Methods for identifying road bump types, vehicle control methods, and related devices

Abstract

This application discloses a method for identifying road bump types, a vehicle control method, and related devices. The method includes: acquiring real-time driving information of a vehicle; determining whether the vehicle is in a bumpy state based on the driving information; determining that the vehicle is traveling on a bumpy road surface when it is in a bumpy state; triggering the collection of a preset dataset when the vehicle is on a bumpy road surface; the preset dataset includes: road surface image information and vehicle driving information; identifying the road bump type using an image recognition algorithm, a neural network recognition algorithm, and a dynamic time planning algorithm based on the collected preset dataset; fusing the identification results of the image recognition algorithm, neural network recognition algorithm, and dynamic time planning algorithm, and using the fused result as the road bump type. Using this application, the road bump type can be identified efficiently and accurately.

Description

Technical Field

[0001] This application relates to the field of automotive technology, specifically to a method for identifying road bump types, a vehicle control method, and related devices. Background Technology

[0002] Vehicles encounter various road surfaces while driving. When the road has undulations, potholes, or other uneven surfaces, it will cause the vehicle to bounce. Bumps will affect driving comfort and pose safety hazards.

[0003] Identifying road bumps and their types allows for better vehicle control, improving driving safety and comfort. Currently, bump identification is primarily done manually, which is inefficient and unreliable.

[0004] Therefore, how to efficiently and accurately identify the types of road bumps is a technical problem that urgently needs to be solved. Summary of the Invention

[0005] This application provides a method for identifying road bump types, a vehicle control method, and related devices, which can efficiently and accurately identify road bump types.

[0006] In a first aspect, embodiments of this application provide a method for identifying the type of road bumps, the method comprising:

[0007] Obtain real-time driving information of the vehicle while it is in motion;

[0008] Determine whether the vehicle is in a bumpy state based on the driving information;

[0009] When the vehicle is in a bumpy state, it is determined that the vehicle is traveling on a bumpy road surface;

[0010] When the vehicle is on a bumpy road, a preset dataset is triggered for collection; the preset dataset includes: road surface image information and vehicle driving information;

[0011] Based on the collected preset dataset, the types of road bumps are identified using image recognition algorithms, neural network recognition algorithms, and Dynamic Time Wrapping (DTW) algorithms, respectively.

[0012] The recognition results of the image recognition algorithm, neural network recognition algorithm, and dynamic time planning algorithm for identifying the road bump type are combined, and the combined result is used as the road bump type.

[0013] In this embodiment, the vehicle's condition is first determined based on real-time driving information. If the vehicle is in a bumpy state, a pre-set dataset is collected. Based on the collected dataset, image recognition algorithms, neural network recognition algorithms, and dynamic time planning algorithms are used to identify the road bump type. Finally, the identification results from these three algorithms are merged, and the merged result is taken as the road bump type. Compared with manual identification in the prior art, this embodiment identifies road bump types more promptly. Furthermore, because the identification results of three different algorithms are merged, this embodiment identifies road bump types more accurately.

[0014] In some possible implementations, the step of identifying the road bump type using image recognition algorithms, neural network recognition algorithms, and dynamic time planning algorithms based on the collected preset dataset includes:

[0015] Based on the pre-set dataset, the confidence levels of various possible bump types of the road are identified using image recognition algorithms, neural network recognition algorithms, and dynamic time planning algorithms, respectively.

[0016] In some possible implementations, the fusion of the recognition results from the image recognition algorithm, neural network recognition algorithm, and dynamic time planning algorithm in identifying the road bump type, and the fused result is used as the road bump type; including:

[0017] The bump type with the highest matching degree after fusion is taken as the bump type of the road; the matching degree corresponding to any bump type i is Mi.

[0018] M i =α i A i +β i B i +γ i C i

[0019] Where, α i A represents the confidence weight of the image recognition algorithm for bump type i; i β represents the matching degree of the image recognition algorithm for bump type i; i B is the confidence weight assigned to bump type i by the neural network algorithm; i γ represents the matching degree of the neural network algorithm for identifying bump type i; i C represents the confidence weight of turbulence type i in the DTW algorithm. i Let be the matching degree of the dynamic time planning algorithm for identifying bump type i; where,

[0020]

[0021]

[0022]

[0023] Where a1 is the illumination condition influence factor of the image recognition algorithm, 0≤a1≤1; a2 is the visual output sharpness influence factor of the image recognition algorithm, with a value range of 0≤a2≤1; W i1 b1 is the initial confidence weight assigned to bump type i by the image recognition algorithm; b2 is the influence factor of neural network input data quality, 0≤b1≤1; b2 is the influence factor of neural network algorithm label data quality, with a value range of 0≤b2≤1; W i2 Here, c represents the initial confidence weight assigned to turbulence type i by the neural network algorithm; c is the standard turbulence data quality impact factor of the DTW algorithm, with a value range of 0 ≤ c ≤ 1; W i3 The initial confidence weights are assigned to bump type i in the dynamic time planning algorithm.

[0024] Among the possible implementations are:

[0025] Obtain location information;

[0026] The location information and the road bump type are uploaded to the server.

[0027] In some possible implementations, the road bump type includes one or more of the following types: bridge access, bridge exit, speed bumps, potholes, small undulations, medium undulations, large undulations, and continuous bumpy sections.

