiment are as follows: the size of the intercepted original training data set is, for example, 2278×9, corresponding to taking the first 70% as the training set and the last 30% as the test set. Figure 4 The following is a fitness value change curve of an example; Figure 5 Shown is a comparison of the test set and predicted vehicle speed in the example.

[0124] Step S24, perform online prediction and optimization. Taking the distance length when the main vehicle and the first vehicle in the team establish V2V communication as the standard, define the main vehicle number in the signal area as Veh_host, the first vehicle number as Veh_1, and the vehicle in front of the main vehicle as Veh_N. If the vehicle in front of the main vehicle initially defined leaves the current lane, the vehicle in front of the main vehicle will be updated forward, and the total number of vehicles in front of the vehicle will be controlled to remain N; similarly, if the vehicle in front of the team initially defined leaves, the vehicle in front of the team will be updated forward, so as to control the total number of vehicles in front of the vehicle to remain N. It can be seen that the total number of vehicles in the lane where the data collection vehicle Veh_host is located is N. Since each vehicle is driven by a human driver and the vehicle models are different, the driver's reaction time t i There is uncertainty. Here, for example, the average reaction time of ordinary drivers is 2s. Therefore, the total time delay from Veh_1 starting to the vehicle in front of the main vehicle Veh_N starting can be expressed as:

[0125] Based on this, the starting time can be defined as the time when the main vehicle and the first vehicle in the team establish V2V communication And from that moment on, record the next N vehicles in front of the main vehicle The speed sequence of seconds is denoted by Will As the training set, it is input into the trained neural network model for prediction. Through this prediction model, we can get Future vehicle speed information At the same time, the time when prediction starts is recorded as the starting time of following the vehicle T0, and the real-time speed sequence of the leading vehicle and the preceding vehicle with a duration of τ is recorded from this time: R N =[v N (T0),v N (T1),v N (T2),…,v N (T τ-1 )] T ,in

[0126] In the above process, the future predicted vehicle speed is obtained by using the BP neural network based on GA optimization. Although the predicted results reflect the traffic flow characteristics well, due to the different models and drivers of the vehicles in front of the main vehicle, there is a large uncertainty in the reaction time, so there is a temporal and spatial misalignment error, which leads to a large temporal and spatial difference between the predicted speed and the actual speed. Among them, the specific steps to eliminate the temporal and spatial errors can be as follows:

[0127] First, define the following three parameter features:

[0128] Vehicle speed matching positioning features: Acceleration features:

[0129] Impact characteristics:

[0130] Turning point characteristics:

[0131]

[0132] Among them, the turning point characteristic condition 1 is: at the i-th moment, the product of the accelerations at moments i-1 and i+1 is greater than zero, and the total number of moments that meet condition 1 is recorded as m1; the turning point characteristic condition 2 is: at the i-th moment, the product of the accelerations at moments i-1 and i+1 is less than or equal to zero, and the total number of moments that meet condition 2 is recorded as m2.

[0133] Secondly, construct an optimization problem: in order to obtain the best time misalignment error compensation value, a simple optimization algorithm is established, including but not limited to particle swarm algorithm, genetic algorithm, ant colony algorithm, etc., and optionally, an optimization algorithm such as genetic algorithm, particle swarm algorithm, etc. is established.

[0134] The three positioning matching errors are defined as follows:

[0135] Acceleration error: δ A (i) = A real (i)-A predict (i);

[0136] Impact error: δ G (i) = G real (i)-G predict (i);

[0137] Feature point error: δ P (i) = -P(i).

[0138] Then, the data can be normalized using the maximum and minimum principle.

[0139] The objective function is:

[0140] It should be added that the above objective function can also be expressed as: Among them, ρ1, ρ2, and ρ3 are the weights corresponding to the acceleration error (value), the impact error (value), and the feature point error (value), respectively. The weight values ​​can be adaptively allocated according to the current road congestion level. If the road is very congested, acceleration and deceleration are frequent, and the acceleration and impact weights can be appropriately increased.

