Noise control method, apparatus and system, and transportation means

Abstract

A noise control method, apparatus and system, and a transportation means, relating to the technical field of noise control, for use in providing stable noise cancellation experience for users. The method comprises: controlling a loudspeaker to play a noise cancellation signal (401); acquiring a first reference signal, a first error signal, and first state information of a current secondary path (402); matching the first state information of the current secondary path with state information of a plurality of pre-stored secondary paths; if there is a matched pre-stored secondary path, determining a filter coefficient on the basis of a path parameter of the matched pre-stored secondary path, the first reference signal, and the first error signal, and using a first step size to adjust the filter coefficient (404); if there is no matched pre-stored secondary path, determining a filter coefficient on the basis of a path parameter of a default secondary path, the first reference signal, and the first error signal, and using a second step size to adjust the filter coefficient, wherein the second step size is less than the first step size (405); and using the adjusted filter coefficient to update the noise cancellation signal (406). The method can take into account both noise cancellation stability and noise cancellation efficiency.

Description

A noise control method, apparatus, system and vehicle

[0001] Cross-references to related applications

[0002] This application claims priority to Chinese Patent Application No. 202411874314.4, filed on December 18, 2024, entitled "A Noise Control Method, Apparatus, System and Vehicle", the entire contents of which are incorporated herein by reference. Technical Field

[0003] This application relates to the field of noise control technology, and in particular to a noise control method, apparatus, system and vehicle. Background Technology

[0004] During vehicle operation, friction between airflow and the vehicle body, tires and the road surface, and engine noise generate significant low-frequency noise inside the car, directly impacting passenger comfort and well-being. To effectively suppress this low-frequency noise, numerous studies have focused on active noise cancellation (ANC) technology. The basic principle of ANC is to generate an inverse noise equal to or opposite to the external noise level to cancel it out, thus achieving noise reduction. ANC technology demonstrates good control over low-frequency noise and is expected to become one of the main solutions for in-vehicle noise control in the future.

[0005] Currently, ANC technology is usually integrated with car seats to form noise-canceling seats. However, the noise-canceling stability of existing mainstream noise-canceling seats is not good, especially when the user's head covers the acoustic components in the noise-canceling seat, such as speakers or microphones. The noise-canceling performance drops significantly, and there may even be a reverse noise amplification phenomenon, which can easily cause the noise-canceling system to diffuse or emit abnormal sounds, affecting the user's riding comfort.

[0006] In summary, providing passengers with a stable noise reduction experience is a pressing technical problem that needs to be solved in the field of in-vehicle noise control. Summary of the Invention

[0007] This application provides a noise control method, apparatus, system, and vehicle to improve the stability of noise reduction effects.

[0008] Firstly, this application provides a noise control method applicable to a noise control device, which may be a cockpit or its components, such as an in-vehicle system, domain controller, or vehicle controller, or it may be an external device or component, such as a user terminal or other vehicle. The method includes: controlling a speaker to play a noise-reducing signal, and acquiring a first reference signal, a first error signal, and first state information of the current secondary path; matching the first state information with the state information of multiple pre-stored secondary paths; if a matching pre-stored secondary path exists, determining filter coefficients based on the path parameters of the matching pre-stored secondary path, the first reference signal, and the first error signal, and adjusting the filter coefficients using a first step size; if no matching pre-stored secondary path exists, determining filter coefficients based on the path parameters of a default secondary path, the first reference signal, and the first error signal, and adjusting the filter coefficients using a second step size, where the second step size is smaller than the first step size; and then updating the noise-reducing signal using the adjusted filter coefficients.

[0009] Based on the above noise control method, multiple pre-configured secondary paths can be configured. When the current secondary path does not belong to one of the pre-configured paths, the filter coefficients are determined by combining the default path parameters and a small step size. This reduces the noise reduction amount per update, avoids abrupt changes in noise that could cause abnormal sounds or divergence, and improves the stability of the noise reduction effect. Conversely, when the current secondary path belongs to one of the pre-configured paths, the filter coefficients are determined by combining the path parameters of that pre-configured path and a large step size. This allows for faster convergence to the target noise reduction amount, achieving greater noise reduction depth and convergence speed. Therefore, this noise control method can balance noise reduction stability and efficiency.

[0010] In one possible design, the default secondary path is one of several pre-stored secondary paths, which are: a secondary path in which neither the speaker nor the error sensor is obstructed; the other pre-stored secondary paths include at least one of the following: a secondary path in which at least one but not all of the speakers are obstructed; a secondary path in which all speakers are obstructed; a secondary path in which at least one but not all of the error sensors are obstructed; and a secondary path in which all error sensors are obstructed.

