A noise warning method and system during vehicle travel

Abstract

The application discloses a noise early warning method and system in vehicle driving process, comprising the following steps: obtaining a real-time noise map of a driving route of the vehicle from a vehicle network; obtaining a real-time noise value of a next section according to the real-time noise map; obtaining a section distance and a traffic flow of the next section through navigation software; and obtaining a vehicle self-noise value through a noise sensor; calculating a noise safety index value of the next section according to the real-time noise value of the next section, the vehicle self-noise value, the section distance of the next section and the traffic flow of the next section; and sending an early warning information to a driver when the noise safety index value exceeds a noise safety index threshold value preset by the driver. The application realizes quantitative monitoring of travel noise safety and guarantees the travel noise safety of the driver.

Description

Technical Field

[0001] This invention relates to the field of noise warning technology, and in particular to a noise warning method and system for vehicles in motion. Background Technology

[0002] Vehicles, as a common means of transportation in modern society, are closely related to people's daily lives. Urban noise pollution in my country is severe, and its impact is constantly expanding in scope and intensity. Urban noise mainly includes traffic noise, industrial noise, construction noise, and social noise. Noise pollution is random, localized, and dispersed, and is ubiquitous in cities. Noise not only affects the driver's travel experience and can even distract the driver, affecting their concentration and creating potential traffic risks, but it also affects the travel experience of pedestrians and passengers in vehicles. Existing systems that provide driver warnings based on real-time noise maps are few and far between, failing to provide adequate protection in this regard.

[0003] Existing in-vehicle systems typically have the following drawbacks:

[0004] (1) Existing in-vehicle systems cannot provide drivers with warnings about noise pollution they will encounter while driving, or allow them to change routes or take precautions, based on real-time noise maps, vehicle noise levels, traffic flow, and road length. This results in severe noise pollution affecting the driver's driving experience and even significantly impacting their concentration, increasing potential traffic risks.

[0005] (2) Existing vehicle systems have not yet calculated an index value that measures the noise environment in which the driver is located by combining the noise value of the noise map, road information, vehicle noise value and other indicators. Summary of the Invention

[0006] This invention provides a noise warning method and system for vehicles in motion, enabling quantitative monitoring of travel noise safety and ensuring driver safety. By acquiring real-time noise levels of the next road segment, the vehicle's own noise level, the distance to the next road segment, and the traffic flow of the next road segment, this invention calculates a travel noise safety index value reflecting the driver's perception of the route's noise environment, thus achieving quantitative monitoring of travel noise safety and ensuring driver safety.

[0007] To address the aforementioned technical problems, embodiments of the present invention provide a noise warning method for vehicles in motion, comprising:

[0008] The system obtains a real-time noise map of the current vehicle route from the Internet of Vehicles, obtains the real-time noise value of the next road segment based on the real-time noise map, obtains the road segment distance and traffic flow of the next road segment through navigation software, and obtains the vehicle's own noise value through noise sensors.

[0009] The noise safety index value of the next road segment is calculated based on the real-time noise value of the next road segment, the vehicle's own noise value, the road segment distance of the next road segment, and the traffic flow of the next road segment.

[0010] When the noise safety index value exceeds the driver's preset noise safety index threshold, a warning message is issued to the driver.

[0011] As one preferred embodiment, the step of calculating the noise safety index value of the next road segment based on the real-time noise value of the next road segment, the vehicle's own noise value, the road segment distance of the next road segment, and the traffic flow of the next road segment specifically involves:

[0012] The real-time noise value of the next road segment, the vehicle's own noise value, the road segment distance of the next road segment, and the road segment traffic flow of the next road segment are input into the noise safety index prediction model based on the convolutional neural network. The noise safety index prediction model outputs the noise safety index value of the next road segment. The noise safety index prediction model includes an embedding layer, a convolutional layer, a pooling layer, and three fully connected layers.

[0013] As one preferred embodiment, at the end of the trip, the driver scores the noise level of each road segment and calculates the error value using a loss function, and adjusts the parameters of each level of the noise safety index prediction model based on the convolutional neural network according to the error value.