[0141] Finally, the optimization problem is solved: The above optimization problem can be described as follows. The solution goal is to obtain a suitable input value u of the time misalignment error compensation value, so as to minimize the matching positioning error J. Here, the time and space misalignment error between the predicted speed and the actual speed is defined as u∈[u min ,u max ], where u min and u max The minimum and maximum spatiotemporal misalignment errors respectively come from the comparison between the test set and the prediction results during the model training process, that is, they correspond to the minimum and maximum values ​​in the error values ​​described in step 204 respectively.

[0142] Thus, the optimal time misalignment error compensation value (preset compensation value) u can be obtained through the optimization algorithm. best , add the time coordinate of the predicted vehicle speed to the time offset error compensation value u best The temporal and spatial consistency of the predicted vehicle speed and the actual vehicle speed can be ensured, thereby reducing the temporal and spatial errors caused by the uncertainty of the driver and vehicle model of the main vehicle, improving the prediction accuracy of the long-term vehicle speed, and realizing the long-term prediction of the vehicle speed in front.

[0143] It can be seen that by using V2V technology to obtain driving data information in a fixed period of time, training the BP neural network to establish a BP neural network long-term vehicle speed prediction model based on GA optimization, and then using feature matching parameters to match and locate the predicted vehicle speed to eliminate the temporal and spatial misalignment error, it finally achieves the long-term prediction of the future speed of the vehicle in front of the target lane. It provides more accurate future long-term vehicle speed information for the vehicle control system, and provides a basis for improving the energy efficiency, safety and comfort performance of vehicles in congested sections. In addition, the predicted speed is optimized using feature matching parameters, which reduces the temporal and spatial misalignment error caused by the uncertainty of the driver and vehicle model of the vehicle in front, and improves the accuracy of the long-term prediction of the speed of the vehicle in front.

[0144] It should be understood that, although the various steps in the flowcharts involved in the above-mentioned embodiments are displayed in sequence according to the indication of the arrows, these steps are not necessarily executed in sequence according to the order indicated by the arrows. Unless there is a clear explanation in this article, the execution of these steps does not have a strict order restriction, and these steps can be executed in other orders. Moreover, at least a part of the steps in the flowcharts involved in the above-mentioned embodiments can include multiple steps or multiple stages, and these steps or stages are not necessarily executed at the same time, but can be executed at different times, and the execution order of these steps or stages is not necessarily to be carried out in sequence, but can be executed in turn or alternately with other steps or at least a part of the steps or stages in other steps.

[0145] Based on the same inventive concept, the embodiment of the present application also provides a device for predicting the speed of a leading vehicle in a following convoy for implementing the method for predicting the speed of a leading vehicle in a following convoy mentioned above. The implementation scheme for solving the problem provided by the device is similar to the implementation scheme recorded in the above method, so the specific limitations in the embodiments of one or more devices for predicting the speed of a leading vehicle in a following convoy provided below can refer to the limitations of the method for predicting the speed of a leading vehicle in a following convoy above, and will not be repeated here.

[0146] In an exemplary embodiment, Figure 6 As shown, a device 300 for predicting the speed of a leading vehicle in a following vehicle group is provided, comprising: a data acquisition module 81, a data processing module 82 and a result correction module 83, wherein:

[0147] The data acquisition module 81 is used to acquire the real-time vehicle speed of each of the at least two preceding vehicles within a period of time when the vehicle establishes communication with the at least two preceding vehicles in the preceding direction, and obtain a real-time vehicle speed set;

[0148] The data processing module 82 is used to input the real-time vehicle speed set into the pre-trained vehicle speed prediction model to obtain a first prediction result of the neighboring vehicle speed output by the vehicle speed prediction model; wherein the first prediction result of the neighboring vehicle speed includes a plurality of preliminary prediction speeds of the neighboring vehicle in the future period; the neighboring vehicle is the vehicle in front of the vehicle in the driving direction and closest to the vehicle;

[0149] The result correction module 83 is used to obtain a preset compensation value, process the first prediction result of the neighboring vehicle speed based on the preset compensation value, and obtain the second prediction result of the neighboring vehicle speed; wherein the preset compensation value is determined based on the historical acceleration error value, the historical impact error value and the historical feature point error value corresponding to the historical following fleet.