[0011] Based on the above design, secondary paths corresponding to various possible scenarios can be pre-configured, including scenarios where neither the speakers nor the error sensors are blocked, scenarios where some (at least one and not all) of all speakers are blocked, scenarios where some (at least one and not all) of all error sensors are blocked, scenarios where all speakers are blocked, scenarios where all error sensors are blocked, and so on. The more scenarios configured, the easier it is to find matching secondary paths when performing noise control, which can improve the probability of fast convergence.

[0012] In one possible design, before controlling the speaker to play the noise-reduced signal, it is also possible to first determine to enable the noise reduction operation, and acquire the second reference signal collected by the reference sensor and the second error signal collected by the error sensor. Based on the second reference signal, the second error signal and the path parameters of the default secondary path, the initial filter coefficients are determined, and the initial filter coefficients are adjusted using the first step length. The noise-reduced signal is then determined based on the adjusted initial filter coefficients and the second reference signal.

[0013] Based on the above design, a large step size can be used to determine the noise reduction signal during the initial noise reduction, thereby improving the noise reduction effect of the initial noise reduction. Furthermore, the noise reduction signal from the initial noise reduction will also serve as the basis for the next noise reduction, and will be updated in the next noise reduction by combining other information, so as to achieve the stability of noise reduction through more comprehensive information.

[0014] In a further possible design, the initial filter coefficients are adjusted using a first-step length. The denoising signal is then determined based on the adjusted initial filter coefficients and a second reference signal. Specifically, this could involve iteratively updating the initial filter coefficients using the first-step length until they no longer diverge. The denoising signal is then determined using these updated initial filter coefficients and the second reference signal. Alternatively, the denoising operation can be partially or completely exited when an exit condition is met. For example, the initial filter coefficients are multiplied by the first-step length, and then the product is convolved with the second reference signal to obtain the output signal. The strength of the output signal is then checked against a set strength. If it is, the first-step length is reduced, and the reduced first-step length is multiplied again by the original initial filter coefficients. This process is repeated until the strength of the output signal is no greater than the set strength. This output signal is then used as the denoising signal. Alternatively, the denoising operation can be partially or completely exited when the step size is less than the set step size.

[0015] Based on the above design, a non-divergent noise reduction signal can be generated iteratively. This noise reduction signal is then output to the speaker so that the speaker plays inverse noise according to the noise reduction signal, thereby effectively reducing noise at the target point with a good noise reduction effect.

[0016] In a further possible design, the noise reduction operation is determined to be activated. Specifically, in response to the noise reduction command, the vehicle speed and the status of the body switch components are obtained, and it is determined that the vehicle speed is greater than a first vehicle speed and that the status of the body switch components is closed. The body switch components include one or more of the following: doors, windows, and sunroof.

[0017] Based on the above design, by determining that the vehicle is in motion and the cabin is sealed before activating noise reduction, road noise during vehicle movement can be reduced, avoiding the impact of wind noise on the noise reduction effect. Furthermore, if the cabin is not sealed, wind noise interference is very significant; in this case, noise reduction can be omitted to save unnecessary resource waste.

[0018] In one possible design, the path parameter of any secondary path is a transfer function. The transfer function characterizes the path of the signal played by the loudspeaker as it propagates to the error sensor. Therefore, using the transfer function as a path parameter can accurately characterize the signal transmission relationship of the secondary path, which can help calculate the filter coefficients.

[0019] In one example of the above design, the path parameters of any secondary path can be obtained as follows: In a silent environment, configure the speaker and error sensor in the scene corresponding to each secondary path, control the speaker to play a test signal, and acquire the error signal collected by the error sensor. Based on the test signal and the error signal, determine the path parameters of each secondary path. For example, the transfer function of each secondary path can be determined based on the test signal and error signal, and this transfer function can be used as the path parameters of each secondary path.

[0020] In one possible design, obtaining the first state information of the current secondary path can specifically involve: controlling the speaker to play a detection signal and acquiring the first error signal collected by the error sensor; determining the first state information of the current secondary path based on the detection signal and the first error signal; the first state information includes one or more of the transfer function, cross-spectral function, and cross-correlation coefficient.

[0021] Based on the above design, the state information of the secondary path can be calculated by using a loudspeaker to play detection signals and an error sensor to collect error signals. In this way, the detection operation of the secondary path can be completed using only the acoustic components in the noise control system, without the need to introduce other components. This reduces the structural complexity of the noise control system and lowers the control cost.

