Railway vehicle noise distinguishing and extracting device and method

Abstract

The invention relates to the technical field of railway vehicle noise analysis and processing, and particularly discloses a railway vehicle noise distinguishing and extracting device and method.The device comprises an acquisition module, a processing module and an output module, and the acquisition module is used for acquiring carriage mixed noise in the running process of a railway vehicle; the processing module is used for performing frequency decomposition, noise classification and independent loudness analysis of various types of classified noise on the collected mixed noise of the carriage; the output module is used for outputting and / or storing the loudness values and the frequency characteristics of various types of noise after classification; according to the method, wheel track noise, passenger voice and train station reporting voice can be accurately separated and subjected to loudness analysis in real time, the problem that the passenger voice and the train station reporting voice are difficult to separate and distinguish is solved, and a scientific noise monitoring and management means is provided for rail transit operation enterprises; the passenger comfort level is improved, the station reporting volume and the operation service quality are optimized, and complaint and transformation cost caused by noise is reduced.

Description

Technical Field

[0001] This invention relates to the field of rail vehicle noise analysis and processing technology, and specifically discloses a rail vehicle noise differentiation and extraction device and method. Background Technology

[0002] With the rapid development of urban rail transit, the ride comfort of subway, light rail, and other rail vehicles has increasingly become an important indicator for operation management and passenger evaluation. Among these factors, the noise level inside the carriage is one of the most critical factors affecting ride comfort. The noise environment inside rail vehicles during operation is complex, with its main sound sources including: wheel-rail noise generated by vehicle operation, passenger voices from conversations inside the carriage, and station announcements broadcast by the vehicle's public address system (PIS). These sound sources superimpose in both the temporal and spatial domains, forming a complex mixed sound field.

[0003] Currently, the industry mainly relies on traditional sound pressure level measurement methods for monitoring and assessing noise inside rail vehicles. For example, invention patent application number 200910169986.4 discloses a method and system for measuring and analyzing noise inside railway trains. This patent uses multiple sound pressure sensors arranged inside the carriage, combined with multi-source information such as acceleration sensors, speed sensors, and GPS outside the vehicle, to perform statistical analysis, spectral analysis, and FFT calculations at the data acquisition and processing terminal, thereby realizing the analysis and evaluation of the overall noise level inside the vehicle. However, it has obvious limitations.

[0004] First, it lacks the ability to distinguish sound sources: Most of the patented technologies treat carriage noise as a whole for measurement and analysis, failing to effectively differentiate between different types of sound sources such as wheel-rail noise, passenger voices, and station announcements. Different types of noise have drastically different impacts on passenger comfort. For example, continuous wheel-rail noise is objective physical noise, while clear station announcements provide necessary service information, but excessive volume can also cause interference. Without distinguishing sound sources, it's difficult to accurately trace and target noise sources. Second, it's difficult to distinguish specific sound sources: In mixed sounds, passenger voices and station announcements (PIS) are both human voices, with highly similar spectral characteristics within conventional frequency bands. Effective differentiation using traditional spectral analysis or filtering methods is extremely difficult. This makes it difficult for operators to accurately assess the clarity and interference of the station announcement system, and also makes it impossible to analyze the noise level generated by passenger conversations separately. Third, this technical solution relies on multi-sensor (such as accelerometers and GPS) data fusion and large-scale centralized data processing platforms (such as LabVIEW), resulting in a complex system structure, high deployment costs, and difficulty in achieving low-cost, large-scale field applications and long-term real-time monitoring.

[0005] Therefore, given the shortcomings of existing methods and systems for measuring and analyzing noise inside railway trains in terms of noise analysis and processing of rail vehicles, this application proposes a device and method that can operate in the on-board environment at low cost and high efficiency, and can accurately distinguish, extract and independently analyze key sound sources in the mixed noise inside rail vehicle carriages. Summary of the Invention

[0006] The purpose of this invention is to provide a device and method for distinguishing and extracting noise from rail vehicles, and to achieve accurate distinction, extraction and independent analysis of key sound sources in mixed noise inside rail vehicle carriages.