[0150] In an exemplary embodiment, the result correction module 83 also includes: obtaining the historical speed of each of the at least two leading vehicles communicating with the vehicle within a historical period of time to obtain a historical speed set; inputting the historical speed set into a pre-trained speed prediction model to obtain the historical prediction result of the neighboring vehicle speed output by the speed prediction model; obtaining the historical actual speed of the neighboring vehicle associated with multiple historical initial prediction speeds in the historical prediction result of the neighboring vehicle speed, and calculating the error value between the historical actual speed of each neighboring vehicle and the corresponding historical initial prediction speed; calling the objective function to process each historical acceleration error value, each historical impact error value and historical feature point error value to obtain a matching error; processing the matching error and error value through the optimal solution algorithm to determine a preset compensation value. For details, please continue to refer to Figure 1 and Figure 2 , and the above Figure 1 and Figure 2 Description.

[0151] In an exemplary embodiment, the result correction module 83 further includes: performing differential processing on the historical actual speed of each neighboring vehicle to obtain the corresponding first acceleration characteristics; performing differential processing on the historical initial predicted speed corresponding to the historical actual speed of each neighboring vehicle to obtain the corresponding second acceleration characteristics; based on the first acceleration characteristics and the second acceleration characteristics, determining the historical acceleration error value corresponding to the historical actual speed of each neighboring vehicle. For details, please continue to refer to Figure 1 and Figure 2 , and the above Figure 1 and Figure 2 Description.

[0152] In an exemplary embodiment, after obtaining each first acceleration feature and each second acceleration feature, the result correction module 83 further includes: performing differential processing on each first acceleration feature and each second acceleration feature to obtain corresponding each first impact degree feature and each second impact degree feature; based on each first impact degree feature and each second impact degree feature, determining the historical impact degree error value corresponding to the historical actual vehicle speed of each neighboring vehicle. For details, please continue to refer to Figure 1 and Figure 2 , and the above Figure 1 and Figure 2 Description.

[0153] In an exemplary embodiment, the result correction module 83 is used to obtain the historical actual speed of the neighboring vehicle associated with multiple historical initial predicted speeds in the historical prediction results of the neighboring vehicle speed, including: obtaining the historical actual speed of the neighboring vehicle associated with K historical initial predicted speeds in the historical prediction results of the neighboring vehicle speed; wherein K ≥ 3 and K is an integer; after obtaining each first acceleration feature and each second acceleration feature, it also includes: obtaining the product of the first acceleration feature at the previous moment and the first acceleration feature at the next moment corresponding to each of the 2nd to K-1th first acceleration features, and obtaining a first number of positive numbers, and a second number of zero and negative numbers in the product; determining the historical turning point feature based on the first number and the second number, and taking the opposite of the historical turning point feature to obtain the corresponding historical feature point error value. For details, please continue to refer to Figure 1 and Figure 2 , and the above Figure 1 and Figure 2 Description.

[0154] In an exemplary embodiment, the result correction module 83 is used to process the first prediction result of the neighboring vehicle speed based on the preset compensation value to obtain the second prediction result of the neighboring vehicle speed, including: adding each preliminary prediction speed in the first prediction result of the neighboring vehicle speed to the preset compensation value to obtain the second prediction result of the neighboring vehicle speed. Figure 1 and Figure 2 , and the above Figure 1 and Figure 2 Description.

[0155] In an exemplary embodiment, the data processing module 82 further includes: training a GA-BP neural network model based on the historical vehicle speed data of the historical following fleet to obtain a pre-trained vehicle speed prediction model. Figure 1 and Figure 2 , and the above Figure 1 and Figure 2 Description.

[0156] Each module in the above-mentioned prediction of the speed of the leading vehicle in the following convoy can be implemented in whole or in part by software, hardware or a combination thereof. Each of the above-mentioned modules can be embedded in or independent of the processor in the vehicle-side control device in the form of hardware, or can be stored in the memory in the vehicle-side control device in the form of software, so that the processor can call and execute the corresponding operations of each of the above modules.

