Method and device for monitoring tire pressure state of vehicle

Abstract

A method for monitoring a tire pressure condition is provided. The method comprises the following steps: receiving data for inferring a tire pressure state of a vehicle, wherein the data comprises audio data generated by a tire when the vehicle runs; and according to the received data, using a machine learning model to infer tire pressure states of the vehicle, the tire pressure states including normal tire pressure or abnormal tire pressure.

Description

Technical Field

[0001] The present disclosure generally relates to methods and apparatus for monitoring tire pressure status of a vehicle, and more particularly to techniques for monitoring tire pressure status of a vehicle. Background Art

[0002] The tire pressure of a vehicle is crucial to the normal safe driving of the vehicle and the service life of the tire. Too high tire pressure will reduce the friction between the tire and the road, affecting the braking effect of the vehicle, and will reduce the rolling resistance of the tire, making it easy for sharp objects on the road to pierce the tire and cause a blowout, thereby causing a traffic accident. Too high tire pressure will also accelerate the local wear of the pattern in the center of the tire tread, affecting the service life of the tire. Too low tire pressure will increase the friction between the tire and the road, resulting in increased fuel consumption, and will cause the tire temperature to rise sharply and the tire strength to decrease, which may lead to a blowout. Too low tire pressure will also accelerate the wear of the tire shoulder, easily cause cracks on the sidewall, and shorten the service life of the tire.

[0003] Therefore, a method and device for monitoring the tire pressure status of a vehicle are needed to ensure safe driving of the vehicle and extend the service life of the tire. Summary of the invention

[0004] A brief description of one or more aspects of the present disclosure is provided below to provide a basic understanding of these aspects. The present disclosure is not a broad overview of all aspects, nor is it intended to determine the key elements of all aspects, nor is it intended to delimit the scope of any or all aspects. Its sole purpose is to present some concepts of one or more aspects in a simplified form as a prelude to a more detailed description later.

[0005] According to one aspect of the present invention, a method for monitoring tire pressure status is provided. The method comprises: receiving data for inferring the tire pressure status of a vehicle, the data comprising audio data generated by the tire when the vehicle is traveling; and inferring the tire pressure status of the vehicle using a machine learning model based on the received data, the tire pressure status comprising normal tire pressure or abnormal tire pressure.

[0006] According to one aspect of the present invention, a device for monitoring tire pressure status is provided. The device comprises: a receiving module for receiving data for inferring the tire pressure status of a vehicle, the data comprising audio data generated by the tire when the vehicle is traveling; and an inference module for inferring the tire pressure status of the vehicle using a machine learning model based on the received data, the tire pressure status comprising normal tire pressure or abnormal tire pressure.

[0007] According to one aspect of the present invention, a device for monitoring tire pressure status is provided. The device includes: a memory; and a processor coupled to the memory. The processor is configured to receive data for inferring the tire pressure status of a vehicle, the data including audio data generated by the tire when the vehicle is traveling; and infer the tire pressure status of the vehicle using a machine learning model based on the received data, the tire pressure status including normal tire pressure or abnormal tire pressure.

[0008] According to one aspect of the present invention, a computer-readable medium is provided. The computer-readable medium stores a computer program code for monitoring tire pressure status. When the computer program code is executed by a processor, the processor receives data for inferring the tire pressure status of a vehicle, the data including audio data generated by the tire when the vehicle is traveling; and infers the tire pressure status of the vehicle using a machine learning model based on the received data, the tire pressure status including normal tire pressure or abnormal tire pressure.

[0009] According to one aspect of the present invention, a computer program product for monitoring tire pressure status is provided. The computer program product includes computer program code executable by a processor. The computer program code is used to receive data for inferring the tire pressure status of a vehicle, the data including audio data generated by the tire when the vehicle is traveling; and infer the tire pressure status of the vehicle using a machine learning model based on the received data, the tire pressure status including normal tire pressure or abnormal tire pressure.