[0007] This invention is achieved through the following technical solution: A noise differentiation and extraction device for rail vehicles includes a data acquisition module, a processing module, and an output module, wherein: The acquisition module is used to collect the mixed noise from the carriages during the operation of rail vehicles; The processing module is used to perform frequency decomposition, noise classification, and independent loudness analysis of each type of noise after classification on the collected mixed noise in the carriage. The output module is used to output and / or store the loudness values ​​and frequency characteristics of various types of noise after classification.

[0008] As a specific feature of the above scheme, the acquisition module is an INMP441 digital microphone with an I2S interface.

[0009] As a specific feature of the above scheme, the processing module is an ESP32 microcontroller, which has built-in fast Fourier transform algorithm, noise classification and loudness analysis functions.

[0010] As a specific setting of the above scheme, the output module can output and / or store information using any one or more of the following methods: serial port output, wireless transmission, or local storage.

[0011] The present invention also discloses a method for distinguishing and extracting noise from rail vehicles based on the above-mentioned rail vehicle noise distinguishing and extraction device, comprising the following steps: (1) Embed identification audio that is beyond the range of human hearing but can be collected by the acquisition module into the audio of the train station announcement system; (2) Collect mixed noise from the carriages during the operation of the rail vehicle using the acquisition module; (3) The sound energy distribution of different frequency bands is obtained by performing fast Fourier transform operation through the processing module; (4) Set a preset frequency threshold for wheel-rail noise, and extract wheel-rail noise features based on the preset frequency threshold to distinguish it from human voice; (5) Detect whether the human voice frequency band contains the identification audio embedded in the train station announcement system audio; (6) The part of the detected audio is identified as the train station announcement voice, and the remaining part is identified as the passenger voice, thus completing the classification of passenger voice and train station announcement voice; (7) Calculate and analyze the loudness values ​​of wheel-rail noise, passenger voice and train station announcement, and output the analysis results through the output module.

[0012] As a specific setting of the above scheme, the frequency range of the identified audio in step (1) is 18kHz~22kHz.

[0013] As a specific setting of the above scheme, the loudness value analysis in step (7) is to calculate and analyze one or more acoustic parameters of the sound signal, including sound pressure level, RMS value, peak value, and spectral energy distribution.

[0014] As a further setting of the above scheme, step (7) also includes data processing, which optimizes the sound signal quality and improves the accuracy of noise separation and classification by using a filtering algorithm. The filtering algorithm is one or more of the Kalman filtering algorithm or the Butterworth filtering algorithm.

[0015] Compared with the prior art, the beneficial effects of the present invention are as follows: The noise differentiation and extraction device and method for rail vehicles disclosed in this invention can separate and analyze the loudness of wheel-rail noise, passenger voices, and train station announcements in real time and accurately. This provides rail transit operators with scientific noise monitoring and management tools, which helps to improve passenger comfort, optimize station announcement volume and operational service quality, and reduce noise-related complaints and renovation costs.

[0016] This invention distinguishes between the mixed noise collected inside the train carriage by using a Fast Fourier Transform algorithm to decompose the noise signal into different frequency bands. This enables the extraction of frequency domain features from wheel-rail noise and human voices (including passenger voices and train announcement voices). Simultaneously, a special identifier audio is embedded into the existing train announcement system audio. This identifier audio is beyond the range of human hearing but can still be collected by the acquisition module. Therefore, without affecting the normal voice broadcast of the train announcement system, it can effectively distinguish between the train announcement voices and passenger voices, solving the problem of distinguishing between passenger voices and train announcement voices in existing technologies.

[0017] The hardware architecture of the noise differentiation and extraction device for rail vehicles disclosed in this invention adopts a combination of ESP32 microcontroller and INMP441 microphone, which has the advantages of low hardware cost, small device size, field deployment and easy maintenance. It can be widely used in various rail transit vehicles such as subway, light rail and high-speed rail, and has high promotion value and market application prospects.