[0022] In one example of the above design, controlling the speaker to play the detection signal can be specifically as follows: First, an initial signal is generated randomly or according to a set rule; then, a third error signal collected by an error sensor is acquired, and the initial signal is spectral sculpted and gain controlled based on the third error signal to obtain a detection signal. The spectral shape of the detection signal is close to the spectral shape of the third error signal. The noise intensity generated at the error sensor after the detection signal is played by the speaker is lower than the noise intensity of the third error signal, and the difference between the noise intensity of the detection signal and the noise intensity of the third error signal is within a set range; finally, the detection signal and the noise-reduced signal are superimposed, and the speaker is controlled to play the superimposed signal.

[0023] Based on the above example, by performing spectral shaping on the detection signal, its spectral shape can be made to approximate the spectral shape of the third error signal. Therefore, the detection signal can be masked by the third error signal. Based on the masking effect of the human ear, this ensures that the human ear cannot subjectively perceive it, or hardly perceives it at all. Furthermore, by controlling the gain of the detection signal, the noise intensity of the detection signal is prevented from being too high to affect human perception, while also preventing it from being too low to result in an excessively low signal-to-noise ratio. While ensuring that the noise is imperceptible to the human ear, the signal-to-noise ratio of the detection can be maximized, thereby enabling a more accurate determination of the secondary path state.

[0024] In one example of the above design, the first state information is matched with the state information of multiple pre-stored secondary paths. Specifically, for each pre-stored secondary path, the deviation between the first state information and the state information of the pre-stored secondary path is determined. If the deviation is within the allowable deviation fluctuation range, the first state information is determined to match the state information of the pre-stored secondary path.

[0025] Based on the above example, an allowable deviation fluctuation range can be configured in advance. Within the allowable deviation fluctuation range, even if there are some fluctuations in the current secondary path, it can be classified as a pre-configured pre-stored secondary path. In this way, a small range of deviation fluctuations can be allowed, avoiding frequent changes in noise reduction parameters and improving the user's noise reduction experience.

[0026] In one example of the above design, the state information of multiple pre-stored secondary paths is obtained as follows: In a silent environment, the speaker and error sensor are configured in the scene corresponding to each pre-stored secondary path. The speaker is controlled to play a test signal, and the error signal collected by the error sensor is acquired. The state information of each pre-stored secondary path is determined based on the test signal and the error signal. For example, one or more of the transfer function, cross-spectral function, and cross-correlation function of each pre-stored secondary path can be determined based on the test signal and error signal of each pre-stored secondary path, and this one or more pieces of information can be used as the state information of each pre-stored secondary path.

[0027] Based on the above design, the two acoustic components, loudspeakers and error sensors, can be used for pre-modeling and testing to obtain the state information of multiple pre-stored secondary paths without the need to set up other components in the cockpit, which can reduce the complexity of modeling and testing.

[0028] In another possible design, the first state information of the current secondary path can also be obtained visually, such as by acquiring the cabin images captured by the cabin camera, identifying the image features of the areas where the speakers and error sensors are located in the cabin images, and matching the image features with the image features of multiple pre-stored secondary paths.

[0029] Based on the above design, the cabin images captured by the cabin camera can be used to calculate the status information of the secondary path. In this way, non-acoustic components in the cabin can be called to complete the detection operation of the secondary path, improving the flexibility of noise control.

[0030] In one possible design, the adjusted filter coefficients are used to update the noise reduction signal. Specifically, the adjusted filter coefficients are subjected to divergence detection. If there is no divergence trend, the adjusted filter coefficients are used to update the noise reduction signal. If there is a divergence trend, the step size is reduced, and the filter coefficients are readjusted using the reduced step size. The readjusted filter coefficients are then used to update the noise reduction signal.

[0031] Based on the above design, by detecting whether there is a divergence trend before updating the noise reduction signal, the adjustment step size can be reduced in time when there is a divergence trend, so as to avoid excessive noise reduction and the generation of obvious abnormal sound.

[0032] In a further possible design, divergence detection can be performed on the adjusted filter coefficients. Specifically, the adjusted filter coefficients and the first reference signal can be convolved to obtain the output signal. If the signal strength of the output signal is less than or equal to the set strength, it is determined that the adjusted filter coefficients do not have a divergence trend. If the signal strength of the output signal is greater than the set strength, it is determined that the adjusted filter coefficients have a divergence trend.

[0033] Based on the above design, the set intensity can be used as the basis for divergence detection. By simply comparing the output signal with the set intensity, it is possible to quickly determine whether there is a divergence trend, resulting in high efficiency in divergence detection.

[0034] In a further possible design, when the reduced step size is less than or equal to the set step size, the noise reduction operation can be completely or partially exited.