[0028] Secondly, this application provides a vehicle control method, including:

[0029] The type of road bumps can be identified using any of the technical solutions described in the first aspect;

[0030] The suspension is controlled to match its state with the identification result, allowing the vehicle to smoothly pass through bumpy road sections.

[0031] Among the possible implementations are:

[0032] The system obtains bump information from the server, which is shared by other vehicle owners, within a preset range of the current location of the vehicle. The bump information includes: location information and bump type.

[0033] Based on the vehicle's real-time driving information and the bump information, the type of bumps in the upcoming road section is predicted.

[0034] The suspension is controlled to match the predicted bump type of the upcoming road segment.

[0035] Thirdly, this application provides a road bump type identification device, the device comprising:

[0036] The acquisition unit is used to acquire real-time driving information of the vehicle while it is in motion;

[0037] The first determining unit is configured to determine whether the vehicle is in a bumpy state based on the driving information; and to determine that the vehicle is traveling on a bumpy road surface when the vehicle is in a bumpy state.

[0038] The acquisition unit is further configured to trigger the collection of a preset dataset when the vehicle is on a bumpy road surface; the preset dataset includes: road surface image information and vehicle driving information;

[0039] The identification unit is used to identify the type of road bumps based on the preset dataset collected by the acquisition unit, using image recognition algorithm, neural network recognition algorithm, and dynamic time planning algorithm respectively.

[0040] The fusion unit is used to fuse the recognition results of the image recognition algorithm, the neural network recognition algorithm, and the dynamic time planning algorithm respectively in identifying the road bump type, and use the fused result as the road bump type.

[0041] Fourthly, this application provides a processing device, including: a memory and a processor, wherein the memory stores a computer program executable by the processor, and the processor executes the computer program to implement any one of the methods described in the first aspect and / or any one of the methods described in the second aspect.

[0042] Fifthly, this application provides a computer-readable storage medium storing a computer program, which, when read and executed, implements any one of the methods described in the first aspect and / or any one of the methods described in the second aspect.

[0043] It should be understood that the technical features of the technical solutions provided in the second to fifth aspects can all be corresponding to the methods for determining wake-up time provided in the first aspect and its possible implementations. Therefore, the beneficial effects that can be achieved are similar, and will not be elaborated here. Attached Figure Description

[0044] Figure 1A This is a schematic flowchart of a method for identifying road bump types according to an embodiment of this application;

[0045] Figure 1B This is a flowchart illustrating a method for identifying road bump types according to another embodiment of this application;

[0046] Figure 2A This is a schematic flowchart of a vehicle control method provided in an embodiment of this application;

[0047] Figure 2B This is a schematic flowchart of a vehicle control method provided in another embodiment of this application;

[0048] Figure 2C This is a schematic flowchart of a vehicle control method provided in another embodiment of this application;

[0049] Figure 3 This is a schematic diagram of the structure of a road bump type identification device provided in an embodiment of this application;

[0050] Figure 4 This is a schematic diagram of the structure of a processing device provided in an embodiment of this application. Detailed Implementation

[0051] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0052] Figure 1A This is a flowchart illustrating the method for identifying road bump types provided in an embodiment of this application. Figure 1A As shown in the embodiment of this application, the method for identifying the type of road bumps includes steps 101 to 106. Steps 101 to 106 will be described in detail below.

[0053] 101. Obtain real-time driving information while the vehicle is in motion.

[0054] The driving information may include indicators such as wheel acceleration, vehicle acceleration, suspension height, or suspension height change rate.

[0055] 102. Determine whether the vehicle is in a bumpy condition based on the driving information.

[0056] In some possible implementations, the acquired driving data can be preprocessed to obtain parameter values ​​such as wheel acceleration, vehicle acceleration, suspension height, vehicle pitch angle, vehicle roll angle, vehicle speed, or slope information. Furthermore, through feature engineering, feature information including mean, standard deviation, integral, eccentricity, extreme values, variance, or power density can be extracted. Specifically, for changes in suspension height, vibration waveforms of bumpy road sections can be generated as feature graphics using methods such as filtering, integration, and Fourier transform.

[0057] Initial turbulence identification based on feature reasoning uses an expert system and compares against thresholds of set parameters to screen for the presence or absence of turbulence events. In some possible implementations, further preliminary identification of turbulence events can be made based on signals such as the amplitude, frequency, shape, and impact variation of characteristic turbulence vibration waveforms, yielding preliminary identification results. It can also preliminarily determine the turbulence category, such as bridge deck turbulence, undulating turbulence, or continuous turbulence.

[0058] 103. When the vehicle is in a bumpy state, determine that the vehicle is traveling on a bumpy road surface.

[0059] 104. When the vehicle is on a bumpy road, trigger the collection of a preset dataset; the preset dataset includes: road surface image information and vehicle driving information.

[0060] 105. Based on the collected preset dataset, use image recognition algorithm, neural network recognition algorithm, and dynamic time planning algorithm to identify the bump type of the road respectively.

[0061] The preset dataset can include multiple parameters from the following: wheel acceleration, vehicle acceleration, suspension height, or suspension height change rate, etc.

[0062] The types of bumps on a road can include one or more of the following: going up a bridge, going down a bridge, speed bumps, potholes, small undulations, medium undulations, large undulations, and continuous bumpy sections.