[0018] This invention, through long-term accumulation and analysis of noise characteristics, can provide data support for vehicle design improvement, vehicle maintenance, and noise control, thereby achieving multiple improvements in social, economic, and environmental benefits. Attached Figure Description

[0019] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0020] Figure 1 This is a schematic diagram of the hardware architecture in this invention; Figure 2 This is a flowchart illustrating the noise differentiation and extraction method for rail vehicles in this invention. Detailed Implementation

[0021] To enable those skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present application, and not all embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative effort should fall within the scope of protection of the present application.

[0022] It should be noted that, unless otherwise specified, the embodiments and features described in this application can be combined with each other. The following will refer to the appendix... Figures 1-2 This application will be described in detail with reference to the embodiments.

[0023] Example 1

[0024] Embodiment 1 of the present invention provides a method for distinguishing and extracting noise from rail vehicles, which is implemented based on a hardware architecture.

[0025] Reference Appendix Figure 1 The hardware architecture includes a data acquisition module 100, a processing module 200, and an output module 300. Specifically, the data acquisition module 100 can be equipped with I / O... 2 The INMP441 digital microphone with an S-interface can be installed in the carriage of a rail vehicle to collect mixed noise from the carriage during operation.

[0026] The processing module 200 can specifically adopt an ESP32 microcontroller, which has a built-in Fast Fourier Transform (FFT) algorithm and classification function. It can perform frequency decomposition on the sound signal of the mixed noise of the carriage collected by the acquisition module 100, and then decompose the mixed noise of the carriage into at least three categories, including train wheel and rail noise, passenger voices and train station announcements.

[0027] The output module 300 can use one or more of the following: serial port output, wireless transmission (Wi-Fi, Bluetooth) or local storage (SD card). The design of the output module 300 can realize the saving and display of data.

[0028] The basic process and logic of the noise differentiation and extraction method for rail vehicles provided in Embodiment 1 are as follows: Figure 2 As shown, the key steps include the following: S1: Install the acquisition module 100 in the carriage of the rail vehicle, and then collect the mixed noise inside the carriage during the operation of the rail vehicle.

[0029] S2: Embed identification audio that is beyond the range of human hearing but can be collected by the acquisition module into the audio of the announcement of the train station announcement system (PIS system). In the specific design, the identification audio is selected from ultrasonic or high-frequency components (18kHz~22kHz) to effectively distinguish the announcement voice from the passenger voice without affecting the normal voice broadcast of the train station announcement system.

[0030] S3: The processing module 200 uses the Fast Fourier Transform (FFT) algorithm to decompose the acoustic signal of the mixed noise in the carriage collected by the acquisition module 100 into different frequency bands, thereby realizing the extraction of frequency domain features of wheel-rail noise, passenger voices and train station announcements.

[0031] S4: In the processing module 200, a frequency threshold (500~2000Hz) for wheel-rail noise is preset, and wheel-rail noise features are extracted according to the preset frequency threshold to distinguish them from human voices (including passenger voices and station announcement voices).

[0032] S5: The processing module 200 further detects the separated human voice to determine whether it contains the identifier audio.

[0033] S6: The detected audio portion is identified as the train announcement voice, and the remaining portion is identified as the passenger voice, thus completing the classification of passenger voice and train announcement voice.

[0034] S7: Perform loudness analysis on the separated train wheel-rail noise, passenger voices, and station announcements (specific loudness analysis includes sound pressure level, RMS value, peak value, spectral energy distribution, etc.), and finally use one or more methods such as serial port transmission, wireless transmission (Wi-Fi, Bluetooth) or local storage (SD card) to display and store the data, and output the analysis results.

[0035] In addition, when the processing module 200 processes the mixed noise in the carriage, it can also use a variety of filtering algorithms (such as Kalman filtering and Butterworth filtering) to optimize signal quality and improve the accuracy of noise separation and classification.

[0036] This embodiment 1 discloses a device and method for distinguishing and extracting noise from urban rail vehicles based on embedded acoustic acquisition and frequency domain analysis. The device uses an ESP32 microcontroller as the core processing unit, employs an INMP441 digital microphone as the sound acquisition device, and utilizes I... 2 The S-interface enables real-time audio data transmission. During the operation of rail vehicles, the sources of sound inside the carriages are complex, mainly including wheel-rail noise generated by the train itself, passenger voices, and announcements from the train station announcement system (PIS system). These different sound sources superimpose on each other within the carriage space, forming a mixed noise environment.