[0014] As one preferred embodiment, the calculation of the error value using the loss function specifically involves:

[0015] The error value is calculated using the following formula:

[0016]

[0017] Where S is the noise safety index value of the road segment obtained in the fully connected layer, and its value is an integer ranging from 1 to 10; C is a constant representing the total number of different values ​​of S; X is the feature vector; y XS The sign function takes a value of 0 when the noise score of the feature vector X equals S, and a value of 1 otherwise; p KS The output of the second fully connected layer represents the predicted probability that X belongs to S; N is the number of input feature vectors in this round of error analysis.

[0018] As one preferred solution, a warning message is issued to the driver, specifically as follows:

[0019] A warning message is sent to the driver to prompt him to change the driving route and recalculate the noise safety index value for the next road segment.

[0020] As one preferred approach, the input feature vector is preprocessed using one-hot encoding in the embedding layer.

[0021] Another embodiment of the present invention provides a noise warning system for vehicles in motion, including a user module, a noise sensor module, a vehicle-mounted module, an analysis module, and a database module;

[0022] The user module is used to preset noise safety index thresholds and perform noise scoring, and input the noise safety index thresholds and noise scores into the analysis module;

[0023] The noise sensor module is used to calculate the vehicle's own noise level.

[0024] The vehicle infotainment module is used to obtain the vehicle's own noise value from the noise sensor module, obtain the real-time noise value of the next road segment from the real-time noise map, obtain the road segment distance and traffic flow of the next road segment from the navigation software, input them into the analysis module, and dynamically adjust the calculation strategy of the noise safety index value.

[0025] The analysis module is used to input the real-time noise value of the next road segment, the vehicle's own noise value, the road segment distance of the next road segment, and the traffic flow of the next road segment into the noise safety index prediction model based on the convolutional neural network to obtain the noise safety index value of the next road segment. It is also used to determine whether to issue a warning message to the driver based on the noise safety index value and the noise safety index threshold, as well as to perform error scoring after the trip and adjust the parameters of the noise safety index prediction model based on the convolutional neural network.

[0026] The database module is used to store the neural network parameters of different drivers.

[0027] Compared with the prior art, the beneficial effects of the embodiments of the present invention are at least one of the following:

[0028] This invention can assess the noise safety of upcoming road sections in advance based on real-time dynamic noise maps, the vehicle's own noise level, and information from other road segments. It can calculate the appropriate noise safety index value for each driver based on their individual circumstances. Furthermore, drivers can rate the noise safety index values ​​for each road segment during their trip, helping to dynamically adjust the correlation coefficients in the noise safety index calculation formula and improve the subsequent user experience. Attached Figure Description

[0029] Figure 1This is a flowchart illustrating a noise warning method based on vehicle driving process in one embodiment of the present invention.

[0030] Figure 2 This is a schematic diagram of the structure of a noise warning system during vehicle operation in one embodiment of the present invention;

[0031] Figure 3 This is a schematic diagram of the structure of the noise safety index prediction model of the noise warning method during vehicle operation in one embodiment of the present invention. Detailed Implementation

[0032] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. The purpose of providing these embodiments is to make the disclosure of the present invention more thorough and comprehensive. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present invention.

[0033] In the description of this application, the terms "first," "second," "third," etc., are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features. Therefore, a feature defined with "first," "second," "third," etc., may explicitly or implicitly include one or more of that feature. In the description of this application, unless otherwise stated, "a plurality of" means two or more.

[0034] In the description of this application, it should be noted that, unless otherwise expressly specified and limited, the terms "installation," "connection," and "linking" should be interpreted broadly. For example, they can refer to fixed connections, detachable connections, or integral connections; they can refer to mechanical connections or electrical connections; they can refer to direct connections or indirect connections through an intermediate medium; and they can refer to the internal communication between two components. The terms "vertical," "horizontal," "left," "right," "upper," "lower," and similar expressions used herein are for illustrative purposes only and do not indicate or imply that the device or component referred to must have a specific orientation, or be constructed and operated in a specific orientation, and therefore should not be construed as limiting the invention. The term "and / or" as used herein includes any and all combinations of one or more of the associated listed items. Those skilled in the art can understand the specific meaning of the above terms in this application based on the specific circumstances.