[0063] Image recognition algorithms can process road images acquired by vehicle cameras, such as performing standardized preprocessing on the images, including cropping, noise reduction, and grayscale conversion, as well as matching the processed image signals with bump types to obtain the matching degree results between the current image and various bump types.

[0064] The neural network recognition algorithm identifies the type of bump based on the input bump feature data. The input feature data may include suspension height, vibration waveform amplitude and frequency, wheel acceleration, vehicle acceleration, vehicle pitch angle, vehicle roll angle, vehicle speed or slope, etc. The output result is the matching degree of various bump types for the current road segment.

[0065] The DTW algorithm performs time-series similarity matching between bumpy segments and bump feature data from a standard set of bump types. The matching data includes suspension height, vibration waveform amplitude and frequency, wheel acceleration, and vehicle vertical acceleration. Before matching, the data is normalized to linearly map it to a reasonable range, and adaptive deformations are performed based on vehicle speed, acceleration, etc., to align the data across time series. Finally, the similarity matching results for various bump types corresponding to the current bumpy segment are obtained.

[0066] It should be noted that while image recognition can capture bumps related to road morphology, it cannot obtain effective image information in adverse weather conditions. Neural network models are highly adaptable to input data, but require a large amount of labeled data for training. The DTW algorithm takes into account the changes in time-series data, but its adaptability is poor in the absence of significant time signals.

[0067] In some possible implementations, based on a pre-collected dataset, image recognition algorithms, neural network recognition algorithms, and dynamic time planning algorithms are used to identify the types of road bumps, including: based on a pre-collected dataset, using image recognition algorithms, neural network recognition algorithms, and dynamic time planning algorithms to identify the confidence levels of various possible types of road bumps.

[0068] 106. The recognition results obtained by the image recognition algorithm, neural network recognition algorithm and dynamic time planning algorithm for identifying the road bump type are fused together, and the fused result is taken as the road bump type.

[0069] In some possible implementations, the recognition results of image recognition algorithms, neural network recognition algorithms, and dynamic time planning algorithms for identifying road bump types are fused, and the fused result is used as the road bump type; including:

[0070] The bump type with the highest matching degree after fusion is taken as the road bump type; the matching degree corresponding to any bump type i is Mi.

[0071] Mi=αiAi+βiBi+γiCi

[0072] Where αi is the confidence weight of the image recognition algorithm for bump type i; Ai is the recognition matching degree of the image recognition algorithm for bump type i; βi is the confidence weight of the neural network algorithm for bump type i; Bi is the recognition matching degree of the neural network algorithm for bump type i; γi is the confidence weight of the DTW algorithm for bump type i; Ci is the recognition matching degree of the dynamic time planning algorithm for bump type i; and so on.

[0073]

[0074]

[0075]

[0076] Where a1 is the illumination condition influence factor of the image recognition algorithm, 0≤a1≤1; a2 is the visual output sharpness influence factor of the image recognition algorithm, with a value range of 0≤a2≤1; W i1b1 is the initial confidence weight assigned to bump type i by the image recognition algorithm; b2 is the influence factor of neural network input data quality, 0≤b1≤1; b2 is the influence factor of neural network algorithm label data quality, with a value range of 0≤b2≤1; W i2 Here, c represents the initial confidence weight assigned to turbulence type i by the neural network algorithm; c is the standard turbulence data quality impact factor of the DTW algorithm, with a value range of 0 ≤ c ≤ 1; W i3 The initial confidence weights are assigned to bump type i in the dynamic time planning algorithm.

[0077] For example, in a bump identification process, if the top two bump types in the image recognition results are "downhill" and "large undulation," the top two bump types in the neural network recognition results are "large undulation" and "downhill," and the top two bump types in the DTW recognition results are "downhill" and "large undulation," then because the image recognition algorithm has higher accuracy in identifying downhill bumps, it receives a larger initial confidence weight for this type. Finally, after weighted fusion of the results from the three algorithms, downhill bumps receive the highest matching score, resulting in the final identification result being "downhill." To improve recognition accuracy, in a given identification process, if the difference in matching scores between the top two bump types after fusion is less than a certain threshold, the identification is considered invalid and not included in the final result. Since different algorithms have varying accuracies in identifying different bump types, fusion can reduce the risk of misidentification caused by a single algorithm and improve the overall accuracy of bump identification.

[0078] In this embodiment, the vehicle's condition is first determined based on real-time driving information. If the vehicle is in a bumpy state, a pre-set dataset is collected. Based on the collected dataset, image recognition algorithms, neural network recognition algorithms, and dynamic time planning algorithms are used to identify the road bump type. Finally, the identification results from these three algorithms are merged, and the merged result is taken as the road bump type. Compared with manual identification in the prior art, this embodiment identifies road bump types more promptly. Furthermore, because the identification results of three different algorithms are merged, this embodiment identifies road bump types more accurately.

[0079] Please see Figure 1B , Figure 1B This is a flowchart illustrating a method for identifying road bump types according to another embodiment of this application. Figure 1B As shown, the method for identifying road bump types provided in this application includes steps 101 to 108. Steps 101 to 108 will be described in detail below.

[0080] 101. Obtain real-time driving information while the vehicle is in motion.

[0081] The driving information may include indicators such as wheel acceleration, vehicle acceleration, suspension height, or suspension height change rate.

[0082] 102. Determine whether the vehicle is in a bumpy condition based on the driving information.

