Vehicle tire wear degree real-time monitoring method, device, equipment and program product

Abstract

The application discloses a vehicle tire wear degree real-time monitoring method, device, equipment and program product, comprising: acquiring a current tire model, tire image data, wheel speed difference time series data, tire pressure time series data and suspension vibration frequency spectrum of a current vehicle; inputting the current tire model, tire image data, wheel speed difference time series data, tire pressure time series data and suspension vibration frequency spectrum into a pre-trained tire wear degree recognition model to obtain a current tire wear degree of the current vehicle; wherein the tire wear degree recognition model is obtained based on meta-learning training, and the current tire model is used to determine feature fusion weights of the tire image data, wheel speed difference time series data, tire pressure time series data and suspension vibration frequency spectrum. The application improves the accuracy and real-time performance of vehicle tire wear degree monitoring, and can be applied to vehicles with different tire models, and can be widely applied to the technical field of vehicle monitoring.

Description

Technical Field

[0001] This invention relates to the field of vehicle monitoring technology, and in particular to a method, device, equipment, and program product for real-time monitoring of vehicle tire wear. Background Technology

[0002] Tire wear is an inevitable loss that occurs during vehicle operation. Excessive wear can significantly reduce vehicle grip, shorten braking distance, and even cause tire blowout accidents.

[0003] Current mainstream tire wear monitoring technologies mainly include: 1) Indirect estimation of tire wear based on tire pressure / temperature changes (such as TPMS), which has low accuracy and cannot distinguish wear from other interfering factors; 2) Visual detection based on image recognition, which is easily affected by light and mud, and usually only trains models for a single preset tire tread pattern, and cannot accurately monitor after changing the current tire model; 3) Tire wear detection based on vibration / acoustic signal analysis, which has poor model generalization ability, and the wear characteristics of different tire structures (such as summer tires, winter tires, and off-road tires) vary greatly, resulting in weak universality.

[0004] In summary, existing tire wear monitoring technologies suffer from low accuracy and only support a single tire type, which urgently need to be addressed. Summary of the Invention

[0005] The purpose of this invention is to at least partially solve one of the technical problems existing in the prior art.

[0006] Therefore, one objective of this invention is to provide a method for real-time monitoring of vehicle tire wear. This method is based on meta-learning pre-training to obtain a tire wear identification model that can adjust the feature fusion weights according to the tire model. By inputting the current vehicle's tire model, tire image data, wheel speed difference time series data, tire pressure time series data, and suspension vibration spectrum into the tire wear identification model, the current tire wear of the vehicle can be obtained. This improves the accuracy and real-time performance of vehicle tire wear monitoring and is applicable to vehicles with different tire models.

[0007] Another objective of this invention is to provide a device for real-time monitoring of vehicle tire wear.

[0008] To achieve the above-mentioned technical objectives, the technical solutions adopted in the embodiments of the present invention include: On one hand, embodiments of the present invention provide a method for real-time monitoring of vehicle tire wear, comprising the following steps: Acquire the current tire model, tire image data, wheel speed difference time series data, tire pressure time series data, and suspension vibration spectrum of the current vehicle; The current tire model, the tire image data, the wheel speed difference time series data, the tire pressure time series data, and the suspension vibration spectrum are input into a pre-trained tire wear recognition model to obtain the current tire wear level of the current vehicle. The tire wear identification model is trained based on meta-learning, and the current tire model is used to determine the feature fusion weights of the tire image data, the wheel speed difference time series data, the tire pressure time series data, and the suspension vibration spectrum.

[0009] Furthermore, in one embodiment of the present invention, the acquisition of the current tire model, tire image data, wheel speed difference time series data, tire pressure time series data, and suspension vibration spectrum of the current vehicle specifically includes: Obtain the current tire model input by the user; Tire image data is acquired by a camera device positioned above each tire of the current vehicle; The wheel speed is obtained by using wheel speed sensors installed on each tire of the current vehicle, and the wheel speed difference timing data is determined based on the wheel speed of the tires on the same axle. The tire pressure timing data is obtained by tire pressure sensors installed in each tire of the current vehicle; The suspension vibration signal is acquired by a vibration sensor installed in the suspension system of the current vehicle, and the suspension vibration spectrum is obtained by performing a Fourier transform on the suspension vibration signal.

[0010] Furthermore, in one embodiment of the present invention, the tire wear recognition model is trained through the following steps: Training samples from multiple test vehicles are obtained, including sample tire models, tire image samples, wheel speed difference time series samples, tire pressure time series samples, and suspension vibration spectrum samples. The tire wear degree label corresponding to each training sample was determined by manual annotation. The sample tire models are divided into multiple meta-learning sub-tasks, and the support set and query set corresponding to each meta-learning sub-task are determined based on the training samples and the tire wear degree labels. A CNN-LSTM meta-learning model is constructed, and the CNN-LSTM meta-learning model is trained on each meta-learning sub-task based on the support set and the query set to obtain the trained tire wear degree recognition model.