[0010] It should be noted that one or more of the above aspects include the features described in detail below and specifically described in the claims. The following description and drawings elaborate on some exemplary features from multiple aspects. These features are merely indicative of the various ways in which the principles of various aspects can be implemented, and the present disclosure is intended to include all such aspects and their equivalents. BRIEF DESCRIPTION OF THE DRAWINGS

[0011] The following drawings describe various embodiments of the present disclosure for illustrative purposes only. Those skilled in the art will readily recognize from the following description that alternative embodiments of the methods and structures disclosed herein can be implemented without departing from the spirit and principles of the disclosure described herein.

[0012] Figure 1 A schematic diagram showing the proportion of noise generated by various components during vehicle driving.

[0013] Figure 2A-2B The pre-processed waveform diagrams of the audio signal generated by the tire under normal tire pressure state and abnormal tire pressure state are shown respectively.

[0014] Figure 3 A flow chart of a method for monitoring tire pressure status according to an embodiment of the present disclosure is shown.

[0015] Figures 4A-4C A schematic diagram showing relevant data for monitoring tire pressure status according to an embodiment of the present disclosure is shown.

[0016] Figure 5 A schematic diagram of a neural network model based on an attention mechanism for monitoring tire pressure status according to an embodiment of the present disclosure is shown.

[0017] Figure 6 A block diagram of a device for monitoring tire pressure status according to an embodiment of the present disclosure is shown.

[0018] Figure 7 A block diagram of a device for monitoring tire pressure status according to another embodiment of the present disclosure is shown. DETAILED DESCRIPTION

[0019] In the following description, many specific details are set forth to provide a thorough understanding of the embodiments of the present invention. However, those skilled in the relevant art will recognize that the present invention can be practiced without one or more of the specific details, or alternative methods, components, etc. can be used to practice the present invention. In some instances, well-known structures and operations are not shown or described in detail to avoid unnecessarily obscuring the present invention.

[0020] Figure 1 A schematic diagram showing the proportion of noise generated by various components of a vehicle during driving. Various components of a vehicle, such as the engine, transmission system, intake system, exhaust system, and tires, will generate some noise during driving. By analyzing these noises, such as Figure 1 As shown, the noise 105 generated by the tires during vehicle driving due to the friction between the tires and the road surface and the disturbance of the surrounding air accounts for about 50% of the total noise. The noise 110 generated by the exhaust system when purifying the exhaust gas and the exhaust gas passing through the exhaust pipe accounts for about 20% of the total noise. The noise 115 generated by the intake system when the air passes through the air filter, intake pipe, valve and other components accounts for about 12% of the total noise. The noise 120 generated by the engine when it is working due to combustion in the cylinder and mechanical movement of the piston accounts for about 10% of the total noise. The noise generated by the vehicle transmission system due to friction and mechanical movement between components accounts for about 5% of the total noise. Other noises generated by the vehicle when it is driving account for about 3% of the total noise. From Figure 1 It can be seen from the analysis that tire noise is the main component of the noise generated when the vehicle is running. Therefore, it is easy to get the tire noise signal from the noise generated when the vehicle is running.

[0021] Figure 2A and 2B The preprocessed waveforms of the tire noise signal under normal tire pressure state and abnormal tire pressure state are shown respectively. Figure 2A The figure shows a partial waveform of the sound signal or audio signal generated by the tire under normal tire pressure (e.g., between 2.2 and 2.4 bar) collected by the sound sensor during vehicle driving, after preprocessing (e.g., denoising and padding to smooth the waveform), where the horizontal axis is the time axis and the vertical axis shows the amplitude of the audio signal (i.e., tire noise signal). Figure 2A It can be seen that the waveform of the sound signal generated when the tire is running under normal tire pressure is relatively continuous, and the amplitude change of the waveform is relatively stable. Figure 2B The figure shows a partial waveform of the sound signal generated by the tire under abnormal tire pressure (for example, 1.6 bar) collected by the sound sensor after preprocessing. Similarly, the horizontal axis is the time axis and the vertical axis is the amplitude of the tire sound signal. Figure 2B It can be seen that the waveform of the sound signal generated when the tire is running under abnormal tire pressure is intermittent and continuous, and the amplitude of the waveform varies greatly. Figure 2A and Figure 2B It can be seen from the comparison that the waveforms of the sound signals generated when the tire is running under normal and abnormal air pressure conditions have different characteristics.