[0083] In some possible implementations, the acquired driving data can be preprocessed to obtain parameter values ​​such as wheel acceleration, vehicle acceleration, suspension height, vehicle pitch angle, vehicle roll angle, vehicle speed, or slope information. Furthermore, through feature engineering, feature information including mean, standard deviation, integral, eccentricity, extreme values, variance, or power density can be extracted. Specifically, for changes in suspension height, vibration waveforms of bumpy road sections can be generated as feature graphics using methods such as filtering, integration, and Fourier transform.

[0084] Initial turbulence identification based on feature reasoning uses an expert system and compares against thresholds of set parameters to screen for the presence or absence of turbulence events. In some possible implementations, further preliminary identification of turbulence events can be made based on signals such as the amplitude, frequency, shape, and impact variation of characteristic turbulence vibration waveforms, yielding preliminary identification results. It can also preliminarily determine the turbulence category, such as bridge deck turbulence, undulating turbulence, or continuous turbulence.

[0085] 103. When the vehicle is in a bumpy state, determine that the vehicle is traveling on a bumpy road surface.

[0086] 104. When the vehicle is on a bumpy road, trigger the collection of a preset dataset; the preset dataset includes: road surface image information and vehicle driving information.

[0087] 105. Based on the collected preset dataset, use image recognition algorithm, neural network recognition algorithm, and dynamic time planning algorithm to identify the bump type of the road respectively.

[0088] The preset dataset can include multiple parameters from the following: wheel acceleration, vehicle acceleration, suspension height, or suspension height change rate, etc.

[0089] The types of bumps on a road can include one or more of the following: going up a bridge, going down a bridge, speed bumps, potholes, small undulations, medium undulations, large undulations, and continuous bumpy sections.

[0090] Image recognition algorithms can process road images acquired by vehicle cameras, such as performing standardized preprocessing on the images, including cropping, noise reduction, and grayscale conversion, as well as matching the processed image signals with bump types to obtain the matching degree results between the current image and various bump types.

[0091] The neural network recognition algorithm identifies the type of bump based on the input bump feature data. The input feature data may include suspension height, vibration waveform amplitude and frequency, wheel acceleration, vehicle acceleration, vehicle pitch angle, vehicle roll angle, vehicle speed or slope, etc. The output result is the matching degree of various bump types for the current road segment.

[0092] The DTW algorithm performs time-series similarity matching between bumpy segments and bump feature data from a standard set of bump types. The matching data includes suspension height, vibration waveform amplitude and frequency, wheel acceleration, and vehicle vertical acceleration. Before matching, the data is normalized to linearly map it to a reasonable range, and adaptive deformations are performed based on vehicle speed, acceleration, etc., to align the data across time series. Finally, the similarity matching results for various bump types corresponding to the current bumpy segment are obtained.

[0093] It should be noted that while image recognition can capture bumps related to road morphology, it cannot obtain effective image information in adverse weather conditions. Neural network models are highly adaptable to input data, but require a large amount of labeled data for training. The DTW algorithm takes into account the changes in time-series data, but its adaptability is poor in the absence of significant time signals.

[0094] In some possible implementations, based on a pre-collected dataset, image recognition algorithms, neural network recognition algorithms, and dynamic time planning algorithms are used to identify the types of road bumps, including: based on a pre-collected dataset, using image recognition algorithms, neural network recognition algorithms, and dynamic time planning algorithms to identify the confidence levels of various possible types of road bumps.

[0095] 106. The recognition results obtained by the image recognition algorithm, neural network recognition algorithm and dynamic time planning algorithm for identifying the road bump type are fused together, and the fused result is taken as the road bump type.

[0096] In some possible implementations, the recognition results of image recognition algorithms, neural network recognition algorithms, and dynamic time planning algorithms for identifying road bump types are fused, and the fused result is used as the road bump type; including:

[0097] The bump type with the highest matching degree after fusion is taken as the road bump type; the matching degree corresponding to any bump type i is Mi.

[0098] M i =α i A i +β i B i +γ i C i

[0099] Where, α i A represents the confidence weight of the image recognition algorithm for bump type i; i β represents the matching degree of the image recognition algorithm for bump type i; i B is the confidence weight assigned to bump type i by the neural network algorithm; i γ represents the matching degree of the neural network algorithm for identifying bump type i; i C represents the confidence weight of turbulence type i in the DTW algorithm. i Let be the matching degree of the dynamic time planning algorithm for identifying bump type i; where,

[0100]

[0101]

[0102]

[0103] Where a1 is the illumination condition influence factor of the image recognition algorithm, 0≤a1≤1; a2 is the visual output sharpness influence factor of the image recognition algorithm, with a value range of 0≤a2≤1; W i1 b1 is the initial confidence weight assigned to bump type i by the image recognition algorithm; b2 is the influence factor of neural network input data quality, 0≤b1≤1; b2 is the influence factor of neural network algorithm label data quality, with a value range of 0≤b2≤1; W i2 Here, c represents the initial confidence weight assigned to turbulence type i by the neural network algorithm; c is the standard turbulence data quality impact factor of the DTW algorithm, with a value range of 0 ≤ c ≤ 1; W i3 The initial confidence weights are assigned to bump type i in the dynamic time planning algorithm.