[0157] Figure 7 FIG. 1 is an internal structure diagram of a vehicle-side control device in an embodiment. In an exemplary embodiment, a vehicle-side control device is provided. The internal structure diagram of the vehicle-side control device can be as follows: Figure 7 As shown. The vehicle-side control device includes a processor and a memory. The processor of the vehicle-side control device is used to provide computing and control capabilities. The memory of the vehicle-side control device includes a non-volatile storage medium, and the non-volatile storage medium stores a computer program. When the computer program is executed by the processor, a method for predicting the speed of a leading vehicle in a following convoy is implemented. The method for predicting the speed of a leading vehicle in a following convoy is any method for predicting the speed of a leading vehicle in a following convoy mentioned in the embodiments of the present application. For relevant embodiments, please refer to the above.

[0158] Those skilled in the art will understand that Figure 7 The structure shown in the figure is merely a block diagram of a partial structure related to the scheme of the present application, and does not constitute a limitation on the vehicle-side control device to which the scheme of the present application is applied. The specific vehicle-side control device may include more or fewer components than those shown in the figure, or combine certain components, or have a different component arrangement.

[0159] Based on the same inventive concept, the present application also provides a computer-readable storage medium on which a computer program is stored. When the computer program is executed by a processor, the aforementioned method for predicting the speed of a leading vehicle in a following convoy is implemented. The method for predicting the speed of a leading vehicle in a following convoy is any one of the methods for predicting the speed of a leading vehicle in a following convoy mentioned in the embodiments of the present application. Relevant embodiments can be found above.

[0160] Based on the same inventive concept, the present application also provides a computer program product, including a computer program, which, when executed by a processor, implements the aforementioned method for predicting the speed of a leading vehicle in a following convoy. The method for predicting the speed of a leading vehicle in a following convoy is any one of the methods for predicting the speed of a leading vehicle in a following convoy mentioned in the embodiments of the present application. Relevant embodiments can be found above.

[0161] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, stored data, displayed data, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of relevant data must comply with relevant regulations.

[0162] Those of ordinary skill in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to the memory, database or other medium used in the embodiments provided in the present application can include at least one of non-volatile memory and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. As an illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM). The database involved in each embodiment provided in this application may include at least one of a relational database and a non-relational database. Non-relational databases may include distributed databases based on blockchains, etc., but are not limited thereto. The processor involved in each embodiment provided in this application may be a general-purpose processor, a central processing unit, a graphics processor, a digital signal processor, a programmable logic device, a data processing logic device based on quantum computing, an artificial intelligence (AI) processor, etc., but are not limited thereto.

[0163] The technical features of the above embodiments may be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this application.

[0164] The above-described embodiments only express several implementation methods of the present application, and the descriptions thereof are relatively specific and detailed, but they cannot be construed as limiting the scope of the present application. It should be noted that, for a person of ordinary skill in the art, several modifications and improvements can be made without departing from the concept of the present application, and these all belong to the protection scope of the present application. Therefore, the protection scope of the present application shall be subject to the attached claims.

Claims

1. A method for predicting the speed of a leading vehicle in a convoy, characterized in that: include: When the vehicle establishes communication with at least two preceding vehicles in the direction of travel of the vehicle, the real-time vehicle speed of each of the at least two preceding vehicles within a period of time is obtained to obtain a real-time vehicle speed set; The real-time vehicle speed set is input into a pre-trained vehicle speed prediction model to obtain a first prediction result of the neighboring vehicle speed output by the vehicle speed prediction model; wherein the first prediction result of the neighboring vehicle speed includes a plurality of preliminary predicted vehicle speeds of the neighboring vehicle in a future time period; the neighboring vehicle is the preceding vehicle that is closest to the own vehicle in the driving direction of the own vehicle; Obtain a preset compensation value, and process the first prediction result of the neighboring vehicle speed based on the preset compensation value to obtain a second prediction result of the neighboring vehicle speed; wherein the preset compensation value is determined based on a historical acceleration error value, a historical impact error value, and a historical feature point error value corresponding to a historical following fleet.