[0022] Although the noise signal generated by the tire during driving contains information such as whether the tire pressure is normal, the waveform of the tire noise signal may also be affected by various other factors and change. For example, the waveform of the sound signal generated by tires of different materials and models under different road or weather conditions and under different working conditions such as vehicle speed or load will also be different. Therefore, if it is determined whether the tire pressure is within the normal range according to fixed rules, such as by analyzing certain fixed features of the audio signal waveform generated by the tire, it may be difficult to obtain correct and reliable results in all the above situations.

[0023] The present disclosure provides a method for inferring tire pressure state by using a machine learning model based on at least audio data generated by tires during vehicle driving. In some embodiments, a large amount of audio data generated by various types of vehicles and tires in various driving states and / or driving environments can be collected to perform unsupervised learning on a machine learning model (e.g., a deep neural network). It is also possible to collect a large amount of audio data generated by various types of vehicles and tires in various driving states and / or driving environments and corresponding tire pressure and / or air pressure state (i.e., normal or abnormal) data to perform supervised learning on the machine learning model. In some embodiments, fixed parameter data of the relevant vehicle (e.g., length, width and height dimensions of the vehicle, suspension type, tire size, etc.), various working condition data of the vehicle during driving (e.g., engine output power, torque, wheel speed, vehicle speed, acceleration, attitude angle, steering wheel steering angle, wheel height, etc.), and working environment data of the vehicle (e.g., temperature, humidity, altitude, geographical location, etc.) can also be collected, and these data together with the collected tire audio data are used as a training data set to train the machine learning model to improve the accuracy and reliability of inferring the tire pressure state. In some embodiments, a machine learning model based on an attention mechanism can be used. Through this attention mechanism, different weights can be applied to various input data during the inference process, thereby effectively utilizing the input data, improving the applicability of machine learning models in various application scenarios, and further improving the accuracy of inference results.

[0024] Figure 3 A flow chart of a method 300 for monitoring tire pressure status according to an embodiment of the present disclosure is shown. The method 300 can be executed on the vehicle side. The method 300 can also be executed on the cloud, for example, on a server side that is in communication with the vehicle. The server can send the monitored tire pressure status, such as normal tire pressure or abnormal tire pressure, to the vehicle via a communication network. The server can also send a tire pressure abnormality message to the vehicle only when an abnormal tire pressure status is detected, so as to save communication network overhead.

[0025] In box 310, method 300 may receive data for inferring the tire pressure state of the vehicle. The data at least includes audio data generated by the tire when the vehicle is running. The audio data generated by each tire when running can be collected separately by a sound sensor installed near each wheel. The noise generated during the vehicle's running can also be collected by an existing sound system on the vehicle, such as a microphone (including a sound sensor), and then the noise data generated by each tire in different directions can be obtained by acoustic processing. Figure 1As shown in , the main part of the noise generated by the vehicle during driving is tire noise, so the sound data collected by the sound sensor can be directly used as tire sound data. In addition, in order to reduce the interference of other noise sources to improve the accuracy of tire pressure monitoring, a frequency filter can be used to perform frequency domain filtering on the original audio data received from one or more sound sensors to obtain the audio data generated by the tire. The frequency filter can be a bandpass filter corresponding to the tire noise frequency band, so that the noise signal in the frequency band outside the tire noise frequency band can be filtered out by performing frequency domain filtering on the vehicle noise signal.