[0035] Based on the above design, the noise reduction operation can be partially or completely exited when the excessive divergence trend is detected. For example, some or all speakers can be turned off, or the noise reduction signal played by some or all speakers can be set to zero, so as to avoid unnecessary noise reduction processes and save noise reduction resources.

[0036] Secondly, this application provides a noise control device, which may be a cockpit, a component within the cockpit (such as a processor, vehicle infotainment chip, or chip system), or a component outside the cockpit (such as a user terminal, other vehicle, or components thereof). The noise control device may include units or modules for performing the various steps of the first aspect or any of the designs or examples in the first aspect.

[0037] In one example, the noise control device may include a control unit, an acquisition unit, a determination unit, and an update unit. This noise control device can be used to perform various steps of the first aspect or any of the designs or examples in the first aspect. Specifically, the control unit is used to control the speaker to play a noise-reduced signal; the acquisition unit is used to acquire a first reference signal, a first error signal, and first state information of the current secondary path; the determination unit is used to match the first state information with the state information of multiple pre-stored secondary paths. If a matching pre-stored secondary path exists, filter coefficients are determined based on the path parameters of the matching pre-stored secondary path, the first reference signal, and the first error signal, and the filter coefficients are adjusted using a first step size. If no matching pre-stored secondary path exists, filter coefficients are determined based on the path parameters of a default secondary path, the first reference signal, and the first error signal, and the filter coefficients are adjusted using a second step size, where the second step size is smaller than the first step size; the update unit is used to update the noise-reduced signal using the adjusted filter coefficients.

[0038] Thirdly, this application provides a noise control device, which can be a cockpit, a component within the cockpit (such as a processor, vehicle infotainment chip, or chip system), or a component outside the cockpit (such as a user terminal, other vehicles, or components thereof). The noise control device may include a processor, and optionally, may also include a memory (or storage medium). The memory stores program instructions; the processor can read the program instructions from the memory, causing the noise control device to execute the methods provided in the first aspect or any of the designs or examples in the first aspect.

[0039] Optionally, there may be one or more processors and one or more memories.

[0040] Optionally, the memory can be integrated with the processor, or the memory can be set up separately from the processor.

[0041] In one possible design, the noise control device may further include a transceiver. The transceiver is used to receive and transmit signals; a processor is used to execute program instructions in response to signals received by the transceiver, causing the noise control device to perform the methods provided in the first aspect or any of the designs or examples in the first aspect. Optionally, the transceiver may include a transmitter and a receiver.

[0042] In another possible design, the noise control device also includes a communication interface, to which the processor is coupled. The processor reads program instructions from memory, invokes the communication interface to communicate with other devices, and executes the methods provided in the first aspect or any of the designs or examples in the first aspect. Optionally, the communication interface can be a transceiver, or an input / output interface. Optionally, the transceiver can be a transceiver circuit. Optionally, the input / output interface can be an input / output circuit.

[0043] Optionally, when the noise control device is a chip or chip system, the communication interface can be an input / output interface, interface circuit, output circuit, input circuit, pin, or related circuit on the chip or chip system. The processor can also be manifested as a processing circuit or logic circuit.

[0044] Fourthly, this application provides a noise control system, including a noise control device, such as the noise control device described in any of the second or third aspects above, which is used to perform a method as described in any of the designs or examples in the first aspect above.

[0045] In one possible design, the noise control system further includes a loudspeaker, a reference sensor, and an error sensor, with the noise control device connected to the loudspeaker, the reference sensor, and the error sensor respectively; the loudspeaker is used to play a noise-reduced signal, the reference sensor is used to acquire a first reference signal, the error sensor is used to acquire a first error signal, and the noise control device is used to perform the method as described in any of the designs or examples of the first aspect above, based on the noise-reduced signal, the first reference signal, and the first error signal.

[0046] In a further possible design, the noise control system also includes a vision sensor. The noise control device is connected to the vision sensor, which is used to acquire images inside the cabin and send them to the noise control device. The noise control device is used to determine the first state information of the current secondary path based on the images inside the cabin.

[0047] In further possible designs, the noise control system is located in the cabin, the error sensor is located in the headrest, and the speakers are located in the headrest, seat, or inside the door.

[0048] Fifthly, this application provides an electronic device connected to a noise control vehicle for communicating with the noise control vehicle to implement the noise control method as described in the first aspect or any of the designs described in the first aspect. The electronic device may include units or modules for implementing the noise control method as described in the first aspect or any of the designs described in the first aspect, such as including the noise control device as described in any of the second or third aspects described above.

[0049] Alternatively, the electronic device may be a user terminal or other vehicle, or other device that can control the vehicle's speakers to produce sound.