[0035] In the description of this application, it should be noted that, unless otherwise defined, all technical and scientific terms used in this invention have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains. The terminology used in this specification is for the purpose of describing specific embodiments only and is not intended to limit the invention. Those skilled in the art can understand the specific meaning of the above terms in this application based on the specific circumstances.

[0036] One embodiment of the present invention provides a noise warning method for vehicles in motion; please refer to [link / reference]. Figure 1 , Figure 1 The diagram shown is a flowchart illustrating a noise warning method for a vehicle during operation, according to one embodiment of the present invention, which includes:

[0037] Step S1: Obtain a real-time noise map of the current driving route from the vehicle network. Based on the real-time noise map, obtain the real-time noise value of the next road segment. Obtain the road segment distance and traffic flow of the next road segment through the navigation software. Obtain the vehicle's own noise value through the noise sensor. Specifically, the driver can obtain the neural network parameters stored in the database through the vehicle network, and can also view the preset noise safety index threshold in the database; if the driver's preset noise safety index threshold cannot be obtained, the driver will be prompted to set the noise safety index threshold.

[0038] As the vehicle approaches the next road segment, it obtains the real-time noise value of the next road segment, the vehicle's own noise value, the distance to the road segment, and the traffic flow of the road segment through the real-time noise map.

[0039] Step S2: Obtain the real-time noise map of the current vehicle driving route from the vehicle network, and obtain the real-time noise value of the next road segment, the vehicle's own noise value, the road segment distance of the next road segment, and the traffic flow of the next road segment based on the real-time noise map.

[0040] As one embodiment, the real-time noise value of the next road segment, the vehicle's own noise value, the road segment distance of the next road segment, and the traffic flow of the next road segment are input into a noise safety index prediction model based on a convolutional neural network. The noise safety index prediction model outputs the noise safety index value of the next road segment. Please refer to [link to relevant documentation]. Figure 3The noise safety index prediction model includes an embedding layer, a convolutional layer, a pooling layer, and three fully connected layers. The two fully connected layers have 64 and 10 neurons, respectively. The embedding layer performs a one-dimensional transformation on the one-dimensional feature vector. This preprocessing uses word2vecc or one-hot encoding, preferably one-hot encoding, which addresses the difficulty of classifiers handling discrete data and expands the features to some extent. Furthermore, since the features obtained from one-hot encoding are discrete and sparse, each category can be represented with fewer dimensions through the embedding layer, and the relationships between different category variables can also be shown. After obtaining a two-dimensional feature matrix through the embedding layer, this two-dimensional feature matrix is ​​input into the convolutional and pooling layers for feature extraction.

[0041] As one embodiment, before inputting the real-time noise value of the next road segment, the vehicle's own noise value, the road segment distance of the next road segment, and the road segment traffic flow of the next road segment into the noise safety index prediction model based on the convolutional neural network, the neural network parameters corresponding to the driver are retrieved from the database through the vehicle network, and the parameters of the noise safety index prediction model based on the convolutional neural network are adjusted according to the neural network parameters.

[0042] As one embodiment, the parameters input to the noise safety index prediction model based on the actual situation reported by the driver are added or reduced.

[0043] In one embodiment, the real-time noise value of the next road segment, the vehicle's own noise value, the road segment distance of the next road segment, and the traffic flow of the next road segment are used as the output parameters of the noise safety index prediction model, and represented by the feature vector X = [A, B, C, D]. The first layer of the fully connected layer is calculated using the ReLU activation function, and the last layer is calculated using the softmax and argmax functions; specifically, the original output values ​​are processed by the softmax function to obtain the probability distribution between different noise safety index values, as shown in formula (1):

[0044]

[0045] Then, using argmax(f(x)), we obtain the variable point x that maximizes f(x):

[0046]

[0047] The value of x is the noise safety index value, and S is the noise safety index value of the road segment obtained in the fully connected layer, which is an integer ranging from 1 to 10; the value of x is the noise safety index value finally calculated by the noise safety index prediction model based on the convolutional neural network.