[0104] For example, in a bump identification process, if the top two bump types in the image recognition results are "downhill" and "large undulation," the top two bump types in the neural network recognition results are "large undulation" and "downhill," and the top two bump types in the DTW recognition results are "downhill" and "large undulation," then because the image recognition algorithm has higher accuracy in identifying downhill bumps, it receives a larger initial confidence weight for this type. Finally, after weighted fusion of the results from the three algorithms, downhill bumps receive the highest matching score, resulting in the final identification result being "downhill." To improve recognition accuracy, in a given identification process, if the difference in matching scores between the top two bump types after fusion is less than a certain threshold, the identification is considered invalid and not included in the final result. Since different algorithms have varying accuracies in identifying different bump types, fusion can reduce the risk of misidentification caused by a single algorithm and improve the overall accuracy of bump identification.

[0105] 107. Obtain location information.

[0106] 108. Upload the location information and road bump type to the server.

[0107] After identifying the type of road bumps using this embodiment, the location information and the type of road bumps are uploaded to the server, allowing the bump types obtained by this embodiment to be shared.

[0108] Please see Figure 2A , Figure 2A This is a schematic flowchart of a vehicle control method provided in another embodiment of this application. For example... Figure 2A As shown, the vehicle control method provided in this application includes steps 201 to 207. Steps 201 to 207 will be described in detail below.

[0109] 201. Obtain real-time driving information of the vehicle while it is in motion.

[0110] The driving information may include indicators such as wheel acceleration, vehicle acceleration, suspension height, or suspension height change rate.

[0111] 202. Determine whether the vehicle is in a bumpy condition based on the driving information.

[0112] In some possible implementations, the acquired driving data can be preprocessed to obtain parameter values ​​such as wheel acceleration, vehicle acceleration, suspension height, vehicle pitch angle, vehicle roll angle, vehicle speed, or slope information. Furthermore, through feature engineering, feature information including mean, standard deviation, integral, eccentricity, extreme values, variance, or power density can be extracted. Specifically, for changes in suspension height, vibration waveforms of bumpy road sections can be generated as feature graphics using methods such as filtering, integration, and Fourier transform.

[0113] Initial turbulence identification based on feature reasoning uses an expert system and compares against thresholds of set parameters to screen for the presence or absence of turbulence events. In some possible implementations, further preliminary identification of turbulence events can be made based on signals such as the amplitude, frequency, shape, and impact variation of characteristic turbulence vibration waveforms, yielding preliminary identification results. It can also preliminarily determine the turbulence category, such as bridge deck turbulence, undulating turbulence, or continuous turbulence.

[0114] 203. When the vehicle is in a bumpy state, determine that the vehicle is traveling on a bumpy road surface.

[0115] 204. When the vehicle is on a bumpy road, trigger the collection of a preset dataset; the preset dataset includes: road surface image information and vehicle driving information.

[0116] 205. Based on the collected preset dataset, use image recognition algorithm, neural network recognition algorithm, and dynamic time planning algorithm to identify the bump type of the road respectively.

[0117] The preset dataset can include multiple parameters from the following: wheel acceleration, vehicle acceleration, suspension height, or suspension height change rate, etc.

[0118] The types of bumps on a road can include one or more of the following: going up a bridge, going down a bridge, speed bumps, potholes, small undulations, medium undulations, large undulations, and continuous bumpy sections.

[0119] Image recognition algorithms can process road images acquired by vehicle cameras, such as performing standardized preprocessing on the images, including cropping, noise reduction, and grayscale conversion, as well as matching the processed image signals with bump types to obtain the matching degree results between the current image and various bump types.

[0120] The neural network recognition algorithm identifies the type of bump based on the input bump feature data. The input feature data may include suspension height, vibration waveform amplitude and frequency, wheel acceleration, vehicle acceleration, vehicle pitch angle, vehicle roll angle, vehicle speed or slope, etc. The output result is the matching degree of various bump types for the current road segment.

[0121] The DTW algorithm performs time-series similarity matching between bumpy segments and bump feature data from a standard set of bump types. The matching data includes suspension height, vibration waveform amplitude and frequency, wheel acceleration, and vehicle vertical acceleration. Before matching, the data is normalized to linearly map it to a reasonable range, and adaptive deformations are performed based on vehicle speed, acceleration, etc., to align the data across time series. Finally, the similarity matching results for various bump types corresponding to the current bumpy segment are obtained.

[0122] It should be noted that while image recognition can capture bumps related to road morphology, it cannot obtain effective image information in adverse weather conditions. Neural network models are highly adaptable to input data, but require a large amount of labeled data for training. The DTW algorithm takes into account the changes in time-series data, but its adaptability is poor in the absence of significant time signals.

[0123] In some possible implementations, based on a pre-collected dataset, image recognition algorithms, neural network recognition algorithms, and dynamic time planning algorithms are used to identify the types of road bumps, including: based on a pre-collected dataset, using image recognition algorithms, neural network recognition algorithms, and dynamic time planning algorithms to identify the confidence levels of various possible types of road bumps.

[0124] 206. The recognition results obtained by the image recognition algorithm, neural network recognition algorithm and dynamic time planning algorithm for identifying the road bump type are integrated, and the integrated result is taken as the road bump type.

[0125] In some possible implementations, the recognition results of image recognition algorithms, neural network recognition algorithms, and dynamic time planning algorithms for identifying road bump types are fused, and the fused result is used as the road bump type; including:

[0126] The bump type with the highest matching degree after fusion is taken as the road bump type; the matching degree corresponding to any bump type i is Mi.