[0011] Further, in one embodiment of the present invention, the step of training the CNN-LSTM meta-learning model on each of the meta-learning sub-tasks based on the support set and the query set to obtain the trained tire wear degree recognition model specifically includes: The CNN-LSTM meta-learning model is locally updated according to the support set of each meta-learning sub-task to obtain the CNN-LSTM meta-learning sub-model corresponding to each meta-learning sub-task; Determine the subtask loss value of each CNN-LSTM meta-learning sub-model on the query set of the corresponding meta-learning sub-task; The global loss value is determined based on the sub-task loss value, and the global parameters of the CNN-LSTM meta-learning model are updated based on the global loss value. Then, the process returns to the step of updating the local parameters of the CNN-LSTM meta-learning model based on the support set of each meta-learning sub-task. When the global loss value is lower than the preset loss threshold, training stops, and the trained tire wear recognition model is obtained.

[0012] Further, in one embodiment of the present invention, the CNN-LSTM meta-learning model includes a first CNN branch network, a second CNN branch network, a first LSTM branch network, a second LSTM branch network, a feature fusion network, and a fully connected layer. The step of updating the local parameters of the CNN-LSTM meta-learning model according to the support set of each meta-learning sub-task to obtain the CNN-LSTM meta-learning sub-model corresponding to each meta-learning sub-task specifically includes: The current training samples contained in the support set of the current meta-learning subtask are input into the CNN-LSTM meta-learning model; The first CNN branch network, the second CNN branch network, the first LSTM branch network, and the second LSTM branch network are used to extract features from the tire image samples, the suspension vibration spectrum samples, the wheel speed difference time series samples, and the tire pressure time series samples corresponding to the current training samples, thereby obtaining tire image features, vibration spectrum features, wheel speed difference time series features, and tire pressure time series features. The CNN-LSTM meta-learning model determines the current sub-task feature fusion weights based on the sample tire model corresponding to the current meta-learning sub-task, and then performs feature fusion on the tire image features, vibration spectrum features, wheel speed difference time series features, and tire pressure time series features through the feature fusion network based on the current sub-task feature fusion weights to obtain a feature fusion vector. The fully connected layer maps the feature fusion vector into a predicted tire wear level. The current loss value is determined based on the predicted tire wear level and the tire wear level label corresponding to the current training sample; Based on the current loss value, the local model parameters and subtask feature fusion weights of the CNN-LSTM meta-learning model are updated using the gradient descent algorithm to obtain the CNN-LSTM meta-learning sub-model corresponding to each meta-learning sub-task.

[0013] Furthermore, in one embodiment of the present invention, the step of determining a global loss value based on the sub-task loss value and updating the global parameters of the CNN-LSTM meta-learning model based on the global loss value specifically includes: The subtask weight parameters corresponding to the meta-learning subtasks are determined based on the market share of each of the sample tire models. The global loss value is obtained by weighted summation of the sub-task loss values ​​based on the sub-task weight parameters. The global model parameters and global feature fusion weights of the CNN-LSTM meta-learning model are updated using the backpropagation algorithm based on the global loss value.

[0014] Furthermore, in one embodiment of the present invention, the real-time vehicle tire wear monitoring method further includes the following steps: Determine whether the current tire wear level is greater than or equal to a preset safety threshold; When the current tire wear level is greater than or equal to the safety threshold, a warning is issued to the driver of the current vehicle.

[0015] On the other hand, embodiments of the present invention provide a real-time vehicle tire wear monitoring device, comprising: The data acquisition module is used to acquire the current tire model, tire image data, wheel speed difference time series data, tire pressure time series data, and suspension vibration spectrum of the current vehicle. The wear degree recognition module is used to input the current tire model, the tire image data, the wheel speed difference time series data, the tire pressure time series data, and the suspension vibration spectrum into a pre-trained tire wear degree recognition model to obtain the current tire wear degree of the current vehicle. The tire wear identification model is trained based on meta-learning, and the current tire model is used to determine the feature fusion weights of the tire image data, the wheel speed difference time series data, the tire pressure time series data, and the suspension vibration spectrum.

[0016] On the other hand, embodiments of the present invention provide an electronic device, including: At least one processor; At least one memory for storing at least one program; When the at least one program is executed by the at least one processor, the at least one processor implements the above-described method for real-time monitoring of vehicle tire wear.

[0017] On the other hand, embodiments of the present invention also provide a computer-readable storage medium storing a processor-executable computer program that, when executed by a processor, implements the above-described method for real-time monitoring of vehicle tire wear.

[0018] On the other hand, embodiments of the present invention also provide a computer program product, including a computer program that, when executed by a processor, implements the above-described method for real-time monitoring of vehicle tire wear.

[0019] The advantages and beneficial effects of the present invention will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of the invention: This invention provides a tire wear identification model that is pre-trained based on meta-learning and can adjust the feature fusion weights according to the tire model. By inputting the current vehicle's tire model, tire image data, wheel speed difference time series data, tire pressure time series data, and suspension vibration spectrum into the tire wear identification model, the current tire wear level of the vehicle can be obtained. This improves the accuracy and real-time performance of vehicle tire wear monitoring and is applicable to vehicles with different tire models. Attached Figure Description

[0020] To more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings used in the embodiments of the present invention are described below. It should be understood that the drawings described below are only for the convenience of clearly describing some embodiments of the technical solutions of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0021] Figure 1 A flowchart illustrating the steps of a method for real-time monitoring of vehicle tire wear according to an embodiment of the present invention; Figure 2 This is a structural block diagram of a vehicle tire wear real-time monitoring device provided in an embodiment of the present invention; Figure 3 This is a structural block diagram of an electronic device provided in an embodiment of the present invention. Detailed Implementation

[0022] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention. In the following description, when referring to the accompanying drawings, unless otherwise indicated, the same numbers in different drawings represent the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with the embodiments of this invention; they are merely examples of apparatuses and methods consistent with some aspects of the embodiments of this invention as detailed in the appended claims.

