Subway cab active noise reduction method based on feedforward algorithm

Abstract

The present application relates to the field of traffic track noise reduction technology, and particularly relates to a subway cab active noise reduction method based on a feedforward algorithm. The present application comprises the following steps: S1, construction of a noise reduction seat; S2, construction of a transfer function model: MTF function model and STF function model are respectively established under a real noise source and a quiet environment, an EC function model is established by collecting a loudspeaker working signal, and a PF function model is calculated by using recorded signals and the MTF function model and the STF function model; S3, active noise reduction of the feedforward algorithm. The present application realizes accurate cancellation of low-frequency structure noise and medium-high frequency air noise at the ear position of a subway driver by real-time collection of vibration and noise signals by an acceleration / reference / error sensor, combination of MTF / STF / EC function models and a PF prediction algorithm, and noise reduction of up to 10-15 dB. Meanwhile, a silent calibration mode can adapt to the acoustic characteristics of different working conditions and improve environmental comfort.

Description

Technical Field

[0001] This invention relates to the field of noise reduction technology in transportation and rail transit, specifically to an active noise reduction method for subway driver's cab based on a feedforward algorithm. Background Technology

[0002] With the improvement of large-scale integrated circuits and the application of algorithms such as LMS in the field of adaptive control, the cost of DSP chips has continued to decrease and the performance has been significantly improved, driving the in-depth development of digital signal processing technology. Active noise cancellation technology has therefore attracted the attention of scholars worldwide and has been effectively applied in noise-canceling headphones, helmets, and aircraft and automobiles. However, its application in the train field remains at the theoretical research stage. Currently, train noise reduction largely relies on passive technologies, such as optimizing component shapes and using new materials to absorb and reduce sound. However, these methods have long research cycles and require a large workload. Therefore, active noise cancellation technology has become an inevitable trend for noise reduction in high-speed trains. However, in rail transit vehicles, active noise cancellation technology is mostly found in high-end automobiles, with very little application in subway vehicles. Limited by the environment of subway driver's cabs, active noise cancellation devices mostly achieve localized noise reduction through headrests. Although there are existing patents, such as patent publication number CN113257217A which provides an adaptive active noise cancellation method, problems such as concentrated sensors and large noise source localization errors still exist. Furthermore, the main controller needs to be installed separately in the electrical cabinet or under the control panel, occupying limited space and increasing the design workload. The wiring between the main controller and the headrest needs to run along the bottom of the car body, increasing cable length and the risk of data transmission packet loss. It also requires increasing the speaker drive power, affecting noise reduction effectiveness. In addition, the active noise cancellation system and the seats are supplied by different manufacturers, making it difficult to standardize product appearance and manufacturing processes, affecting visual appeal. Therefore, the existing solution still needs optimization in the field of active noise cancellation for subways. Summary of the Invention

[0003] The technical problem to be solved by the present invention is to overcome the shortcomings of the prior art and provide an active noise reduction method for subway driver's cab based on feedforward algorithm.

[0004] The technical solution adopted in this invention is as follows:

[0005] An active noise reduction method for subway driver's cab based on a feedforward algorithm includes the following steps:

[0006] S1. Construction of the noise-reducing seat: First, assemble the metal frame and insert corrugated tubing to protect the wiring harness. Then, install the seat components and apply foam and fabric coverings. Finally, install the lumbar support, fore-and-aft movement mechanism, backrest tilt adjustment mechanism, height adjustment mechanism, and rotation adjustment mechanism. Additionally, the following are also included:

[0007] An accelerometer, mounted on the seat base flange, is used to detect vibration signals transmitted from the subway car body and vibration signals generated by the vehicle's traction device.

[0008] Reference sensors are installed on the left and right sides of the seat cushion support, the left and right sides of the backrest, and the upper right and upper left sides of the headrest. A total of 6 reference microphones are used to install them in the optimal positions based on the actual noise distribution through coherence analysis.

[0009] Error sensors are installed at the front of the head near the left and right ears, with one error sensor installed 10cm above and below the center position.