[0026] As above combined Figure 2A and 2B As described above, the audio signal generated by the tire when driving can reflect the state of the tire pressure, but whether the tire pressure state is normal is also related to factors such as the parameters of the vehicle, the working condition of the vehicle, and the driving environment. For example, for a heavier truck, the normal tire pressure range is higher, while for a lighter family car, the normal tire pressure range is lower. When the vehicle is accelerating or turning sharply, the pressure on some wheels will also increase, and the normal tire pressure range will also increase accordingly. In addition, in rainy and snowy weather, the road surface is slippery, and in order to increase friction, the normal tire pressure range can be lower. Therefore, in order to more accurately and reasonably monitor the tire pressure state and avoid false alarms (for example, mistaking the instantaneous tire pressure exceeding the normal range due to the driving state of the vehicle for abnormal tire pressure), the data received in box 310 for inferring the tire pressure state of the vehicle can also include one or more of the vehicle's fixed parameter data, the vehicle's working condition data, and the vehicle's working environment data.

[0027] The fixed parameter data of the vehicle may include one or more of the vehicle's length, width and height, track width, wheelbase, vehicle weight, vehicle suspension type, and vehicle tire size. Vehicle weight data can be used to adjust the normal range of tire pressure. The various dimensions of the vehicle and the tire size can be used to calculate the dynamic dimensions of the tire in combination with the vehicle's driving status data. For example, the dynamic radius (i.e., the actual radius) of the tire under stress will be smaller than the nominal radius of the tire. The calculated dynamic dimensions of the tire can also be further used to adjust the normal range of tire pressure. The vehicle's suspension type may include, for example, McPherson suspension, multi-link suspension, torsion beam suspension, air spring suspension, and the like. Figure 4AA schematic diagram of an independent suspension is shown, wherein 410 is a coil spring, 412 is an upper suspension arm, and 414 is a lower suspension arm. As shown in the figure, in the case of independent suspension, the instantaneous tire pressure of the tire 420 will change when it is on a bumpy road (for example, running over a stone), but it will not have or will only have a relatively small impact on the tire pressure of the tire 430 that is driving normally. In addition, different suspension types will also have different effects on the changes in the air pressure of each tire of the vehicle under different driving conditions (for example, sharp turns). In one embodiment, these suspension types can be converted into specific values ​​by table mapping to assist in monitoring whether the tire pressure state is normal.

[0028] The working condition data of the vehicle may include one or more of the engine output power of the vehicle, the torque of the vehicle, the speed of the vehicle, the wheel speed, the acceleration of the vehicle, the attitude angle of the vehicle, the steering angle of the steering wheel, and the height of the wheel. The working status data here refers to the real-time parameter data of the vehicle during operation, rather than the fixed parameter data of the vehicle. For example, the engine output power refers to the actual current output power of the engine, rather than the maximum output power of the engine. The torque of the vehicle refers to the torque that each wheel currently actually bears, rather than the maximum torque of the vehicle. The engine output power and the torque of the wheel can be used to calculate the acceleration of the vehicle, or directly used as input data for the machine learning model. The speed of the vehicle can be obtained through the vehicle positioning system (e.g., GPS, Beidou, etc.). The wheel speed of each wheel can be obtained through the wheel speed sensor of each wheel. The wheel speed can be used to calculate the vehicle speed.