[0023] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains. The terminology used herein is for the purpose of describing embodiments of the invention only and is not intended to limit the invention.

[0024] The real-time vehicle tire wear monitoring method provided in this invention can be applied to a terminal, a server, or software running on either a terminal or a server. In some embodiments, the terminal can be a smartphone, tablet, laptop, desktop computer, smart speaker, smartwatch, or in-vehicle terminal, but is not limited to these. The server can be configured as an independent physical server, a server cluster or distributed system composed of multiple physical servers, or a cloud server providing basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communication, middleware services, domain name services, security services, CDN, and big data and artificial intelligence platforms. The server can also be a node server in a blockchain network. The software can be an application that implements the real-time vehicle tire wear monitoring method, but is not limited to the above forms.

[0025] This invention can be used in a wide variety of general-purpose or special-purpose computer system environments or configurations. Examples include: personal computers, server computers, handheld or portable devices, tablet devices, multiprocessor systems, microprocessor-based systems, set-top boxes, programmable consumer electronics, network PCs, minicomputers, mainframe computers, and distributed computing environments including any of the above systems or devices. This invention can be described in the general context of computer-executable instructions, such as program modules, that are executed by a computer. Generally, program modules include routines, programs, objects, components, data structures, etc., that perform specific tasks or implement specific abstract data types. This invention can also be practiced in distributed computing environments where tasks are performed by remote processing devices connected via a communication network. In distributed computing environments, program modules can reside in local and remote computer storage media, including storage devices.

[0026] It should be noted that in various specific embodiments of the present invention, when processing data related to user identity or characteristics, such as user information, user behavior data, user historical data, and user parking space location information, user permission or consent is obtained first. Furthermore, the collection, use, and processing of this data comply with relevant laws, regulations, and standards. In addition, when embodiments of the present invention require access to sensitive personal information of users, separate permission or consent from the user is obtained through pop-ups or redirection to a confirmation page. Only after obtaining the user's separate permission or consent is the necessary user-related data for the normal operation of the embodiments of the present invention acquired.

[0027] Reference Figure 1 This invention provides a method for real-time monitoring of vehicle tire wear, specifically including the following steps: S101. Obtain the current tire model, tire image data, wheel speed difference time series data, tire pressure time series data, and suspension vibration spectrum of the current vehicle. S102. Input the current tire model, tire image data, wheel speed difference time series data, tire pressure time series data and suspension vibration spectrum into the pre-trained tire wear degree recognition model to obtain the current tire wear degree of the current vehicle. Among them, the tire wear identification model is trained based on meta-learning, and the current tire model is used to determine the feature fusion weights of tire image data, wheel speed difference time series data, tire pressure time series data, and suspension vibration spectrum.

[0028] This invention provides a tire wear identification model that is pre-trained based on meta-learning and can adjust the feature fusion weights according to the tire model. By inputting the current vehicle's tire model, tire image data, wheel speed difference time series data, tire pressure time series data, and suspension vibration spectrum into the tire wear identification model, the current tire wear level of the vehicle can be obtained. This improves the accuracy and real-time performance of vehicle tire wear monitoring and is applicable to vehicles with different tire models.

[0029] As a further optional implementation, the current tire model, tire image data, wheel speed difference time series data, tire pressure time series data, and suspension vibration spectrum of the current vehicle are acquired, specifically including: S1011. Obtain the current tire model input by the user; S1012. Obtain tire image data by means of a camera device installed above each tire of the current vehicle; S1013. Obtain the corresponding tire speed by using wheel speed sensors installed on each tire of the current vehicle, and determine the wheel speed difference timing data based on the tire speed of the tires on the same axle. S1014. Obtain tire pressure timing data through tire pressure sensors installed in each tire of the current vehicle; S1015. Obtain suspension vibration signals by using vibration sensors installed in the suspension system of the current vehicle, and perform Fourier transform on the suspension vibration signals to obtain the suspension vibration spectrum.

[0030] Specifically, after replacing tires, users can input the currently installed tire model via the vehicle's HMI, mobile app, or OBD interface, which is stored in the vehicle controller. This tire model determines the feature fusion weights for different subsequent characteristics. Tire image data is acquired through cameras positioned above each tire, visually reflecting changes in tread depth caused by tire wear. Wheel speed sensors on each tire acquire corresponding tire rotation speeds, determining wheel speed difference time-series data based on the rotation speeds of tires on the same axle. This time-series data reflects the difference in rolling radii between the left and right tires on the same axle, indicating different wear levels. Tire pressure sensors inside each tire acquire tire pressure time-series data, reflecting tire pressure fluctuations caused by changes in the tire's contact area with the ground, thus indicating tire wear. Vibration sensors in the suspension system acquire suspension vibration signals, which are then subjected to Fourier transform to obtain the suspension vibration spectrum, reflecting vehicle imbalance caused by tire wear.