[0127] M i =α i A i +β i B i +γ i C i

[0128] Where, α i A represents the confidence weight of the image recognition algorithm for bump type i; i β represents the matching degree of the image recognition algorithm for bump type i; i B is the confidence weight assigned to bump type i by the neural network algorithm; i γ represents the matching degree of the neural network algorithm for identifying bump type i; i C represents the confidence weight of turbulence type i in the DTW algorithm. i Let be the matching degree of the dynamic time planning algorithm for identifying bump type i; where,

[0129]

[0130]

[0131]

[0132] Where a1 is the illumination condition influence factor of the image recognition algorithm, 0≤a1≤1; a2 is the visual output sharpness influence factor of the image recognition algorithm, with a value range of 0≤a2≤1; W i1 b1 is the initial confidence weight assigned to bump type i by the image recognition algorithm; b2 is the influence factor of neural network input data quality, 0≤b1≤1; b2 is the influence factor of neural network algorithm label data quality, with a value range of 0≤b2≤1; W i2 Here, c represents the initial confidence weight assigned to turbulence type i by the neural network algorithm; c is the standard turbulence data quality impact factor of the DTW algorithm, with a value range of 0 ≤ c ≤ 1; W i3The initial confidence weights are assigned to bump type i in the dynamic time planning algorithm.

[0133] For example, in a bump identification process, if the top two bump types in the image recognition results are "downhill" and "large undulation," the top two bump types in the neural network recognition results are "large undulation" and "downhill," and the top two bump types in the DTW recognition results are "downhill" and "large undulation," then because the image recognition algorithm has higher accuracy in identifying downhill bumps, it receives a larger initial confidence weight for this type. Finally, after weighted fusion of the results from the three algorithms, downhill bumps receive the highest matching score, resulting in the final identification result being "downhill." To improve recognition accuracy, in a given identification process, if the difference in matching scores between the top two bump types after fusion is less than a certain threshold, the identification is considered invalid and not included in the final result. Since different algorithms have varying accuracies in identifying different bump types, fusion can reduce the risk of misidentification caused by a single algorithm and improve the overall accuracy of bump identification.

[0134] 207. Control the suspension to match its state with the identification results so that the vehicle can smoothly pass through bumpy road sections.

[0135] Controlling the suspension includes adjusting the suspension height or the suspension stiffness.

[0136] In this embodiment, after obtaining the identification result of the road bump type, the suspension is controlled according to the identification result to match the suspension state with the identification result, allowing the vehicle to smoothly pass through bumpy road sections. This helps improve driving comfort and safety.

[0137] Please see Figure 2B , Figure 2B This is a schematic flowchart of a vehicle control method provided in another embodiment of this application. For example... Figure 2B As shown, the vehicle control method provided in this application includes steps 201 to 209. Steps 201 to 209 will be described in detail below.

[0138] Steps 201 to 207 and Figure 2A Steps 201 to 207 are the same, and will not be repeated here for the sake of simplicity.

[0139] 208. Obtain location information.

[0140] 209. Upload the location information and road bump type to the server.

[0141] After identifying the type of road bumps using this embodiment, the location information and the type of road bumps are uploaded to the server, allowing the bump types obtained by this embodiment to be shared.

[0142] Please see Figure 2C, Figure 2C This is a schematic flowchart of a vehicle control method provided in another embodiment of this application. For example... Figure 2C As shown, the vehicle control method provided in this application includes steps 201 to 212. Steps 201 to 212 will be described in detail below.

[0143] Steps 201 to 207 and Figure 2A Steps 201 to 207 are the same, and steps 208 to 209 are the same. Figure 2B Steps 208 to 209 are the same, and for simplicity, they will not be repeated here.

[0144] 210. Obtain bump information from the server within a preset range of the current location of other vehicle owners' vehicles. The bump information includes: location information and bump type.

[0145] 211. Predict the type of bumps on the road section you are about to pass through based on the vehicle's real-time driving and bump information.

[0146] 212. Control the suspension to match the predicted bump type of the upcoming road section.

[0147] Controlling the suspension includes adjusting the suspension height or the suspension stiffness.

[0148] This embodiment can predict the type of bumps on the road to be traversed and adjust the suspension control in advance to improve driving comfort and safety.

[0149] like Figure 3 The diagram shown is a structural schematic of a road bump type identification device provided in an embodiment of this application. See also... Figure 3 The road bump type identification device 300 includes: an acquisition unit 301, a first determination unit 302, an identification unit 303, and a fusion unit 304.

[0150] The system includes: an acquisition unit 301 for acquiring real-time driving information of the vehicle; a first determination unit 302 for determining whether the vehicle is in a bumpy state based on the driving information; and for determining that the vehicle is traveling on a bumpy road surface when it is in a bumpy state; the acquisition unit 301 is also used to trigger the collection of a preset dataset when the vehicle is on a bumpy road surface; the preset dataset includes: road surface image information and vehicle driving information; an identification unit 303 for identifying the bumpy type of the road using image recognition algorithm, neural network recognition algorithm, and dynamic time planning algorithm based on the preset dataset collected by the acquisition unit 301; and a fusion unit 304 for fusing the identification results of the bumpy type of the road identified by the image recognition algorithm, neural network recognition algorithm, and dynamic time planning algorithm, and using the fused result as the bumpy type of the road.