[0050] Sixthly, this application provides a means of transportation that includes a noise control device provided in either the second or third aspect above, or a noise control system provided in the fourth aspect above.

[0051] In a seventh aspect, this application provides a computer-readable storage medium storing a computer program that, when executed by a computer, causes the computer to perform the method provided in the first aspect or any of the designs in the first aspect. Optionally, the computer may be a cockpit or a component thereof.

[0052] Eighthly, this application provides a computer program product that, when run on a computer, causes the computer to perform the method provided in the first aspect or any of the designs in the first aspect. Optionally, the computer may be a cockpit or a component thereof.

[0053] Ninthly, this application provides a chip for reading a computer program stored in a memory and executing the method provided in the first aspect or any of the designs in the first aspect.

[0054] Alternatively, the chip can be an in-vehicle infotainment chip.

[0055] Optionally, the chip may include a processor coupled to a memory for reading a computer program stored in the memory to implement the method provided by the first aspect or any of the designs in the first aspect.

[0056] Optionally, the chip may also include components such as memory, communication interface, and power supply module. The memory is used to store computer programs; the communication interface is used to receive and send data; and the power supply unit is used to supply power to the processor.

[0057] In a tenth aspect, this application provides a chip system including a processor for supporting a computer to implement the methods provided in the first aspect or any of the designs in the first aspect.

[0058] In one possible design, the chip system also includes memory for storing the computer's necessary programs and data. The chip system can consist of chips or include chips and other discrete components.

[0059] The technical effects that can be achieved in aspects two through ten above can be referred to the description of the beneficial effects in aspect one above, and will not be repeated here. Attached Figure Description

[0060] Figure 1a illustrates a possible application scenario provided by this application;

[0061] Figure 1b illustrates another possible application scenario provided by this application;

[0062] Figure 2 illustrates a possible architecture diagram of a noise control system provided in this application;

[0063] Figure 3a illustrates an exemplary schematic diagram of the installation location of an error sensor provided in this application;

[0064] Figure 3b illustrates an exemplary installation location diagram of a speaker provided in this application;

[0065] Figure 3c illustrates a schematic diagram of an error sensor and a speaker both installed in a soft headrest according to this application.

[0066] Figure 3d illustrates an exemplary diagram of the installation location of an in-cabin camera provided in this application;

[0067] Figure 4 illustrates a flowchart of a noise control method provided in this application;

[0068] Figure 5 is an exemplary schematic diagram of the presentation form of a human-computer interaction interface provided in this application;

[0069] Figure 6 illustrates an exemplary noise reduction process for the first cycle provided in this application;

[0070] Figure 7 illustrates a scenario diagram of various secondary paths provided in this application;

[0071] Figure 8 illustrates a flowchart of a noise control method provided in Implementation Scheme 1;

[0072] Figure 9 illustrates an exemplary implementation flow diagram of spectrum shaping and gain control provided in Implementation Scheme 1;

[0073] Figure 10a illustrates an exemplary flowchart of another noise control method provided in Implementation Scheme 1;

[0074] Figure 10b illustrates, for example, a functional principle diagram of another noise control method provided in Implementation Scheme 1;

[0075] Figure 11 illustrates a flowchart of a noise control method provided in Implementation Scheme 2;

[0076] Figure 12a illustrates an exemplary flowchart of another noise control method provided in Implementation Scheme 2;

[0077] Figure 12b illustrates, for example, a functional principle diagram of another noise control method provided in Implementation Scheme 2;

[0078] Figure 13 illustrates a schematic diagram of the structure of a noise control device provided in this application;

[0079] Figure 14 illustrates a schematic diagram of another noise control device provided in this application. Detailed Implementation

[0080] The embodiments of this application will now be described in detail with reference to the accompanying drawings.

[0081] The following provides explanations for some of the terms used in this application. It should be noted that these explanations are for the convenience of those skilled in the art and do not constitute a limitation on the scope of protection claimed in this application.

[0082] I. ANC Technology

[0083] Typically, the components used to implement ANC (Asynchronous Noise Cancellation) technology include an error sensor, a reference sensor, a speaker, and a controller. The error sensor collects ambient noise at the target point and sends it to the controller. The controller combines the signals from the error sensor and the reference sensor to generate anti-phase noise that is 180 degrees out of phase with the ambient noise and sends it to the speaker, which plays this anti-phase noise. The spectrum of this anti-phase noise has the same or similar amplitude as the ambient noise at the target point, only with opposite phase. Therefore, superimposing this anti-phase noise onto the ambient noise at the target point effectively suppresses and eliminates ambient noise.