[0031] The current tire model, tire image data, wheel speed difference time series data, tire pressure time series data, and suspension vibration spectrum are input into a pre-trained tire wear recognition model. This tire wear recognition model is pre-trained based on meta-learning using training samples of multiple different tire models. It automatically selects the corresponding feature fusion weights according to the input current tire model, extracts features from the tire image data, wheel speed difference time series data, tire pressure time series data, and suspension vibration spectrum, and then performs feature fusion based on the feature fusion weights, thereby identifying the current tire wear level of the current vehicle based on the fused features.

[0032] As an optional further implementation, the tire wear recognition model is trained through the following steps: S201. Obtain training samples from multiple test vehicles. The training samples include sample tire models, tire image samples, wheel speed difference time series samples, tire pressure time series samples, and suspension vibration spectrum samples. S202. Determine the tire wear label corresponding to each training sample through manual annotation; S203. Divide the sample tire models into multiple meta-learning sub-tasks, and determine the support set and query set corresponding to each meta-learning sub-task based on the training samples and tire wear degree labels. S204. Construct a CNN-LSTM meta-learning model. Train the CNN-LSTM meta-learning model on each meta-learning sub-task based on the support set and query set to obtain a trained tire wear degree recognition model.

[0033] Specifically, training samples from multiple test vehicles were acquired, including tire models, tire image samples, time-series samples of wheel speed differences, time-series samples of tire pressure, and suspension vibration spectrum samples. Professional technicians were invited to label the wear levels of the collected training samples according to tire wear limit standards (e.g., tread depth less than 1.6mm), using different scores to represent tire wear levels, with higher scores indicating higher wear levels. Meta-learning tasks were divided around tire models, with each model corresponding to a sub-task. Each sub-task contained multi-source data samples of that tire model under different wear levels and road conditions. Based on the aforementioned training samples, a support set and a query set were assigned to each sub-task. The support set contained a small number of samples with wear labels (e.g., 5-10 samples per wear level), while the query set contained a large number of training samples of the same tire model. A CNN-LSTM meta-learning model was constructed and trained on each meta-learning sub-task using the support set and query set, thereby obtaining a tire wear level recognition model capable of handling different tire models.

[0034] As an optional further implementation, the CNN-LSTM meta-learning model is trained on each meta-learning subtask based on the support set and query set to obtain a trained tire wear recognition model, which specifically includes: S2041. Update the local parameters of the CNN-LSTM meta-learning model according to the support set of each meta-learning sub-task to obtain the CNN-LSTM meta-learning sub-model corresponding to each meta-learning sub-task. S2042. Determine the subtask loss value of each CNN-LSTM meta-learning sub-model on the query set of the corresponding meta-learning sub-task; S2043. Determine the global loss value based on the sub-task loss value, and update the global parameters of the CNN-LSTM meta-learning model based on the global loss value. Then return to the step of updating the local parameters of the CNN-LSTM meta-learning model based on the support set of each meta-learning sub-task. S2044. When the global loss value is lower than the preset loss threshold, stop training and obtain the trained tire wear recognition model.

[0035] Specifically, the embodiments of the present invention use the MAML (Model Independent Meta-Learning) algorithm as the core framework, and achieve cross-model transfer through a two-stage training of rapid adaptation and global update.

[0036] Specifically, in the rapid adaptation phase, the CNN-LSTM meta-learning model is locally updated based on the support set of each meta-learning sub-task, resulting in a CNN-LSTM meta-learning sub-model corresponding to each sub-task. This involves adjusting the model parameters with a small number of gradient updates on the support set of each sub-task, enabling the model to quickly adapt to the feature distribution of that tire model. In the global update phase, the sub-task loss value of each CNN-LSTM meta-learning sub-model on the query set of its corresponding meta-learning sub-task is determined. The global loss value is then determined based on the sub-task loss value, and the CNN-LSTM meta-learning model's global parameters are updated based on this global loss value. This involves summing the losses of all models after rapid adaptation on the query set and updating the initial parameters of the meta-learning model in reverse, improving the model's generalization ability across different tasks. Rapid adaptation and global updates are performed alternately until the global loss value falls below a preset loss threshold, at which point training stops, resulting in a well-trained tire wear recognition model.