[0029] Figure 4B Schematic diagram showing various attitude angles of a vehicle for monitoring tire pressure status according to an embodiment of the present disclosure. The attitude angle of the vehicle may include a yaw angle, a tilt angle, and a pitch angle. Figure 4B The vehicle coordinate system shown in Figure 1 is the origin of the vehicle coordinate system, which can be the center of gravity of the vehicle. The x-axis is parallel to the vehicle body and points to the front of the vehicle. The y-axis is perpendicular to the left and right of the vehicle body and points to the left. The z-axis is perpendicular to the top of the vehicle body and points to the top of the vehicle. The vehicle yaw angle refers to the angle of rotation of the vehicle along the z-axis, the tilt angle refers to the angle of rotation of the vehicle along the x-axis, and the pitch angle refers to the angle of rotation of the vehicle along the y-axis. The vehicle's attitude angle data also includes the yaw angular velocity ω. r , tilt angular velocity ω p and the pitch angular velocity ω q. The vehicle's attitude angle data can be obtained through attitude sensors, such as yaw angle sensors, inclination angle sensors, and pitch angle sensors. These attitude angle data will have different effects on the vehicle's tire pressure. For example, the greater the vehicle's pitch angle indicates that the vehicle is leaning forward, the more the tire pressure of the vehicle's front wheels will increase than during normal driving, and the tire pressure of the vehicle's rear wheels will decrease more than during normal driving. The greater the vehicle's roll angle indicates that the vehicle is leaning to the left, the more the tire pressure of the vehicle's left wheels will increase than during normal driving, and the tire pressure of the vehicle's right wheels will decrease more than during normal driving. The vehicle's attitude angle can also be combined with the vehicle speed or wheel speed to calculate the friction coefficient between the tire and the road.

[0030] Figure 4B The speed components of the vehicle in different directions are also shown. For example, the speed of the vehicle may include a forward speed u, a lateral speed v, and a vertical speed w. The different speed components of the vehicle may also have different effects on the air pressure of different tires of the vehicle.

[0031] The steering wheel steering angle can be obtained by a steering wheel sensor, which can be used to calculate the vehicle's yaw angle.

[0032] Figure 4C A schematic diagram of wheel height for monitoring tire pressure status according to an embodiment of the present disclosure is shown. Figure 4C Schematic diagrams showing the tire status under different wheel heights h1, h2 and h3 are shown respectively. On the one hand, during normal driving or at rest, the wheel height can directly reflect the tire pressure status. For example, when the wheel height is h1, the tire pressure may be too high, when the wheel height is h2, the tire pressure may be normal, and when the wheel height is h3, the tire pressure may be too low. On the other hand, changes in wheel height caused by instantaneous changes in the road surface or vehicle body posture during vehicle driving will also affect the tire pressure status. For example, when the wheel height suddenly increases from h2 to h1 during vehicle driving, the tire pressure of the wheel may decrease. When the wheel height suddenly decreases from h2 to h3 during vehicle driving, the tire pressure of the wheel may increase.

[0033] Back to Figure 3The data received in box 310 for inferring the tire pressure state of the vehicle may also include the working environment data of the vehicle. The working environment data of the vehicle may include one or more of the ambient temperature, ambient humidity, altitude, geographical location, and road conditions. The temperature and humidity data of the vehicle's driving environment may be obtained through temperature and humidity sensors, or through meteorological information received via a communication network. The altitude and geographical location of the vehicle may be obtained through a positioning system. The altitude and geographical location data may also be used to infer the climatic conditions (e.g., temperature, humidity, etc.) and road conditions (e.g., highways, urban roads, rural roads, etc.) of the vehicle's driving environment.

[0034] The vehicle fixed parameter data, working condition data, and working environment data received in block 310 for inferring the tire pressure state of the vehicle can assist the audio data generated by the tire to improve the accuracy of tire pressure state monitoring. In addition, when certain data cannot be obtained due to sensor failure or vehicle communication failure, the other received data can be used as a supplement or replacement to ensure that relatively high accuracy tire pressure state monitoring is continued, thereby improving the robustness of the tire pressure monitoring method.

[0035] In box 320, method 300 may perform a preprocessing operation on the received data. In one embodiment, data preprocessing may include deleting abnormal data and / or noise data from the received data. Due to various reasons during data acquisition and data transmission, the received data may contain data that interferes with subsequent data analysis, that is, noise data. In addition, data generated by the vehicle under certain abnormal conditions may also lead to erroneous analysis results. For example, when a vehicle passes through a bumpy road at high speed, the tire pressure may have abnormal values, and the tires may also generate abnormal audio data. These data may also affect the accurate monitoring of the tire pressure status. Therefore, deleting abnormal data and / or noise data from the received data can improve the accuracy of tire pressure status monitoring.