[0010] S2. Construction of the transfer function model: MTF function model and STF function model are established in real noise source and quiet environment respectively. EC function model is established by collecting loudspeaker working signal. PF function model is calculated using the recorded signal and MTF function model and STF function model.

[0011] S3. Active noise reduction using feedforward algorithm: Acquire signals from accelerometer and reference sensor, process the predicted signal using feedforward algorithm and evaluate the cancellation noise, finally use speaker to emit sound to implement active noise reduction, and adjust and optimize the effect through feedback from error sensor.

[0012] This technical solution achieves active noise reduction at the subway driver's ear position and optimizes the acoustic comfort of the driving environment by constructing a noise-reducing seat integrating multiple types of sensors, establishing a multi-dimensional transfer function model and a feedforward algorithm. Through modular assembly of the seat structure and integrated sensors and functional components, it provides stable and reliable hardware support for the noise reduction system while ensuring seat comfort and adjustment functions. By establishing an MTF function model under real noise conditions, it quantifies the acoustic transfer characteristics between the reference sensor and the error sensor, providing a basis for noise prediction for the feedforward algorithm. Furthermore, by establishing an STF function model under silent conditions, it quantifies the acoustic transfer characteristics between the speaker input and the error sensor. This provides a basis for evaluating the noise reduction effect of the feedforward algorithm; by establishing an EC function model, the signal coupling relationship between the speaker output and the reference sensor is quantified, eliminating the interference of the speaker's own acoustic feedback on noise evaluation; by recording multi-sensor signals and calculating the PF prediction function, the system signal processing delay is reduced, and the real-time tracking capability of the feedforward algorithm for dynamic noise is improved; by fusing multi-model parameters through the feedforward algorithm to generate an anti-phase drive signal, the speaker emits canceling sound waves, and the noise reduction effect is optimized in a closed loop through error sensor feedback; by using the silent calibration mode when the vehicle starts, an environmental acoustic feature database is established, providing a benchmark for adaptive adjustment of noise reduction parameters under different operating conditions.

[0013] In addition, the active noise reduction method for subway driver's cab based on the feedforward algorithm proposed above according to the present invention also has the following additional technical features:

[0014] According to an embodiment of the present invention, the construction of the noise-reducing seat in step S1 includes the following sub-steps:

[0015] S11. Assembly of basic seat structure: First, assemble the metal frame of the seat, and thread the power cord and sensor wiring into the corrugated pipe. Use the corrugated pipe to protect the power cord and sensor wiring from under the seat to the seat back.

[0016] S12. Seat Component Installation: Fix the corrugated tube with the wire harness in the appropriate position, and install the foam model and seat cushion of the seat back to ensure seat comfort; weld the headrest bracket to the seat back frame, and then fix the headrest shell to the bracket with screws. The headrest shell is made of PC material and is injection molded. It is divided into front and rear ends, and two sealed cavities are formed on the left and right sides as echo chambers for speakers. The power conversion module and main controller are installed in the middle of the headrest shell; attach foam to the headrest shell, and then cover the seat back and headrest shell together with the seat cover cloth to form an integrated set. The rear end of the headrest shell and the cover cloth are reserved for maintenance.

[0017] S13. Installation of Seat Function Adjustment Mechanism: The seat back is equipped with a lumbar support mechanism, and a lumbar support adjustment handwheel is installed, allowing the driver to adjust the lumbar support height according to their own needs; a fore-and-aft movement handle is installed, allowing the seat to be adjusted fore-and-aft within a range of 0-200mm; a backrest tilt angle adjustment handle is installed, allowing the backrest to tilt within a range of 45-135°, and automatically locking at the extreme tilting positions to ensure stability; a height adjustment handle is installed, allowing the seat to be raised or lowered within a range of 0-120mm; a rotation adjustment handle is installed, allowing the seat to be rotated left and right within a range of ±90°, and locking at the rotation position.

[0018] This technical solution ensures sensor signal stability through modular assembly of metal brackets and integrated wiring protection design. The integrated installation of sponge seat cushion and PC headrest shell optimizes the acoustic structure. At the same time, it is equipped with a multi-dimensional adjustment mechanism, ultimately creating an intelligent noise-reducing seat platform that combines comfort, acoustic performance and ergonomic adaptability.