[0037] As an optional implementation, the CNN-LSTM meta-learning model includes a first CNN branch network, a second CNN branch network, a first LSTM branch network, a second LSTM branch network, a feature fusion network, and a fully connected layer. Local parameter updates are performed on the CNN-LSTM meta-learning model based on the support set of each meta-learning sub-task to obtain the CNN-LSTM meta-learning sub-model corresponding to each meta-learning sub-task, which specifically includes: S20411. Input the current training samples contained in the support set of the current meta-learning subtask into the CNN-LSTM meta-learning model; S20412. The tire image samples, suspension vibration spectrum samples, wheel speed difference time series samples and tire pressure time series samples corresponding to the current training samples are extracted through the first CNN branch network, the second CNN branch network, the first LSTM branch network and the second LSTM branch network to obtain tire image features, vibration spectrum features, wheel speed difference time series features and tire pressure time series features. S20413. Determine the feature fusion weights of the current sub-task of the CNN-LSTM meta-learning model based on the sample tire model corresponding to the current meta-learning sub-task, and perform feature fusion on the tire image features, vibration spectrum features, wheel speed difference time series features and tire pressure time series features through the feature fusion network according to the current sub-task feature fusion weights to obtain the feature fusion vector. S20414. The feature fusion vector is mapped to the predicted value of tire wear through a fully connected layer; S20415. Determine the current loss value based on the predicted tire wear level and the tire wear level label corresponding to the current training sample; S20416. Based on the current loss value, update the local model parameters and sub-task feature fusion weights of the CNN-LSTM meta-learning model using the gradient descent algorithm to obtain the CNN-LSTM meta-learning sub-model corresponding to each meta-learning sub-task.

[0038] Specifically, the CNN-LSTM meta-learning model in this embodiment of the invention includes a first CNN branch network, a second CNN branch network, a first LSTM branch network, a second LSTM branch network, a feature fusion network, and a fully connected layer. For tire images, a lightweight convolutional neural network (first CNN branch network) is used to extract image features, focusing on capturing visual features such as tread wear, cracks, and bulges. For vehicle vibration spectra, a one-dimensional convolutional network (second CNN branch network) is used to extract spectral features, filtering feature frequency bands related to tire wear. For wheel speed difference and tire pressure time-series data, a bidirectional LSTM network (first LSTM branch network and second LSTM branch network) is used to extract time-series features and identify the data... The long-term trend and periodic fluctuations of tire wear are analyzed. Tire image features, vibration spectrum features, wheel speed difference time series features, and tire pressure time series features are input into a feature fusion network. The weights of different features are dynamically adjusted according to the current tire model (e.g., vibration spectrum features are more important for off-road tires, while image features are more important for passenger car tires), resulting in a feature fusion vector. The feature fusion vector is mapped to a predicted tire wear level through a fully connected layer. The current loss value is determined based on the predicted tire wear level and the tire wear level label corresponding to the current training sample. Based on the current loss value, the local model parameters and sub-task feature fusion weights of the CNN-LSTM meta-learning model are updated using the gradient descent algorithm to obtain the CNN-LSTM meta-learning sub-model corresponding to each meta-learning sub-task.

[0039] It should be noted that the process of determining the subtask loss value of each CNN-LSTM meta-learning sub-model on the query set of the corresponding meta-learning sub-task is similar to the process of calculating the current loss value of the training samples of the support set, and will not be described in detail here in this embodiment of the invention.

[0040] As a further optional implementation, a global loss value is determined based on the sub-task loss value, and the global parameters of the CNN-LSTM meta-learning model are updated based on the global loss value. Specifically, this includes: S20431. Determine the subtask weight parameters of the corresponding meta-learning subtask based on the market share of each sample tire model. S20432. The sub-task loss values ​​are weighted and summed according to the sub-task weight parameters to obtain the global loss value; S20433. Update the global model parameters and global feature fusion weights of the CNN-LSTM meta-learning model using the backpropagation algorithm based on the global loss value.

[0041] Specifically, the subtask weight parameters of the corresponding meta-learning subtasks are determined based on the market share of each sample tire model. The higher the market share of a tire model, the greater the weight of the corresponding subtask, ensuring that the tire wear identification model is more accurate in identifying the wear level of most tires. The subtask loss values ​​are weighted and summed according to the subtask weight parameters to obtain the global loss value. The global model parameters and global feature fusion weights of the CNN-LSTM meta-learning model are updated according to the global loss value through the backpropagation algorithm.

[0042] It should be noted that each global parameter update is performed on the initial CNN-LSTM meta-learning model or the CNN-LSTM meta-learning model after the last global parameter update, rather than on the CNN-LSTM meta-learning sub-model.

[0043] It can be understood that in the embodiments of the present invention, the feature fusion weights corresponding to different tire models are updated in two stages of each iteration of training, as follows: Phase 1: Rapid Adaptation Phase for Subtasks (Partial Updates) For a specific tire model subtask, the model is updated 1-5 times using a small amount of labeled support set data (e.g., 5-10 samples) for that model. During this process, the weights of the feature fusion module are fine-tuned based on the data features of that tire model. For example, off-road tires will have their vibration spectrum features weighted more, while passenger car tires will have their image visual features weighted more. This update is local and only affects the temporary model parameters for the current subtask; it does not directly modify the global initial parameters of meta-learning.

[0044] Phase Two: Global Update of the Metamodel (Global Optimization) Collect the loss values ​​of the models after rapid adaptation for all subtasks on their respective query sets, perform a weighted sum of the losses for all subtasks, and backpropagate to update the global initial parameters of the meta-learning. This process indirectly optimizes the initial weight configuration of the feature fusion network, making the global initial parameters more suitable for rapid adaptation to different tire models. The globally updated initial parameters will serve as the starting point for the next round of subtask training.