[0084] II. Secondary Paths

[0085] In ANC technology, the secondary path refers to the transmission path from the speaker to the error sensor.

[0086] III. The Masking Effect of the Human Ear

[0087] The masking effect of the human ear refers to the phenomenon where the threshold for perceiving a sound increases due to the presence of another sound. The masking effect is mainly related to factors such as the frequency, intensity, and timing of the sound. High-frequency sounds are generally more easily masked by low-frequency sounds, loud sounds usually mask weak sounds, and sounds occurring simultaneously are more likely to mask sounds that occur later. These factors work together to influence the human ear's perception of sound.

[0088] IV. Soft headrests and seat headrests

[0089] In this application, "soft headrest" refers to headrests or neck pillows that are purchased by the user or included with the vehicle purchase, and are considered non-original vehicle parts. "Seat headrest," on the other hand, refers to a headrest that is part of the seat itself; it can be a floating headrest or one integrated with the seat back, and is considered an original vehicle part. In some scenarios, seat headrests may be equipped with speakers and error microphones for active noise cancellation.

[0090] The preceding text introduced some of the terms used in this application. The following text introduces the possible application scenarios of this application.

[0091] In one possible implementation, the noise control method of this application can be applied to vehicles, such as cars, trucks, buses, trains, recreational vehicles, station wagons, vans, amusement park vehicles, construction vehicles, trams, golf carts, sightseeing vehicles, patrol cars, intelligent vehicles, and digital cars. Please refer to Figure 1a, which exemplifies a possible application scenario of this application. In this scenario, a car is used as an example. One or more seats in the car can be configured as noise-reducing seats, and speakers are installed within these seats. Figure 1a only shows an example of a speaker integrated into the headrest, but the speaker can also be placed in other locations in the cabin, such as inside the doors or on the instrument panel (IP). The speaker can play inverse noise of the in-vehicle environment to the user based on the noise control method provided in this application, thereby providing a quieter environment for the user sitting in the noise-reducing seat, reducing discomfort caused by in-vehicle noise when resting or sleeping, or it can be used to achieve immersive sound playback, such as improving the playback effect of audio-visual content and enhancing the user's viewing experience. When the driver's seat is set to noise-canceling mode, it can also improve the driver's ability to hear voice information such as navigation reminders or driving reminders by playing contrasting noise, thereby achieving safe and efficient driving.

[0092] It should be understood that the above application scenarios are merely examples, and the noise control method provided in this application can also be applied to other possible scenarios, not limited to those listed above. For example, the noise control method can also be applied to other modes of transportation, such as subways, high-speed trains, ships, ferries, passenger ships, airplanes, or helicopters, to reduce low-frequency noise inside the cabin, providing users with a relatively quiet environment and making their journey more comfortable. Furthermore, the noise control method can also be applied to smart home scenarios, especially in homes near subway stations, train stations, airports, or construction sites, as an auxiliary means of home noise reduction, improving the user's smart home experience. Additionally, the noise control method can be applied to office scenarios to reduce the impact of other users' typing or talking on the user's workstation, improving the user's concentration. Moreover, the noise control method can also be applied to public areas, such as cinemas, shopping malls, high-speed rail stations, airports, bus stations, hospitals, schools, parks, communities, squares, churches, etc., to reduce the noise level of the user's surrounding environment by isolating external noise, allowing the user to relax and feel more comfortable. And so on. These are just a few examples.

[0093] It should be noted that the application scenarios described in this application are for the purpose of more clearly illustrating the technical solutions of this application, and do not constitute a limitation on the technical solutions provided in this application.

[0094] For ease of understanding, the following description uses the application of noise control methods to vehicles as an example. However, it should be understood that the noise control content discussed below is also applicable to other vehicles, other equipment, or other fields, and this application does not specifically limit it in this regard.

[0095] In current mainstream automotive ANC solutions, a fixed set of secondary path parameters is used to update the noise reduction filter coefficients throughout the vehicle's driving process. These secondary path parameters are obtained by pre-modeling the vehicle's noise reduction process under ideal driving scenarios, where all acoustic components inside the vehicle are unobstructed. These secondary path parameters are configured in the vehicle before it leaves the factory. After the vehicle leaves the factory, whenever active noise cancellation is activated, these secondary path parameters are used to calculate the noise reduction signal. These secondary path parameters demonstrate good noise reduction performance for in-vehicle driving noise under ideal driving scenarios.