[0019] According to an embodiment of the present invention, the MTF function model in step S2 includes the following construction steps:

[0020] The operation is performed under conditions with a real noise source. The transfer function MTF between the reference sensor and the error sensor is calculated based on the acquired signal. The transfer function MTF is required to execute the predictive filter function PF method. Once there is an acoustic medium filter between the reference sensor and the error sensor, the reference sensor signal is transmitted to the filter, and the original noise is evaluated at the error sensor.

[0021] According to an embodiment of the present invention, the STF function model in step S2 includes the following construction steps:

[0022] Operating in a quiet environment, the speaker emits white noise. The transfer function STF between the speaker input and the error sensor is calculated based on the signals collected by the speaker input and the error sensor. The transfer function STF is required to execute the predictive filter function PF method. Once there is an acoustic medium filter between the speaker and the error sensor, the speaker input signal is passed to the filter to obtain an evaluation of noise reduction at the error sensor.

[0023] According to an embodiment of the present invention, the EC function establishment in step S2 includes the following construction steps:

[0024] During operation, the sound emitted by the speaker is collected by a reference sensor. The speaker input signal and the signal collected by the reference sensor are recorded. Based on the recorded signal, a transfer function EC between the speaker input and the reference sensor is established. The speaker input signal is passed to the transfer function EC to obtain an evaluation of noise reduction at the reference sensor. The evaluation signal is then filtered out by the reference sensor to generate an evaluation of the original noise.

[0025] According to one embodiment of the present invention, the signal recording in step S2 includes the following construction steps:

[0026] During the MTF function model establishment process, the reference sensor and error sensor signals are recorded. The PF function is calculated using the MTF function, STF function, and the recorded signals. The PF function is used to predict future sample signals from the signal obtained from the reference sensor, thereby reducing latency. The PF function should be calculated after the MTF and STF are set. In practice, the PF function replaces the actual error sensor.

[0027] According to an embodiment of the present invention, the active noise reduction of the feedforward algorithm in step S3 includes the following sub-steps:

[0028] S31. The main controller predicts the future sample signal of the reference sensor signal based on the PF function, and uses the MTF, STF and EC functions to evaluate the original noise and the noise signal that needs to be canceled by combining the predicted signal and the acquired signal.

[0029] S32. Generate the drive signal for the loudspeaker based on the calculated noise signal that needs to be canceled.

[0030] S33: The main controller sends a drive signal to the speaker, and the speaker emits a sound with the opposite phase and similar amplitude to the noise, thus achieving active noise reduction.

[0031] S34. The noise reduction effect is monitored in real time by an error sensor, and the signal is fed back to the main controller. The main controller adjusts the parameters of the feedforward algorithm and the drive signal of the speaker according to the feedback signal to continuously optimize the noise reduction effect.

[0032] According to one embodiment of the present invention, the active noise reduction of the feedforward algorithm in step S3 automatically enters the silent calibration mode after the subway vehicle starts, emits multi-band test sounds through the speaker, collects the attenuation characteristics of each frequency band using the error sensor, and establishes an acoustic feature database of the current environment; automatically performs sensor sensitivity detection every 24 hours, judges sensor performance drift by comparing historical data, and triggers an early warning mechanism and automatically compensates parameters when the error exceeds ±3dB.

[0033] Compared with the prior art, the present invention has the following advantages:

[0034] By collecting vibration and noise signals in real time using acceleration / reference / error sensors, and combining MTF / STF / EC function models and PF prediction algorithms, the system can accurately cancel low-frequency structural noise and mid-to-high-frequency air noise at the subway driver's ear position, achieving a noise reduction of 10-15dB. At the same time, the silent calibration mode can adapt to the acoustic characteristics of different working conditions, improving environmental comfort. Attached Figure Description

[0035] Figure 1 This is a flowchart illustrating the principle of the method of the present invention.

[0036] Figure 2 This is a side view of the noise-reducing seat of the present invention.

[0037] Figure 3 This is the front view of the noise-reducing seat of the present invention.