[0045] The training process of the tire wear degree recognition model of the present invention has been described in detail above. The tire wear degree recognition model will automatically select the corresponding feature fusion weight according to the input current tire model, extract features from tire image data, wheel speed difference time series data, tire pressure time series data and suspension vibration spectrum respectively, and then perform feature fusion based on the feature fusion weight, thereby identifying the current tire wear degree of the current vehicle based on the fused features.

[0046] As an optional implementation, the real-time vehicle tire wear monitoring method further includes the following steps: S103. Determine whether the current tire wear level is greater than or equal to the preset safety threshold. S104. When the current tire wear is greater than or equal to the safety threshold, a warning is issued to the driver of the current vehicle.

[0047] Specifically, three warning thresholds can be preset based on the tire wear level: a reminder message is sent for light wear, a scheduled inspection is recommended for moderate wear, and an alarm is issued immediately for heavy wear, with tire replacement recommended. In addition, the warning rules can be dynamically adjusted based on the tire model's lifespan parameters; for example, the heavy wear threshold for high-performance tires can be appropriately lowered.

[0048] In some optional embodiments, warning information is displayed on the vehicle's central control screen and simultaneously pushed to the owner's mobile app; for severe wear, an audible and visual alarm is triggered. Furthermore, the time, wear level, and vehicle status of each warning can be recorded to generate a tire wear history report, providing the owner with a reference for tire replacement.

[0049] The method steps of the embodiments of the present invention have been described above. It can be understood that the embodiments of the present invention are based on meta-learning pre-training to obtain a tire wear degree recognition model that can adjust the feature fusion weights according to the tire model. By inputting the current vehicle's tire model, tire image data, wheel speed difference time series data, tire pressure time series data, and suspension vibration spectrum into the tire wear degree recognition model, the current tire wear degree of the current vehicle can be obtained, which improves the accuracy and real-time performance of vehicle tire wear degree monitoring, and is applicable to vehicles with different tire models.

[0050] Reference Figure 2 This invention provides a real-time vehicle tire wear monitoring device, comprising: The data acquisition module is used to acquire the current tire model, tire image data, wheel speed difference time series data, tire pressure time series data, and suspension vibration spectrum of the current vehicle. The wear degree recognition module is used to input the current tire model, tire image data, wheel speed difference time series data, tire pressure time series data and suspension vibration spectrum into the pre-trained tire wear degree recognition model to obtain the current tire wear degree of the current vehicle. Among them, the tire wear identification model is trained based on meta-learning, and the current tire model is used to determine the feature fusion weights of tire image data, wheel speed difference time series data, tire pressure time series data, and suspension vibration spectrum.

[0051] It is understood that the content of the above method embodiments is applicable to the present device embodiments. The specific functions implemented by the present device embodiments are the same as those of the above method embodiments, and the beneficial effects achieved are also the same as those achieved by the above method embodiments.

[0052] Reference Figure 3 This invention provides an electronic device, comprising: At least one processor; At least one memory for storing at least one program; When the above-mentioned at least one program is executed by the above-mentioned at least one processor, the above-mentioned at least one processor implements the above-mentioned method for real-time monitoring of vehicle tire wear.

[0053] It is understood that the content of the above method embodiments is applicable to this device embodiment. The specific functions implemented by this device embodiment are the same as those of the above method embodiments, and the beneficial effects achieved are also the same as those achieved by the above method embodiments.

[0054] This invention also provides a computer-readable storage medium storing a processor-executable computer program that, when executed by a processor, implements the above-described method for real-time monitoring of vehicle tire wear.

[0055] This invention provides a computer-readable storage medium that can execute a real-time vehicle tire wear monitoring method provided in the method embodiments of this invention. It can execute any combination of the implementation steps of the method embodiments and has the corresponding functions and beneficial effects of the method.

[0056] This invention also provides a computer program product, including a computer program that, when executed by a processor, implements the above-described method for real-time monitoring of vehicle tire wear.

[0057] It is understood that the content of the above method embodiments is applicable to the embodiments of this program product. The specific functions implemented by the embodiments of this program product are the same as those of the above method embodiments, and the beneficial effects achieved are also the same as those achieved by the above method embodiments.

[0058] Memory, as a non-transitory computer-readable storage medium, can be used to store non-transitory software programs and non-transitory computer-executable programs. Furthermore, memory may include high-speed random access memory, and may also include non-transitory memory, such as at least one disk storage device, flash memory device, or other non-transitory solid-state storage device. In some embodiments, memory may optionally include memory remotely located relative to the processor, and these remote memories can be connected to the processor via a network. Examples of such networks include, but are not limited to, the Internet, intranets, local area networks, mobile communication networks, and combinations thereof.

[0059] The embodiments described in this invention are for the purpose of more clearly illustrating the technical solutions of the embodiments of this invention, and do not constitute a limitation on the technical solutions provided by the embodiments of this invention. As those skilled in the art will know, with the evolution of technology and the emergence of new application scenarios, the technical solutions provided by the embodiments of this invention are also applicable to similar technical problems.