2. The method according to claim 1, characterized in that Also includes: Obtaining a historical prediction result of the speed of an adjacent vehicle corresponding to a historical speed set; the historical speed set includes the historical speeds of each of the at least two preceding vehicles communicating with the vehicle within a period of historical time; the historical prediction result of the speed of an adjacent vehicle is obtained by using the output of the pre-trained speed prediction model with the historical speed set as a model input; Obtaining historical actual vehicle speeds of neighboring vehicles associated with a plurality of historical initial predicted vehicle speeds in the historical prediction results of the neighboring vehicle speeds, and calculating an error value between each of the historical actual vehicle speeds of the neighboring vehicles and the corresponding historical initial predicted vehicle speeds; Calling the objective function to process each of the historical acceleration error values, each of the historical impact error values, and the historical feature point error values ​​to obtain a minimum matching error in the matching error; The matching error minimum value, and the maximum and minimum values ​​of the error values ​​are processed by an optimal solution algorithm to determine the preset compensation value.

3. The method according to claim 2, characterized in that Also includes: Performing differentiation processing on the historical actual vehicle speeds of each neighboring vehicle to obtain corresponding first acceleration characteristics; Performing differentiation processing on the historical initial predicted vehicle speed corresponding to the historical actual vehicle speed of each neighboring vehicle to obtain corresponding second acceleration characteristics; Based on each of the first acceleration characteristics and the second acceleration characteristics, a historical acceleration error value corresponding to each of the neighboring vehicles' historical actual vehicle speeds is determined.

4. The method according to claim 3, characterized in that After obtaining the first acceleration characteristics and the second acceleration characteristics, the method further includes: Performing differentiation processing on each of the first acceleration characteristics and each of the second acceleration characteristics to obtain corresponding each of the first impact degree characteristics and each of the second impact degree characteristics; Based on each of the first impact degree characteristics and each of the second impact degree characteristics, a historical impact degree error value corresponding to the historical actual vehicle speed of each of the neighboring vehicles is determined.

5. The method according to claim 3, characterized in that: The step of obtaining the historical actual vehicle speed of the neighboring vehicle associated with the plurality of historical initial predicted vehicle speeds in the historical prediction result of the neighboring vehicle speed comprises: obtaining the historical actual vehicle speed of the neighboring vehicle associated with K historical initial predicted vehicle speeds in the historical prediction result of the neighboring vehicle speed; wherein K≥3 and K is an integer; After obtaining the first acceleration characteristics and the second acceleration characteristics, the method further includes: Obtaining the product of the first acceleration feature at a previous moment and the first acceleration feature at a next moment, respectively corresponding to the second to K-1th first acceleration features, and obtaining a first number of positive numbers, and a second number of zero and negative numbers in the product; A historical turning point feature is determined based on the first number and the second number, and the historical turning point feature is negated to obtain a corresponding historical feature point error value.

6. The method according to any one of claims 1 to 5, characterized in that: The step of processing the first prediction result of the adjacent vehicle speed based on the preset compensation value to obtain the second prediction result of the adjacent vehicle speed includes: Each of the preliminary predicted vehicle speeds in the first prediction result of the neighboring vehicle speed is added to the preset compensation value to obtain a second prediction result of the neighboring vehicle speed.

7. The method according to any one of claims 1 to 5, characterized in that: Also includes: The GA-BP neural network model is trained based on the historical vehicle speed data of the historical car-following fleet to obtain a pre-trained vehicle speed prediction model.

8. A device for predicting the speed of a leading vehicle in a convoy, characterized in that: The device comprises: A data acquisition module, for acquiring the real-time vehicle speed of each of the at least two preceding vehicles within a period of time when the vehicle establishes communication with the at least two preceding vehicles in the preceding direction, and obtaining a real-time vehicle speed set; A data processing module is used to input the real-time vehicle speed set into a pre-trained vehicle speed prediction model to obtain a first prediction result of the neighboring vehicle speed output by the vehicle speed prediction model; wherein the first prediction result of the neighboring vehicle speed includes a plurality of preliminary predicted vehicle speeds of the neighboring vehicle in a future period; the neighboring vehicle is the preceding vehicle that is closest to the own vehicle in the driving direction of the own vehicle; The result correction module is used to obtain a preset compensation value, process the first prediction result of the neighboring vehicle speed based on the preset compensation value, and obtain a second prediction result of the neighboring vehicle speed; wherein the preset compensation value is determined based on the historical acceleration error value, historical impact error value and historical feature point error value corresponding to the historical following fleet.

9. A vehicle-side control device, comprising a memory and a processor, wherein the memory stores a computer program, characterized in that: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 7 are implemented.

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

Images