[0038] Figure 4 This is a rear view of the noise-reducing seat of the present invention.

[0039] Figure 5 This is a model diagram of the active noise reduction algorithm of the feedforward algorithm.

[0040] In the diagram: 1. Accelerometer; 2. Reference sensor; 3. Error sensor. Detailed Implementation

[0041] 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, not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0042] Example 1

[0043] like Figure 1 As shown, this embodiment provides an active noise reduction method for a subway driver's cab based on a feedforward algorithm, including the following steps:

[0044] S1. Construction of noise-reducing seats: such as Figures 2 to 4 As shown, first assemble the metal frame and insert corrugated tubing to protect the wiring harness. Next, install the seat components and apply foam and fabric coverings. Finally, install the lumbar support, fore-and-aft movement mechanism, backrest tilt adjustment mechanism, height adjustment mechanism, and rotation adjustment mechanism. Additionally, the following are also included:

[0045] Accelerometer 1, mounted on the seat base flange, is used to detect vibration signals transmitted from the subway car body and vibration signals generated by the vehicle's traction device.

[0046] Reference sensor 2 is installed on the left and right sides of the seat cushion bracket, the left and right sides of the backrest, and the upper right and upper left sides of the headrest. A total of 6 reference microphones are used to install them in the optimal position based on the actual noise distribution through coherence analysis.

[0047] Error sensor 3 is installed at the left and right ears near the front of the head, with one error sensor 3 installed 10cm above and below the center position.

[0048] S2. Construction of the transfer function model: such as Figure 5 As shown, MTF and STF function models are established in real noise sources and quiet environments, respectively. The EC function model is established by collecting the loudspeaker's operating signal. The PF function model is calculated using the recorded signal and the MTF and STF function models.

[0049] S3, Active noise reduction using feedforward algorithms: such as Figure 5 As shown, the signals from accelerometer 1 and reference sensor 2 are collected, the feedforward algorithm processes the predicted signal and evaluates the cancellation noise, and finally the speaker emits sound to implement active noise reduction, and the effect is adjusted and optimized through feedback from error sensor 3.

[0050] like Figures 1 to 5As shown, this technical solution achieves active noise reduction at the subway driver's ear position and optimizes the acoustic comfort of the driving environment by constructing a noise-reducing seat integrating multiple types of sensors, establishing a multi-dimensional transfer function model and a feedforward algorithm. Through modular assembly of the seat structure and integrated sensors and functional components, stable and reliable hardware support is provided for the noise reduction system while ensuring seat comfort and adjustment functions. By establishing an MTF function model under real noise conditions, the acoustic transfer characteristics between reference sensor 2 and error sensor 3 are quantified, providing a noise prediction basis for the feedforward algorithm. Furthermore, by establishing an STF function model under silent conditions, the acoustic transfer between the speaker input and error sensor 3 is quantified. The system features include: providing a basis for evaluating the noise reduction effect of the feedforward algorithm; quantifying the signal coupling relationship between the speaker output and reference sensor 2 by establishing an EC function model to eliminate the interference of the speaker's own acoustic feedback on noise evaluation; reducing system signal processing latency and improving the real-time tracking capability of the feedforward algorithm for dynamic noise by recording multi-sensor signals and calculating the PF prediction function; generating an anti-phase drive signal by fusing multi-model parameters through the feedforward algorithm to drive the speaker to emit canceling sound waves, and achieving closed-loop optimization of the noise reduction effect through feedback from error sensor 3; and establishing an environmental acoustic feature database through the silent calibration mode when the vehicle starts, providing a benchmark for adaptive adjustment of noise reduction parameters under different operating conditions.

[0051] In addition, the active noise reduction method for subway driver's cab based on the feedforward algorithm proposed above according to the present invention also has the following additional technical features:

[0052] According to an embodiment of the present invention, the construction of the noise-reducing seat in step S1 includes the following sub-steps:

[0053] S11. Assembly of basic seat structure: First, assemble the metal frame of the seat, and thread the power cord and sensor wiring into the corrugated pipe. Use the corrugated pipe to protect the power cord and sensor wiring from under the seat to the seat back.