[0060] The terms "first," "second," "third," "fourth," etc. (if present) in the specification and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that embodiments of the invention described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0061] In some alternative embodiments, the functions / operations mentioned in the block diagrams may not occur in the order shown in the operation diagrams. For example, depending on the functions / operations involved, two consecutively shown blocks may actually be executed substantially simultaneously, or the aforementioned blocks may sometimes be executed in reverse order. Furthermore, the embodiments presented and described in the flowcharts of this invention are provided by way of example to provide a more comprehensive understanding of the technology. The disclosed methods are not limited to the operations and logic flows presented herein. Alternative embodiments are contemplated in which the order of various operations is changed and sub-operations described as part of a larger operation are executed independently.

[0062] Furthermore, although the invention has been described in the context of functional modules, it should be understood that, unless otherwise stated, one or more of the aforementioned functions and / or features may be integrated into a single physical device and / or software module, or one or more functions and / or features may be implemented in a separate physical device or software module. It is also understood that a detailed discussion of the actual implementation of each module is unnecessary for understanding the invention. Rather, given the properties, functions, and internal relationships of the various functional modules in the apparatus disclosed herein, the actual implementation of the module will be understood within the scope of conventional skill of an engineer. Therefore, those skilled in the art can implement the invention as set forth in the claims using ordinary techniques without excessive experimentation. It is also understood that the specific concepts disclosed are merely illustrative and not intended to limit the scope of the invention, which is determined by the full scope of the appended claims and their equivalents.

[0063] If the aforementioned functions are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this invention, or the part that contributes to the prior art, or a portion of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0064] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as a sequenced list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by, or in conjunction with, an instruction execution system, apparatus, or device (such as a computer-based system, a processor-including system, or other system that can fetch and execute instructions from, an instruction execution system, apparatus, or device). For the purposes of this specification, "computer-readable medium" can be any means that can contain, store, communicate, propagate, or transmit programs for use by, or in conjunction with, an instruction execution system, apparatus, or device.

[0065] More specific examples (a non-exhaustive list) of computer-readable media include: electrical connections (electronic devices) having one or more wires, portable computer disk drives (magnetic devices), random access memory (RAM), read-only memory (ROM), erasable and editable read-only memory (EPROM or flash memory), fiber optic devices, and portable optical disc read-only memory (CDROM). Furthermore, computer-readable media can even be paper or other suitable media on which the aforementioned program can be printed, because the aforementioned program can be obtained electronically, for example, by optically scanning the paper or other medium, followed by editing, interpreting, or otherwise processing as necessary, and then stored in computer memory.

[0066] It should be understood that various parts of the present invention can be implemented in hardware, software, firmware, or a combination thereof. In the above embodiments, multiple steps or methods can be implemented in software or firmware stored in memory and executed by a suitable instruction execution system. For example, if implemented in hardware, as in another embodiment, it can be implemented using any one or a combination of the following techniques known in the art: discrete logic circuits having logic gates for implementing logical functions on data signals, application-specific integrated circuits (ASICs) having suitable combinational logic gates, programmable gate arrays (PGAs), field-programmable gate arrays (FPGAs), etc.

[0067] In the foregoing description of this specification, references to terms such as "one embodiment," "another embodiment," or "some embodiments" indicate that a specific feature, structure, material, or characteristic described in connection with an embodiment or example is included in at least one embodiment or example of the present invention. In this specification, illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples.

[0068] Although embodiments of the invention have been shown and described, those skilled in the art will understand that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the claims and their equivalents.

[0069] The above is a detailed description of the preferred embodiments of the present invention. However, the present invention is not limited to the above embodiments. Those skilled in the art can make various equivalent modifications or substitutions without departing from the spirit of the present invention. All such equivalent modifications or substitutions are included within the scope defined by the claims of the present invention.

Claims

1. A method for real-time monitoring of vehicle tire wear, characterized in that, Includes the following steps: Acquire the current tire model, tire image data, wheel speed difference time series data, tire pressure time series data, and suspension vibration spectrum of the current vehicle; The current tire model, the tire image data, the wheel speed difference time series data, the tire pressure time series data, and the suspension vibration spectrum are input into a pre-trained tire wear recognition model to obtain the current tire wear level of the current vehicle. The tire wear identification model is trained based on meta-learning, and the current tire model is used to determine the feature fusion weights of the tire image data, the wheel speed difference time series data, the tire pressure time series data, and the suspension vibration spectrum.

2. The method for real-time monitoring of vehicle tire wear according to claim 1, characterized in that, The acquisition of the current tire model, tire image data, wheel speed difference time series data, tire pressure time series data, and suspension vibration spectrum of the current vehicle specifically includes: Obtain the current tire model input by the user; Tire image data is acquired by a camera device positioned above each tire of the current vehicle; The wheel speed is obtained by using wheel speed sensors installed on each tire of the current vehicle, and the wheel speed difference timing data is determined based on the wheel speed of the tires on the same axle. The tire pressure timing data is obtained by tire pressure sensors installed in each tire of the current vehicle; The suspension vibration signal is acquired by a vibration sensor installed in the suspension system of the current vehicle, and the suspension vibration spectrum is obtained by performing a Fourier transform on the suspension vibration signal.