[0036] In another embodiment, data preprocessing may include data padding of the received data. During vehicle driving, certain data for monitoring tire pressure status may not be obtained due to sensor or communication line failure. In order to still be able to monitor the tire pressure status normally and reliably in this case, data padding may be performed on the received data. For example, in an embodiment in which a machine learning model for inferring tire pressure status requires the vehicle yaw angle as input data, when the vehicle's yaw angle sensor fails, the steering angle data obtained from the steering wheel sensor may be used to calculate and fill in the yaw angle data. For another example, in an embodiment in which a machine learning model for inferring tire pressure status requires wheel speed as input data, when the latest wheel speed data cannot be obtained due to a communication failure, the previously obtained historical wheel speed data may be used to infer and fill in the current wheel speed data.

[0037] In another embodiment, data preprocessing may include data merging of received data with collinearity. The received data for inferring the tire pressure state may contain data with a linear relationship with each other. For example, the output power of the engine and the torque of the wheel may have collinearity. The vehicle speed, wheel speed, and acceleration may have collinearity. The yaw angle and steering wheel steering angle of the vehicle may have collinearity. By merging the data with collinearity in the received data, the data dimension input to the machine learning model for inferring the tire pressure state can be reduced, thereby reducing the complexity of the machine learning model and increasing the inference speed. Data preprocessing may also include any combination of the above-mentioned embodiments. For example, preprocessing operations such as denoising and padding can be performed on the audio data of the sound signal generated by the collected tire during driving to make the waveform of the sound signal generated by the tire as smooth as possible.

[0038] At box 330, method 300 may use a machine learning model to infer the tire pressure status of the vehicle based on the preprocessed data. The inferred tire pressure status may include normal tire pressure or abnormal tire pressure. Depending on different machine learning models, the inferred tire pressure status may also include more grades. In some other embodiments, method 300 may also skip box 320 and directly use a machine learning model to infer the tire pressure status of the vehicle based on the data received in box 310. In one embodiment, the machine learning model may be a deep neural network model. The specific structure of the deep neural network can be set and adjusted according to different design requirements and the type of input data. In another embodiment, the machine learning model may be a neural network model based on an attention mechanism.

[0039] Figure 5A schematic diagram of a neural network model 500 based on an attention mechanism for monitoring tire pressure status according to an embodiment of the present disclosure is shown. The neural network model 500 may receive one or more of the audio data 510 generated by the tire when driving, the fixed parameter data 520 of the vehicle, the working condition data 530 of the vehicle, and the working environment data 540 of the vehicle as input data. The input data of the neural network model 500 may also be one or more of the data 510-540 that have undergone the preprocessing operation described above.

[0040] The neural network model 500 may include an input module 550. The input module 550 may include an input layer and several hidden layers. The number of neurons in the input layer may depend on the type and dimension of the data used to infer the tire pressure state. Different numbers of hidden layers may be selected in different specific implementations. The neural network model 500 may also include an attention module 560. The attention module 560 may set different weights for different data used to infer the tire pressure state in different inference processes. For example, the attention module 560 may use a softmax function to calculate the weights of different data used to infer the tire pressure state. By utilizing the attention mechanism, the input data may be more effectively analyzed and utilized, thereby inferring the tire pressure state more accurately. The neural network model 500 may also include an output module 570. The output module 570 may include an output layer. The output layer may be a decision layer. The output module 570 may also include several hidden layers located before the output layer. Through the output module 570, the neural network 500 may infer the tire pressure state of the vehicle based on the received data 510-540, and output an inference result 580, for example, normal tire pressure or abnormal tire pressure.