[0037] This invention utilizes a Fast Fourier Transform (FFT) algorithm running on an ESP32 microcontroller to perform frequency domain decomposition on the acquired mixed noise acoustic signal, extracting the sound energy distribution characteristics of different frequency bands. Analysis results show that wheel-rail noise differs significantly from human voice in its spectrum and can be distinguished and loudness analyzed. However, both passenger voices and station announcement system voices are human voices with similar spectral characteristics, making direct differentiation difficult.

[0038] To address the challenge of distinguishing human voices, this embodiment 1 also embeds or pre-sets specific ultrasonic or high-frequency components in the broadcast audio of the train announcement system (PIS system). These ultrasonic or high-frequency signals are beyond the range of human hearing, but can be effectively collected by the INMP441 microphone. Then, the embedded ultrasonic or high-frequency information is extracted using the Fast Fourier Transform (FFT) algorithm, thereby effectively distinguishing between passenger voices and announcement voices. This allows for independent loudness analysis and statistics of wheel-rail noise, passenger voices, and train announcement voices, providing reliable technical support for rail vehicle noise monitoring, passenger comfort evaluation, and operation management.

[0039] The above are merely preferred embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A noise differentiation and extraction device for rail vehicles, characterized in that, It includes a data acquisition module, a processing module, and an output module, wherein: The acquisition module is used to collect the mixed noise from the carriages during the operation of rail vehicles; The processing module is used to perform frequency decomposition, noise classification, and independent loudness analysis of each type of noise after classification on the collected mixed noise in the carriage. The output module is used to output and / or store the loudness values ​​and frequency characteristics of various types of noise after classification.

2. The noise differentiation and extraction device for rail vehicles according to claim 1, characterized in that, The acquisition module uses I 2 INMP441 digital microphone with S-interface.

3. The noise differentiation and extraction device for rail vehicles according to claim 1, characterized in that, The processing module is an ESP32 microcontroller, which has built-in fast Fourier transform algorithm, noise classification and loudness analysis functions.

4. The noise differentiation and extraction device for rail vehicles according to claim 1, characterized in that, The output module can output and / or store information using one or more of the following methods: serial port output, wireless transmission, or local storage.

5. A method for distinguishing and extracting noise from rail vehicles based on the rail vehicle noise distinguishing and extraction device according to any one of claims 1 to 4, characterized in that, Includes the following steps: Embed identification audio that is beyond the range of human hearing but can be collected by the acquisition module into the audio of the train station announcement system; The mixed noise from the carriages during the operation of the rail vehicle is collected using a data acquisition module. The sound energy distribution in different frequency bands is obtained by performing a fast Fourier transform operation through the processing module. The system presets a frequency threshold for wheel-rail noise and extracts wheel-rail noise features based on this threshold to distinguish them from human voices. Detect whether the audio signals embedded in the train station announcement system are present in the human voice frequency band; The detected audio portion is identified as the train announcer's voice, while the remaining portion is identified as the passenger's voice, thus completing the classification of passenger voices and train announcer voices. The loudness values ​​of wheel-rail noise, passenger voice, and train station announcement are calculated and analyzed separately, and the analysis results are output through the output module.

6. The method for distinguishing and extracting noise from rail vehicles according to claim 5, characterized in that, The frequency range of the identified audio in step (1) is 18kHz to 22kHz.

7. The method for distinguishing and extracting noise from rail vehicles according to claim 5, characterized in that, The loudness analysis in step (7) involves calculating and analyzing one or more acoustic parameters of the sound signal, including sound pressure level, RMS value, peak value, and spectral energy distribution.

8. The method for distinguishing and extracting noise from rail vehicles according to claim 4, characterized in that, Step (7) also includes data processing, which optimizes the sound signal quality and improves the accuracy of noise separation and classification by using a filtering algorithm. The filtering algorithm is one or more of the Kalman filtering algorithm or the Butterworth filtering algorithm.

Images