[0054] S12. Seat Component Installation: Fix the corrugated tube with the wire harness in the appropriate position, and install the foam model and seat cushion of the seat back to ensure seat comfort; weld the headrest bracket to the seat back frame, and then fix the headrest shell to the bracket with screws. The headrest shell is made of PC material and is injection molded. It is divided into front and rear ends, and two sealed cavities are formed on the left and right sides as echo chambers for speakers. The power conversion module and main controller are installed in the middle of the headrest shell; attach foam to the headrest shell, and then cover the seat back and headrest shell together with the seat cover cloth to form an integrated set. The rear end of the headrest shell and the cover cloth are reserved for maintenance.

[0055] S13. Installation of Seat Function Adjustment Mechanism: The seat back is equipped with a lumbar support mechanism, and a lumbar support adjustment handwheel is installed, allowing the driver to adjust the lumbar support height according to their own needs; a fore-and-aft movement handle is installed, allowing the seat to be adjusted fore-and-aft within a range of 0-200mm; a backrest tilt angle adjustment handle is installed, allowing the backrest to tilt within a range of 45-135°, and automatically locking at the extreme tilting positions to ensure stability; a height adjustment handle is installed, allowing the seat to be raised or lowered within a range of 0-120mm; a rotation adjustment handle is installed, allowing the seat to be rotated left and right within a range of ±90°, and locking at the rotation position.

[0056] This technical solution ensures sensor signal stability through modular assembly of metal brackets and integrated wiring protection design. The integrated installation of sponge seat cushion and PC headrest shell optimizes the acoustic structure. At the same time, it is equipped with a multi-dimensional adjustment mechanism, ultimately creating an intelligent noise-reducing seat platform that combines comfort, acoustic performance and ergonomic adaptability.

[0057] According to an embodiment of the present invention, the MTF function model in step S2 includes the following construction steps:

[0058] The operation is performed under conditions with a real noise source. The transfer function MTF between the reference sensor 2 and the error sensor 3 is calculated based on the acquired signal. The transfer function MTF is required to execute the predictive filter function PF method. Once there is an acoustic medium filter between the reference sensor 2 and the error sensor 3, the signal from the reference sensor 2 is transmitted to the filter, and the original noise is evaluated at the error sensor 3.

[0059] According to an embodiment of the present invention, the STF function model in step S2 includes the following construction steps:

[0060] Operating in a quiet environment, the speaker emits white noise. The transfer function STF between the speaker input and the error sensor 3 is calculated based on the signals collected by the speaker input and the error sensor 3. The transfer function STF is required to execute the predictive filter function PF method. Once there is an acoustic medium filter between the speaker and the error sensor 3, the speaker input signal is passed to the filter to obtain an evaluation of noise reduction at the error sensor 3.

[0061] According to an embodiment of the present invention, the EC function establishment in step S2 includes the following construction steps:

[0062] During operation, the sound emitted by the speaker is collected by the reference sensor 2. The speaker input signal and the signal collected by the reference sensor 2 are recorded. The transfer function EC between the speaker input and the reference sensor 2 is established based on the recorded signal. The speaker input signal is passed to the transfer function EC to obtain an evaluation of noise reduction at the reference sensor 2. The evaluation signal is then filtered out by the reference sensor 2 to generate an evaluation of the original noise.

[0063] According to one embodiment of the present invention, the signal recording in step S2 includes the following construction steps:

[0064] During the establishment of the MTF function model, the signals of reference sensor 2 and error sensor 3 are recorded. The MTF function, STF function and the recorded signal are used to calculate the PF function. The PF function is used to predict the future sample signal of the signal obtained from reference sensor 2, thereby reducing the delay. The PF function should be calculated after the MTF and STF are set. In operation, the PF function replaces the actual error sensor 3.

[0065] According to an embodiment of the present invention, the active noise reduction of the feedforward algorithm in step S3 includes the following sub-steps:

[0066] S31. The main controller predicts the future sample signal of the reference sensor 2 signal based on the PF function, and uses the MTF, STF and EC functions to evaluate the original noise and the noise signal that needs to be canceled by combining the predicted signal and the acquired signal.