[0096] However, in real-world driving scenarios, the position or state of passengers inside the vehicle is constantly changing. At any given moment, a user's body may obstruct an acoustic component, such as a speaker or error sensor. This causes the actual secondary path to differ significantly from the secondary path in an ideal driving scenario. For example, taking the scenario shown in Figure 1a, assuming the speakers are integrated into the sides of the headrest, when the user is seated in the center, their head is in the central area of ​​the headrest and will not obstruct the speakers on either side. However, if the user moves their head significantly to the left, they are likely to obstruct the speaker on the left side of the headrest, and vice versa. In some scenarios, error sensors are also integrated into the headrest; therefore, head movement may also obstruct the error sensor. In other scenarios, it may not be the user directly obstructing the speaker or error sensor, but rather other objects inside the vehicle. For example, as shown in Figure 1b, many users often place a self-purchased or gifted soft headrest or neck pillow on the seat after purchasing a vehicle for the sake of comfort. However, the position of the soft headrest or neck pillow happens to cover the headrest area of ​​the seat, which means that the speakers and microphones set on the left and right sides of the headrest are blocked at the same time. Actual tests have found that this scenario has the greatest impact on secondary paths.

[0097] Understandably, when the secondary path from the speaker to the error sensor changes, the path parameters of this secondary path will deviate significantly from the pre-modeled parameters of the secondary path under the ideal driving scenario. Continuing to use the secondary path parameters from the ideal driving scenario to update the noise reduction filter coefficients will lead to a decrease in the noise reduction performance of the noise reduction system, and may even result in not only a failure to reduce noise but also amplification. Continuous amplification can easily cause the noise reduction system to diverge, while excessive amplification can easily cause abnormal sounds in the noise reduction system. Both of these phenomena reduce the stability of the noise reduction system.

[0098] To address the issue of noise reduction systems easily scattering or generating abnormal sounds, the industry has proposed several solutions, primarily including the following two:

[0099] Solution 1 involves pre-modeling and storing several sets of secondary path parameters. During active noise cancellation control, the vehicle audio signal is used as the detection signal to identify the secondary path. The pre-stored secondary path parameter that best matches the identified result is then selected to update the noise cancellation filter coefficients. Because multiple sets of secondary path parameters are pre-stored, this solution can achieve good noise reduction for various pre-stored secondary path scenarios, reducing the probability of divergence or abnormal sound phenomena in the noise cancellation system. However, this solution uses the vehicle audio signal as the detection signal, therefore, it can only identify secondary paths when there is an in-vehicle audio signal stream. For example, it can only identify secondary paths when the user is listening to music, but it cannot identify them when the user is not listening to music, severely limiting its application scenarios. Furthermore, when the user is not listening to music, changes in secondary paths cannot be detected, and divergence or abnormal sound phenomena will still occur.

[0100] Solution 2, in the active noise cancellation control process, no longer uses pre-stored secondary path parameters to update the filter coefficients. Instead, it utilizes online identification signals for real-time online secondary path identification or recognition. To ensure that the online identification signal is not audible to the human ear and thus affects the noise cancellation effect, an adaptive gain control method is first used to adjust the gain of the online identification signal. Then, the vehicle's speakers are controlled to emit the online identification signal. Subsequently, error signals collected by error sensors are acquired, and online secondary path identification is performed in real time based on these error signals. Secondary path parameters are then modeled and used to update the noise cancellation filter coefficients. However, in practical applications, the gain of the online identification signal needs to be adjusted to be very small, such as less than or equal to 5dB, to be imperceptible to the human. Therefore, the signal-to-noise ratio of the online identification signal is very low. When the signal-to-noise ratio is too low, the accuracy of the identified secondary path is often insufficient, affecting the noise cancellation depth and convergence speed, and may even be unusable for updating the filter coefficients. Therefore, this solution cannot effectively avoid the phenomenon of divergence or abnormal sounds in the noise cancellation system.

[0101] In view of this, this application provides a noise control method. This method pre-configures multiple pre-stored secondary paths. When a pre-stored secondary path is matched, a larger step size and the path parameters of the pre-stored secondary path are used to update the filter coefficients, ensuring faster noise reduction depth and convergence speed. When no pre-stored secondary path is matched, a smaller step size and the path parameters of the default secondary path are used to update the filter coefficients, thereby reducing the amount of noise reduction and ensuring the stability of noise reduction in unknown scenarios, avoiding divergence or abnormal sounds. In other words, this method can effectively avoid divergence or abnormal sounds while maximizing noise reduction depth and speed, balancing noise reduction stability and efficiency.

[0102] Based on the above, the noise control scheme provided in the embodiments of this application will be described in detail below with reference to Figures 2 to 14.

[0103] In the various embodiments of this application, unless otherwise specified or in case of logical conflict, the terminology and / or descriptions of different embodiments are consistent and can be referenced by each other. The technical features of different embodiments can be combined to form new embodiments according to their inherent logical relationship.