[0041] The machine learning model described in the present disclosure (e.g., a neural network model based on an attention mechanism) is a trained model and can be updated regularly or irregularly. The training data set used to train the machine learning model may include driving audio data and air pressure status data of tires of various models and sizes measured for various models of vehicles under various working conditions and driving environments. After the machine learning model is trained, the trained machine learning model can be verified using a test data set. The test data set may be data different from the training data set acquired in the same manner as the training data set. When the accuracy of the machine learning model on the test data set meets the design requirements, the machine learning model can be deployed to the vehicle or to a platform server that provides tire pressure status monitoring services. The platform server can continue to collect new data during the use of the machine learning model, and provide these big data to the server used to train the machine learning model for updating the machine learning model, so as to continuously improve the performance of the machine learning model.

[0042] Figure 6 A block diagram of an apparatus 600 for monitoring tire pressure status according to an embodiment of the present disclosure is shown. The apparatus 600 may include a receiving module 610. The receiving module 610 may receive data for inferring the tire pressure status of a vehicle. The received data may include audio data generated by the tire when the vehicle is traveling. The received data may also include fixed parameter data of the vehicle (e.g., length, width, height, suspension type, tire size, etc.), operating condition data of the vehicle (e.g., engine output power, wheel torque, vehicle speed, wheel speed, acceleration, attitude angle, steering wheel steering angle, wheel height, etc.), and operating environment data of the vehicle (e.g., ambient temperature, ambient humidity, geographic location, altitude, etc.).

[0043] The apparatus 600 may further include a preprocessing module 620 for performing preprocessing operations on the received data, for example, deleting abnormal data and noise data from the received data; performing data padding on the received data; and / or performing data merging on the received data with collinearity, etc. The preprocessing module 620 is optional. In some other embodiments, the apparatus 600 may also not include the preprocessing module 620. Figure 6 As not shown in the figure, the device 600 may also include a frequency filter for performing frequency domain filtering on the original audio data received from one or more sound sensors to obtain the audio data generated by the tire. Figure 1 As described, raw audio data received from one or more sound sensors may also be used as tire generated audio data.

[0044] The apparatus 600 may further include an inference module 630. The inference module 630 may use a machine learning model (e.g., in combination with a machine learning model) based on the received data or the pre-processed received data. Figure 5 The neural network model based on the attention mechanism described in the present invention is used to infer the tire pressure status of the vehicle. The tire pressure status includes normal tire pressure or abnormal tire pressure.

[0045] Figure 7 FIG. 1 is a block diagram of an apparatus 700 for monitoring tire pressure status according to another embodiment of the present disclosure. The apparatus 700 may include a memory 710 and a processor 720 coupled to the memory 710, for example, via an internal communication bus. The processor 720 may include one or more processors or processor cores. The processor 720 may execute the computer program code in the memory 710 to implement the above-mentioned combination Figure 3 A method for monitoring tire pressure status is described in detail. Examples of processor 720 may include a central processing unit (CPU), a microcontroller, etc. Memory 710 suitable for tangibly embodying computer program code and data may include various forms of memory, such as EPROM, EEPROM, and flash memory devices, etc. Although in Figure 7 Not shown, the device 700 may also include an input interface and an output interface. The input interface may receive input signals and data, including, for example, data for inferring the tire pressure state. The output interface may output the inferred tire pressure state to the user. For example, when the tire pressure is abnormal, the output interface may prompt the user by means of an alarm sound.

[0046] In one embodiment of the present disclosure, a computer program product for monitoring tire pressure status may include a processor executable computer code for performing the above method. In another embodiment of the present disclosure, a computer readable medium may store a computer program code for monitoring tire pressure status. The computer program code, when executed by a processor, may cause the processor to perform the above method. A computer program may include instructions or programs executable by a computer. The instructions or programs may be recorded on any data storage medium including a memory. For example, the program may be implemented in a digital electronic circuit, or in computer hardware, firmware, software, or a combination thereof. The process / method steps described in the present disclosure may be performed by a programmable processor that executes program instructions to perform methods, steps, operations by operating on input data and generating outputs.