[0067] S32. Generate the drive signal for the loudspeaker based on the calculated noise signal that needs to be canceled.

[0068] S33: The main controller sends a drive signal to the speaker, and the speaker emits a sound with the opposite phase and similar amplitude to the noise, thus achieving active noise reduction.

[0069] S34. The noise reduction effect is monitored in real time by error sensor 3, and the signal is fed back to the main controller. The main controller adjusts the parameters of the feedforward algorithm and the drive signal of the speaker according to the feedback signal to continuously optimize the noise reduction effect.

[0070] According to one embodiment of the present invention, the active noise reduction of the feedforward algorithm in step S3 automatically enters the silent calibration mode after the subway vehicle starts, emits multi-band test sounds through the speaker, collects the attenuation characteristics of each frequency band using error sensor 3, and establishes an acoustic feature database of the current environment; automatically performs sensor sensitivity detection every 24 hours, judges sensor performance drift by comparing historical data, and triggers an early warning mechanism and automatically compensates parameters when the error exceeds ±3dB.

[0071] Although the present invention has been described in detail with reference to the accompanying drawings and preferred embodiments, the invention is not limited thereto. Various equivalent modifications or substitutions can be made to the embodiments of the invention by those skilled in the art without departing from the spirit and essence of the invention, and such modifications or substitutions should all be within the scope of the invention. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the invention should also be covered within the protection scope of the invention. Therefore, the protection scope of the invention should be determined by the scope of the claims.

Claims

1. A method for active noise reduction in a subway driver's cab based on a feedforward algorithm, characterized in that, Includes the following steps: S1. Construction of the noise-reducing seat: First, assemble the metal frame and insert corrugated tubing to protect the wiring harness. Then, install the seat components and apply foam and fabric coverings. Finally, install the lumbar support, fore-and-aft movement mechanism, backrest tilt adjustment mechanism, height adjustment mechanism, and rotation adjustment mechanism. Additionally, the following are also included: An acceleration sensor (1) is installed on the seat base flange to detect vibration signals transmitted from the subway car body and vibration signals generated by the vehicle traction device; Reference sensors (2) are installed on the left and right sides of the seat cushion bracket, the left and right sides of the backrest, the upper right and upper left sides of the headrest, and a total of 6 reference microphones. They are used to install the microphones in the optimal positions based on the actual noise distribution through coherence analysis. Error sensor (3) is installed at the left and right ears near the front of the head, with one error sensor (3) installed 10cm above and below the center position. S2. Construction of the transfer function model: MTF function model and STF function model are established in real noise source and quiet environment respectively. EC function model is established by collecting loudspeaker working signal. PF function model is calculated using the recorded signal and MTF function model and STF function model. The MTF function model in step S2 includes the following construction steps: The operation is performed under the condition of real noise source. The transfer function MTF between the reference sensor (2) and the error sensor (3) is calculated based on the acquired signal. The transfer function MTF is used to perform the prediction filter function PF method. Once there is an acoustic medium filter between the reference sensor (2) and the error sensor (3), the signal of the reference sensor (2) is transmitted to the filter, and the original noise is evaluated at the error sensor (3). The STF function model in step S2 includes the following construction steps: In a quiet environment, the speaker emits white noise. The transfer function STF between the speaker input and the error sensor (3) is calculated based on the signals collected by the speaker input and the error sensor (3). The transfer function STF is used to perform the prediction filter function PF method. Once there is an acoustic medium filter between the speaker and the error sensor (3), the speaker input signal is passed to the filter to obtain an evaluation of noise reduction at the error sensor (3). During the establishment of the MTF function model, the signals of the reference sensor (2) and the error sensor (3) are recorded. The MTF function, STF function and the recorded signal are used to calculate the PF function. The PF function is used to predict the future sample signal of the signal obtained from the reference sensor (2), thereby reducing the delay. The PF function should be calculated after the MTF and STF are set. In operation, the PF function replaces the actual error sensor (3). S3. Active noise reduction using feedforward algorithm: signals from acceleration sensor (1) and reference sensor (2) are collected, the feedforward algorithm processes the predicted signal and evaluates the cancellation noise, and finally the speaker emits sound to implement active noise reduction, and the effect is adjusted and optimized by feedback through error sensor (3).