[0048] Step S3: When the noise safety index value exceeds the driver's preset noise safety index threshold, a warning message is issued to the driver. Specifically, a warning message is issued to the driver to prompt them to change the driving route and recalculate the noise safety index value for the next road segment. After the driver changes the driving route, steps S1 and S2 are repeated.

[0049] In one embodiment, at the end of the trip, the driver scores the noise levels of each road segment and calculates the error value using a loss function. Based on this error value, the parameters of each level of the noise safety index prediction model based on a convolutional neural network are adjusted. Specifically, if the driver finds the noise safety protection experience satisfactory and gives a default positive review, it indicates that the driver accepts the automatically generated noise safety index value, and the analysis module will not initiate the neural network error analysis process. If the driver is dissatisfied with the noise safety protection, error analysis is performed based on the driver's noise scores and noise safety index values ​​for each road segment. Specifically, the data (i.e., noise scores and noise safety index values) is uploaded to a remote server, where the remote server calculates the error value L using a convolutional neural network, specifically based on the following cross-entropy loss function:

[0050]

[0051] Where S is the noise safety index value of the road segment obtained in the fully connected layer, and its value is an integer ranging from 1 to 10; C is a constant representing the total number of different values ​​of S. If S is an integer in the range [1, 10], then C = 10; X is the feature vector; y XS The sign function takes a value of 0 when the noise score of the feature vector X (i.e., the driver's noise score S0) equals S, and a value of 1 otherwise; p KS The output of the second fully connected layer represents the predicted probability that X belongs to S; N is the number of input feature vectors in this round of error analysis. The convolutional neural network-based noise safety index prediction model backpropagates the error value L and automatically adjusts the corresponding parameters of each layer using the gradient descent algorithm. After multiple adjustments to the convolutional neural network-based noise safety index prediction model based on the error value, the prediction result of the convolutional neural network-based noise safety index prediction model will be closer to the driver's subjective rating.

[0052] After the driver completes the evaluation, the relevant parameters of each level of the noise safety index prediction model based on the convolutional neural network are uploaded to the database for use when passing the same road segment again. At the same time, the remote server can also set up a neural network to train the coefficients of the noise safety index value calculation formula (i.e., formula (2)) using the massive amount of data uploaded to the database by different drivers, and generate a general road segment noise safety index value calculation formula that can meet the initial use of most new drivers. Subsequently, a specific road segment noise safety index value calculation formula will be trained and generated based on the driver's evaluation in actual use.

[0053] This invention analyzes real-time urban noise maps and vehicle-generated noise to calculate a travel noise safety index value reflecting the driver's perception of the noise environment along the route. This enables quantitative monitoring of travel noise safety and ensures driver safety. Simultaneously, it promotes the intelligent development of vehicle scene control, specifically including acquiring a real-time noise map of the vehicle's route and a pre-set noise safety index threshold value by the driver; based on the input vehicle route and current location, and according to real-time road segment information, it pre-calculates the noise safety index value of the upcoming road segment; if the noise safety index value of the upcoming road segment exceeds the preset threshold value, it generates a warning to the driver to change the route or take preventative measures.

[0054] Please see Figure 1 Another embodiment of the present invention is a noise warning system for a vehicle during driving, including a user module 101, a noise sensor module 102, a vehicle-mounted module 103, an analysis module 104, and a database module 105.

[0055] The user module is used to preset noise safety index thresholds and perform noise scoring, and input the noise safety index thresholds and noise scores into the analysis module;

[0056] The noise sensor module is used to calculate the vehicle's own noise level.

[0057] The vehicle infotainment module is used to obtain the vehicle's own noise value from the noise sensor module, obtain the real-time noise value of the next road segment from the real-time noise map, obtain the road segment distance and traffic flow of the next road segment from the navigation software, input them into the analysis module, and dynamically adjust the calculation strategy of the noise safety index value.

[0058] The analysis module is used to input the real-time noise value of the next road segment, the vehicle's own noise value, the road segment distance of the next road segment, and the traffic flow of the next road segment into the noise safety index prediction model based on the convolutional neural network to obtain the noise safety index value of the next road segment. It is also used to determine whether to issue a warning message to the driver based on the noise safety index value and the noise safety index threshold, as well as to perform error scoring after the trip and adjust the parameters of the noise safety index prediction model based on the convolutional neural network.