[0047] In addition to what is described herein, various modifications may be made to the disclosed embodiments and implementations of the present invention without departing from the scope of the disclosed embodiments and implementations of the present invention. Therefore, the descriptions and examples herein should be interpreted as illustrative rather than limiting. The scope of the present invention should be measured only by reference to the claims.

Claims

1. A method for monitoring tire pressure status, comprising: receiving data for inferring tire pressure status of a vehicle, the data comprising audio data generated by the tire when the vehicle is traveling; Based on the received data, a machine learning model is used to infer the tire pressure status of the vehicle, where the tire pressure status includes normal tire pressure or abnormal tire pressure.

2. The method according to claim 1, wherein: The tire-generated audio data is obtained by frequency-domain filtering of raw audio data received from one or more sound sensors.

3. The method according to claim 1, wherein: The data for inferring the tire pressure state of the vehicle also includes at least one of the fixed parameter data of the vehicle, the working condition data of the vehicle, and the working environment data of the vehicle. The fixed parameter data of the vehicle includes at least one of the length, width and height of the vehicle, the suspension type of the vehicle, and the tire size of the vehicle; The vehicle operating condition data includes at least one of the engine output power of the vehicle, the torque of the vehicle, the speed of the vehicle, the acceleration of the vehicle, the attitude angle of the vehicle, the steering angle of the steering wheel of the vehicle, and the wheel height of the vehicle; The vehicle's operating environment data includes at least one of ambient temperature, ambient humidity, and altitude.

4. The method according to claim 1 further comprises performing at least one of the following pre-processing operations on the received data: Remove abnormal data and noise data from the received data; Performing data padding on the received data; Data merging is performed on the received data having collinearity.

5. The method according to claim 1, wherein: The machine learning model is a neural network model based on the attention mechanism.

6. A device for monitoring tire pressure status, comprising: A receiving module, used for receiving data for inferring the tire pressure state of the vehicle, wherein the data includes audio data generated by the tire when the vehicle is traveling; The inference module is used to infer the tire pressure status of the vehicle using a machine learning model based on the received data, and the tire pressure status includes normal tire pressure or abnormal tire pressure.

7. The apparatus according to claim 6, further comprising a frequency filter for performing frequency domain filtering on raw audio data received from one or more sound sensors to obtain the audio data generated by the tire.

8. The device according to claim 6, wherein: The data for inferring the tire pressure state of the vehicle also includes at least one of the fixed parameter data of the vehicle, the working condition data of the vehicle, and the working environment data of the vehicle. The fixed parameter data of the vehicle includes at least one of the length, width and height of the vehicle, the suspension type of the vehicle, and the tire size of the vehicle; The vehicle operating condition data includes at least one of the engine output power of the vehicle, the torque of the vehicle, the speed of the vehicle, the acceleration of the vehicle, the attitude angle of the vehicle, the steering angle of the steering wheel of the vehicle, and the wheel height of the vehicle; The vehicle's operating environment data includes at least one of ambient temperature, ambient humidity, and altitude.

9. The apparatus according to claim 6, further comprising a preprocessing module, configured to perform at least one of the following preprocessing operations on the received data: Remove abnormal data and noise data from the received data; Performing data padding on the received data; Data merging is performed on the received data having collinearity.

10. The device according to claim 6, wherein: The machine learning model is a neural network model based on the attention mechanism.

11. A device for monitoring tire pressure status, comprising: Memory; A processor is coupled to the memory and configured to execute the method according to any one of claims 1-5.

12. A computer readable medium storing a computer program code for monitoring tire pressure status, wherein when the computer program code is executed by a processor, the processor executes the method according to any one of claims 1 to 5.

13. A computer program product for monitoring tire pressure status, comprising a computer program code executable by a processor, wherein the computer program code is used to execute the method according to any one of claims 1-5.

Images