2. The active noise reduction method for subway driver's cab based on feedforward algorithm as described in claim 1, characterized in that, The construction of the noise-reducing seat in step S1 includes the following sub-steps: S11. Assembly of basic seat structure: First, assemble the metal frame of the seat, and thread the power cord and sensor wiring into the corrugated pipe. Use the corrugated pipe to protect the power cord and sensor wiring from under the seat to the seat back. S12. Seat Component Installation: Fix the corrugated tube with the wire harness in the appropriate position, and install the foam model and seat cushion of the seat back to ensure seat comfort; weld the headrest bracket to the seat back frame, and then fix the headrest shell to the bracket with screws. The headrest shell is made of PC material and is injection molded. It is divided into front and rear ends, and two sealed cavities are formed on the left and right sides as echo chambers for speakers. The power conversion module and main controller are installed in the middle of the headrest shell; attach foam to the headrest shell, and then cover the seat back and headrest shell together with the seat cover cloth to form an integrated set. The rear end of the headrest shell and the cover cloth are reserved for maintenance. S13. Installation of seat function adjustment mechanism: The seat back is equipped with a lumbar support mechanism and a rotating lumbar support adjustment handwheel is installed so that the driver can adjust the lumbar support height according to his own situation. Install a fore-and-aft adjustment handle to allow the seat to be adjusted from 0-200mm forward and backward; install a backrest tilt adjustment handle to adjust the backrest tilt angle, with an adjustment range of 45-135°, and it can automatically lock at the extreme positions of forward and backward tilt to ensure stability; install a height adjustment handle to raise and lower the seat height, with an adjustment range of 0-120mm vertically; install a rotation adjustment handle to rotate the seat left and right, with a rotation range of ±90°, and it locks at the rotation position.

3. The active noise reduction method for subway driver's cab based on feedforward algorithm as described in claim 1, characterized in that, The EC function establishment in step S2 includes the following construction steps: When working, the sound emitted by the speaker is collected by the reference sensor (2). The speaker input signal and the signal collected by the reference sensor (2) are recorded. The transfer function EC between the speaker input and the reference sensor (2) is established based on the recorded signal. The speaker input signal is passed to the transfer function EC to obtain an evaluation of noise reduction at the reference sensor (2).

4. The active noise reduction method for subway driver's cab based on feedforward algorithm as described in claim 1, characterized in that, The active noise reduction of the feedforward algorithm in step S3 includes the following sub-steps: S31. The main controller predicts the future sample signal of the reference sensor (2) signal based on the PF function, and uses the MTF, STF and EC functions to evaluate the original noise and the noise signal that needs to be canceled by combining the predicted signal and the acquired signal. S32. Generate the drive signal for the loudspeaker based on the calculated noise signal that needs to be canceled. S33: The main controller sends a drive signal to the speaker, and the speaker emits a sound with the opposite phase and similar amplitude to the noise, thus achieving active noise reduction. S34. The noise reduction effect is monitored in real time by the error sensor (3), and the signal is fed back to the main controller. The main controller adjusts the parameters of the feedforward algorithm and the drive signal of the speaker according to the feedback signal, and continuously optimizes the noise reduction effect.

5. The active noise reduction method for subway driver's cab based on feedforward algorithm as described in claim 1 or 4, characterized in that, The active noise reduction of the feedforward algorithm in step S3 automatically enters the silent calibration mode after the subway vehicle starts. It emits multi-band test sounds through the loudspeaker and uses the error sensor (3) to collect the attenuation characteristics of each band to establish an acoustic feature database of the current environment.

6. The active noise reduction method for subway driver's cab based on feedforward algorithm as described in claim 5, characterized in that, The feedforward algorithm in step S3 performs active noise reduction and automatically performs sensor sensitivity detection every 24 hours. It judges the sensor performance drift by comparing historical data and triggers an early warning mechanism and automatically compensates parameters when the error exceeds ±3dB.

Images