[0059] The database module is used to store the neural network parameters of different drivers.

[0060] It should be noted that the above-described system embodiments of the present invention correspond to the method embodiments of the present invention, and can realize the noise warning method during vehicle operation described in any one of the method embodiments of the present invention. For the sake of brevity, they will not be described in detail here.

[0061] The embodiments described above are merely illustrative of several implementations of the present invention, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of the present invention. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of the present invention, and these modifications and improvements all fall within the scope of protection of the present invention. Therefore, the scope of protection of this patent should be determined by the appended claims.

Claims

1. A noise warning method for vehicles in motion, characterized in that, Includes the following steps: The system obtains a real-time noise map of the current vehicle route from the Internet of Vehicles, obtains the real-time noise value of the next road segment based on the real-time noise map, obtains the road segment distance and traffic flow of the next road segment through navigation software, and obtains the vehicle's own noise value through noise sensors. The noise safety index value of the next road segment is calculated based on the real-time noise value of the next road segment, the vehicle's own noise value, the road segment distance of the next road segment, and the traffic flow of the next road segment. Specifically, the real-time noise value of the next road segment, the vehicle's own noise value, the road segment distance of the next road segment, and the road segment traffic flow of the next road segment are input into a noise safety index prediction model based on a convolutional neural network. The noise safety index prediction model outputs the noise safety index value of the next road segment. The noise safety index prediction model includes an embedding layer, a convolutional layer, a pooling layer, and three fully connected layers. When the noise safety index value exceeds the driver's preset noise safety index threshold, a warning message is issued to the driver.

2. The noise warning method for vehicles during operation as described in claim 1, characterized in that, At the end of the trip, the driver scores the noise level of each road segment and calculates the error value using a loss function. Based on the error value, the driver adjusts the parameters of each level of the noise safety index prediction model based on the convolutional neural network.

3. The noise warning method during vehicle operation as described in claim 2, characterized in that, The specific steps for calculating the error value using the loss function are as follows: The error value is calculated using the following formula: in, The noise safety index value of the road segment obtained in the fully connected layer, ranging from 1 to 10, is an integer; C is a constant, representing... The total number of different values; X is the feature vector; Let X be the sign function, when the noise score of the feature vector X equals The value is 0 if the condition is met, and 1 otherwise. The result of the second fully connected layer indicates that X belongs to The predicted probability; N is the number of feature vectors input in this round of error analysis.

4. The noise warning method during vehicle operation as described in claim 3, characterized in that, A warning message is issued to the driver, specifically: A warning message is sent to the driver to prompt him to change the driving route and recalculate the noise safety index value for the next road segment.

5. The noise warning method for a vehicle during operation as described in any one of claims 1 to 4, characterized in that, The embedding layer employs one-hot encoding to preprocess the input feature vector.

6. A noise warning system for vehicles in motion, characterized in that, It includes a user module, a noise sensor module, an in-vehicle infotainment module, an analysis module, and a database module; The user module is used to preset noise safety index thresholds and perform noise scoring, and input the noise safety index thresholds and noise scores into the analysis module; The noise sensor module is used to calculate the vehicle's own noise level. The vehicle module is used to obtain the vehicle's own noise value from the noise sensor module, obtain the real-time noise value of the next road segment from the real-time noise map, obtain the road segment distance and traffic flow of the next road segment from the navigation software, input them into the analysis module, and dynamically adjust the calculation strategy of the noise safety index value. The analysis module is used to input the real-time noise value of the next road segment, the vehicle's own noise value, the road segment distance of the next road segment, and the traffic flow of the next road segment into a noise safety index prediction model based on a convolutional neural network to obtain the noise safety index value of the next road segment. It is also used to determine whether to issue a warning to the driver based on the noise safety index value and a noise safety index threshold, as well as for error scoring after the trip and adjusting the parameters of the noise safety index prediction model based on the convolutional neural network. The noise safety index prediction model includes an embedding layer, a convolutional layer, a pooling layer, and three fully connected layers. The database module is used to store the neural network parameters of different drivers.

Images