[0151] In some possible implementations, the types of road bumps include one or more of the following: bridges going up and down, speed bumps, potholes, small undulations, medium undulations, large undulations, and continuous bumpy sections.

[0152] In some possible implementations, in terms of identifying the road bump type using image recognition algorithm, neural network recognition algorithm, and dynamic time planning algorithm based on the collected preset dataset, the identification unit 303 is specifically used to identify the confidence level of various possible road bump types using image recognition algorithm, neural network recognition algorithm, and dynamic time planning algorithm based on the collected preset dataset.

[0153] In some possible implementations, the fusion unit 304 is specifically used to: fuse the identification results of image recognition algorithms, neural network recognition algorithms, and dynamic time planning algorithms respectively to identify the road bump type, and use the fused result as the road bump type.

[0154] The bump type with the highest matching degree after fusion is taken as the bump type of the road; the matching degree corresponding to any bump type i is Mi.

[0155] Mi=αiAi+βiBi+γiCi

[0156] Where αi is the confidence weight of the image recognition algorithm for bump type i; Ai is the recognition matching degree of the image recognition algorithm for bump type i; βi is the confidence weight of the neural network algorithm for bump type i; Bi is the recognition matching degree of the neural network algorithm for bump type i; γi is the confidence weight of the DTW algorithm for bump type i; Ci is the recognition matching degree of the dynamic time planning algorithm for bump type i; and so on.

[0157]

[0158]

[0159]

[0160] Where a1 is the illumination condition influence factor of the image recognition algorithm, 0≤a1≤1; a2 is the visual output sharpness influence factor of the image recognition algorithm, with a value range of 0≤a2≤1; W i1 b1 is the initial confidence weight assigned to bump type i by the image recognition algorithm; b2 is the influence factor of neural network input data quality, 0≤b1≤1; b2 is the influence factor of neural network algorithm label data quality, with a value range of 0≤b2≤1; W i2 Here, c represents the initial confidence weight assigned to turbulence type i by the neural network algorithm; c is the standard turbulence data quality impact factor of the DTW algorithm, with a value range of 0 ≤ c ≤ 1; Wi3 The initial confidence weights are assigned to bump type i in the dynamic time planning algorithm.

[0161] In some possible implementations, the acquisition unit 301 can also be used to acquire location information, and the road bump type identification device can also include a communication unit for uploading the location information and the road bump type to a server.

[0162] For simplicity, the specific implementation of each unit module can be found in the description of the relevant steps in the previous method embodiments, and will not be repeated here.

[0163] It should be noted that a "unit" can be a software program, a hardware circuit, or a combination of both that implements the above functions. The hardware circuit may include an application-specific integrated circuit (ASIC), electronic circuitry, a processor (e.g., a shared processor, a proprietary processor, or a group processor) and memory for executing one or more software or firmware programs, integrated logic circuitry, and / or other suitable components that support the described functions.

[0164] Therefore, the units of the various examples described in the embodiments of this application can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0165] Corresponding to the road bump type identification method and vehicle control method provided in the embodiments of this application, this application also provides a processing device, such as... Figure 4 As shown, the processing device 400 includes a memory 401 and a processor 402. The memory 401 stores a computer program executable by the processor 402, which, when executing the computer program, implements the method described in any of the above-described method embodiments.

[0166] This application also provides a computer program product that, when executed by processor 402, implements the methods described in any of the method embodiments of this application. This computer program product may be stored in memory 401.

[0167] This application also provides a computer-readable storage medium storing a computer program thereon, which, when executed by a computer, implements the road bump type identification method or vehicle control method described in any of the method embodiments of this application. The computer program may be a high-level language program or an executable object program.

[0168] The computer-readable storage medium can be volatile or non-volatile memory. Non-volatile memory can be read-only memory (ROM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), electrically erasable programmable read-only memory (EEPROM), or flash memory. Volatile memory can be random access memory (RAM) used as an external cache. By way of example, but not limitation, many forms of RAM are available, such as static random access memory (SRAM), dynamic random access memory (DRAM), synchronous dynamic random access memory (SDRAM), double data rate synchronous dynamic random access memory (DDRSDRAM), enhanced synchronous dynamic random access memory (ESDRAM), synchronous linked dynamic random access memory (SLDRAM), and direct rambus RAM (DR RAM).

[0169] In the above embodiments, the descriptions of each embodiment have different focuses. For parts that are not described in detail or recorded in a certain embodiment, please refer to the relevant descriptions of other embodiments.

[0170] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0171] In the embodiments provided in this application, it should be understood that the disclosed methods and terminal devices can be implemented in other ways. For example, the device / network device embodiments described above are merely illustrative. For instance, the division of modules or units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between devices or units may be electrical, mechanical, or other forms.

[0172] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0173] It should be understood that, when used in this application specification and the appended claims, the term "comprising" indicates the presence of the described features, integrals, steps, operations, elements and / or components, but does not exclude the presence or addition of one or more other features, integrals, steps, operations, elements, components and / or a collection thereof.

[0174] It should also be understood that the term “and / or” as used in this application specification and the appended claims means any combination of one or more of the associated listed items and all possible combinations, and includes such combinations.