3. The method for real-time monitoring of vehicle tire wear according to claim 1, characterized in that, The tire wear identification model is trained through the following steps: Training samples from multiple test vehicles are obtained, including sample tire models, tire image samples, wheel speed difference time series samples, tire pressure time series samples, and suspension vibration spectrum samples. The tire wear degree label corresponding to each training sample was determined by manual annotation. The sample tire models are divided into multiple meta-learning sub-tasks, and the support set and query set corresponding to each meta-learning sub-task are determined based on the training samples and the tire wear degree labels. A CNN-LSTM meta-learning model is constructed, and the CNN-LSTM meta-learning model is trained on each meta-learning sub-task based on the support set and the query set to obtain the trained tire wear degree recognition model.

4. The method for real-time monitoring of vehicle tire wear according to claim 3, characterized in that, The step of training the CNN-LSTM meta-learning model on each meta-learning subtask based on the support set and the query set to obtain the trained tire wear recognition model specifically includes: The CNN-LSTM meta-learning model is locally updated according to the support set of each meta-learning sub-task to obtain the CNN-LSTM meta-learning sub-model corresponding to each meta-learning sub-task; Determine the subtask loss value of each CNN-LSTM meta-learning sub-model on the query set of the corresponding meta-learning sub-task; The global loss value is determined based on the sub-task loss value, and the global parameters of the CNN-LSTM meta-learning model are updated based on the global loss value. Then, the process returns to the step of updating the local parameters of the CNN-LSTM meta-learning model based on the support set of each meta-learning sub-task. When the global loss value is lower than the preset loss threshold, training stops, and the trained tire wear recognition model is obtained.

5. The method for real-time monitoring of vehicle tire wear according to claim 4, characterized in that, The CNN-LSTM meta-learning model includes a first CNN branch network, a second CNN branch network, a first LSTM branch network, a second LSTM branch network, a feature fusion network, and a fully connected layer. The step of updating the local parameters of the CNN-LSTM meta-learning model according to the support set of each meta-learning sub-task to obtain the CNN-LSTM meta-learning sub-model corresponding to each meta-learning sub-task specifically includes: The current training samples contained in the support set of the current meta-learning subtask are input into the CNN-LSTM meta-learning model; The first CNN branch network, the second CNN branch network, the first LSTM branch network, and the second LSTM branch network are used to extract features from the tire image samples, the suspension vibration spectrum samples, the wheel speed difference time series samples, and the tire pressure time series samples corresponding to the current training samples, thereby obtaining tire image features, vibration spectrum features, wheel speed difference time series features, and tire pressure time series features. The CNN-LSTM meta-learning model determines the current sub-task feature fusion weights based on the sample tire model corresponding to the current meta-learning sub-task, and then performs feature fusion on the tire image features, vibration spectrum features, wheel speed difference time series features, and tire pressure time series features through the feature fusion network based on the current sub-task feature fusion weights to obtain a feature fusion vector. The fully connected layer maps the feature fusion vector into a predicted tire wear level. The current loss value is determined based on the predicted tire wear level and the tire wear level label corresponding to the current training sample; Based on the current loss value, the local model parameters and subtask feature fusion weights of the CNN-LSTM meta-learning model are updated using the gradient descent algorithm to obtain the CNN-LSTM meta-learning sub-model corresponding to each meta-learning sub-task.

6. The method for real-time monitoring of vehicle tire wear according to claim 4, characterized in that, The step of determining a global loss value based on the sub-task loss value and updating the global parameters of the CNN-LSTM meta-learning model based on the global loss value specifically includes: The subtask weight parameters corresponding to the meta-learning subtasks are determined based on the market share of each of the sample tire models. The global loss value is obtained by weighted summation of the sub-task loss values ​​based on the sub-task weight parameters. The global model parameters and global feature fusion weights of the CNN-LSTM meta-learning model are updated using the backpropagation algorithm based on the global loss value.

7. A method for real-time monitoring of vehicle tire wear according to any one of claims 1 to 6, characterized in that, The real-time monitoring method for vehicle tire wear also includes the following steps: Determine whether the current tire wear level is greater than or equal to a preset safety threshold; When the current tire wear level is greater than or equal to the safety threshold, a warning is issued to the driver of the current vehicle.

8. A device for real-time monitoring of vehicle tire wear, characterized in that, include: The data acquisition module is used to acquire the current tire model, tire image data, wheel speed difference time series data, tire pressure time series data, and suspension vibration spectrum of the current vehicle. The wear degree recognition module is used to input the current tire model, the tire image data, the wheel speed difference time series data, the tire pressure time series data, and the suspension vibration spectrum into a pre-trained tire wear degree recognition model to obtain the current tire wear degree of the current vehicle. The tire wear identification model is trained based on meta-learning, and the current tire model is used to determine the feature fusion weights of the tire image data, the wheel speed difference time series data, the tire pressure time series data, and the suspension vibration spectrum.

9. An electronic device, characterized in that, include: At least one processor; At least one memory for storing at least one program; When the at least one program is executed by the at least one processor, the at least one processor implements a method for real-time monitoring of vehicle tire wear as described in any one of claims 1 to 7.

10. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by the processor, it implements a method for real-time monitoring of vehicle tire wear as described in any one of claims 1 to 7.

Images