[0175] References to "one embodiment" or "some embodiments" as described in this specification mean that one or more embodiments of this application include a specific feature, structure, or characteristic described in connection with that embodiment. Therefore, the phrases "in one embodiment," "in some embodiments," "in other embodiments," "in still other embodiments," etc., appearing in different parts of this specification do not necessarily refer to the same embodiment, but rather mean "one or more, but not all, embodiments," unless otherwise specifically emphasized. The terms "comprising," "including," "having," and variations thereof mean "including but not limited to," unless otherwise specifically emphasized.

[0176] The above-described embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application, and should all be included within the protection scope of this application.

Claims

1. A method for identifying the type of road bumps, characterized in that, The method includes: Obtain real-time driving information while the vehicle is in motion; Determine whether the vehicle is in a bumpy state based on the driving information; When the vehicle is in a bumpy state, it is determined that the vehicle is traveling on a bumpy road surface; When the vehicle is on a bumpy road, a preset dataset is triggered for collection; the preset dataset includes: road surface image information and vehicle driving information; Based on the collected preset dataset, the types of road bumps are identified using image recognition algorithms, neural network recognition algorithms, and dynamic time planning algorithms, respectively. The recognition results of the image recognition algorithm, neural network recognition algorithm and dynamic time planning algorithm for identifying the road bump type are combined, and the fused result is taken as the road bump type. The step of identifying road bump types using image recognition algorithms, neural network recognition algorithms, and dynamic time planning algorithms based on the collected preset dataset includes: Based on the collected preset dataset, the confidence levels of various possible bump types of the road are identified using image recognition algorithms, neural network recognition algorithms, and dynamic time planning algorithms, respectively.

2. The method according to claim 1, characterized in that, The method of fusing the recognition results of the image recognition algorithm, neural network recognition algorithm, and dynamic time planning algorithm to identify the road bump type, and using the fused result as the road bump type, includes: The bump type with the highest matching degree after fusion is taken as the bump type of the road; the matching degree corresponding to any bump type i is Mi. M i =a i A i +b i B i +g i C i Where, α i A represents the confidence weight of the image recognition algorithm for bump type i; i β represents the matching degree of the image recognition algorithm for bump type i; i B is the confidence weight assigned to bump type i by the neural network recognition algorithm; i γ represents the matching degree of the neural network recognition algorithm for bump type i; i C represents the confidence weights for turbulence type i in the dynamic time planning algorithm. i Let be the matching degree of the dynamic time planning algorithm for identifying bump type i; where, Where a1 is the illumination condition influence factor of the image recognition algorithm, 0≤a1≤1; a2 is the visual output sharpness influence factor of the image recognition algorithm, with a value range of 0≤a2≤1; W i1 b1 is the initial confidence weight assigned to bump type i by the image recognition algorithm; b2 is the input data quality impact factor of the neural network recognition algorithm, 0≤b1≤1; b2 is the label data quality impact factor of the neural network recognition algorithm, with a value range of 0≤b2≤1; W i2 W represents the initial confidence weight assigned to bump type i by the neural network identification algorithm; c is the standard bump data quality impact factor of the dynamic time programming algorithm, with a value range of 0 ≤ c ≤ 1; i3 The initial confidence weights are assigned to bump type i in the dynamic time planning algorithm.

3. The method according to claim 1, characterized in that, Also includes: Obtain location information; The location information and the road bump type are uploaded to the server.

4. The method according to any one of claims 1 to 3, characterized in that, The road bump types include one or more of the following: going up a bridge, going down a bridge, potholes, small undulations, medium undulations, large undulations, and continuous bumpy sections.

5. A vehicle control method, characterized in that, include: The method described in any one of claims 1 to 4 is used to identify the type of road bumps; The suspension is controlled to match its state with the identification result, allowing the vehicle to smoothly pass through bumpy road sections.

6. The method according to claim 5, characterized in that, Also includes: Obtain bump information within a preset range of the current location of the vehicle, shared by other vehicle owners. The bump information includes: location information and bump type. Based on the vehicle's real-time driving information and the bump information, the type of bumps in the upcoming road section is predicted. The suspension is controlled to match the predicted bump type of the upcoming road segment.

7. A device for identifying the type of road bumps, characterized in that, The device includes: The acquisition unit is used to acquire real-time driving information of the vehicle while it is in motion; The first determining unit is configured to determine whether the vehicle is in a bumpy state based on the driving information; and to determine that the vehicle is traveling on a bumpy road surface when the vehicle is in a bumpy state. The acquisition unit is further configured to trigger the collection of a preset dataset when the vehicle is on a bumpy road surface; the preset dataset includes: road surface image information and vehicle driving information; The identification unit is used to identify the type of road bumps based on the preset dataset collected by the acquisition unit, using image recognition algorithm, neural network recognition algorithm and dynamic time planning algorithm respectively. The fusion unit is used to fuse the recognition results of the image recognition algorithm, the neural network recognition algorithm and the dynamic time planning algorithm respectively in identifying the bump type of the road, and use the fused result as the bump type of the road. The identification unit is also used to identify the confidence level of various possible bump types of the road based on the collected preset dataset, using image recognition algorithm, neural network recognition algorithm and dynamic time planning algorithm respectively.

8. A processing apparatus, characterized in that, include: A memory and a processor, the memory storing a computer program executable by the processor, the processor executing the computer program to implement the method of any one of claims 1 to 4 and / or claim 5 or 6.

9. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program that, when read and executed, implements the method of any one of claims 1 to 4 and / or claim 5 or 6.

Images