[0104] Furthermore, in the various embodiments of this application, whenever terms such as "location," "value," "time," "range," or similar terms are used, they do not refer to absolute locations, values, times, or ranges, and a certain degree of engineering error is permissible. For example, they can refer to approximate locations, approximate values, approximate times, approximate ranges, etc., which have a certain degree of error compared to absolute locations, values, times, and ranges. This error is mainly limited by factors such as manufacturing capabilities and measurement accuracy.

[0105] Taking the application of noise control schemes in vehicles as an example, we will first introduce a type of noise control system to which the noise control scheme is applicable.

[0106] Please refer to Figure 2, which shows a possible architecture diagram of a noise control system provided in this application. This architecture includes a noise control device, a reference sensor, an error sensor, and a loudspeaker. Optionally, a vision sensor is also included. Each component is described below by example.

[0107] A noise control device, a core component of a noise control system, can be a device specifically designed for noise control or a device that performs other functions in addition to noise control. For example, in one scenario, the noise control device could be a control unit within a vehicle, such as the infotainment system, cockpit domain controller (CDC), other domain controllers, or vehicle control unit (VCU). This utilizes the existing control unit within the vehicle to achieve noise control, improving the utilization rate of in-vehicle components. Alternatively, to reduce the workload of the in-vehicle control unit, the noise control device could be an additional control unit specifically designed for active noise cancellation, such as a standalone digital signal processing (DSP) chip. This DSP chip contains all the components necessary for digital signal processing, including but not limited to: power amplifiers, analog-to-digital converters (ADCs), digital-to-analog converters (DACs), and processing units. This DSP chip is independent of the vehicle and can achieve noise control functionality by connecting to relevant vehicle components, such as reference sensors, error sensors, and speakers. Alternatively, in another example, the noise control device could be a terminal device or other vehicle. This device could connect via a network to relevant components in the vehicle to be noise-reduced, such as reference sensors, error sensors, and speakers, to implement a noise control scheme for the vehicle. Understandably, in some examples, the noise control device could be a combination of the above examples; for instance, some functions of the noise control device might be implemented by a separately configured control unit, while other functions might be implemented by a control unit within the vehicle. And so on, without further listing.

[0108] A reference sensor, a hardware input unit of a noise control system, is primarily responsible for acquiring raw noise signals. Raw noise signals refer to the signals generated by the interaction between noise sources and the vehicle. During vehicle operation, the most common noise source is road noise, corresponding to vibration signals generated by the interaction between the ground and the wheels. Therefore, to acquire vibration signals, the reference sensor can be an accelerometer or vibration sensor. In some scenarios, a microphone can also be used. Taking an accelerometer as an example, accelerometers are typically mounted on the vehicle body, hence also called body accelerometers. Their installation locations include, but are not limited to, the front end of the vehicle's center of gravity, wheels, and B-pillars. Accelerometers can acquire the vehicle's acceleration signals during operation, and the vehicle's vibration signals can be further calculated from these acceleration signals. Optionally, when accelerometers are mounted on the wheels, considering that the vibrations of each wheel are independent, accelerometers can be installed on each wheel separately to obtain comprehensive vibration signals.

[0109] Error sensors, also known as microphones, are another type of hardware input unit in noise control systems. Error sensors are typically implemented using microphones or microphone arrays, hence the name error microphones, and are primarily responsible for collecting noise signals near the user's ear. To improve the accuracy of noise collection, error sensors can be placed in the cabin close to the user's ear, such as integrated into the seat headrest. The headrest can be a floating headrest, suspended above the seat back, as shown in Figure 3a(A). Alternatively, the area on the seat back intended for head support can itself serve as a headrest, as shown in Figure 3a(B). In some scenarios, users can also place a soft headrest or neck pillow in this area to increase comfort, as shown in Figure 3a(C).

[0110] Optionally, the number of error sensors can be one or more. For example, taking integration into a floating headrest as an example, in one example, to achieve separate acquisition of noise in the user's left and right ears, an error sensor can be installed in the left and right areas of the floating headrest, respectively, as shown in Figure 3b(A). Alternatively, in another example, to further improve the acquisition accuracy, multiple error sensors can be installed in and around the left and right areas of the floating headrest, respectively, as shown in Figure 3b(B). Or, in yet another example, multiple error sensors can be directly integrated into the seat headrest, as shown in Figure 3c(A). The more error sensors there are, the more error signals are acquired, resulting in higher noise reduction accuracy, but also higher cost and structural complexity, and lower noise reduction efficiency. Therefore, if the actual scenario requires high noise reduction accuracy, more error sensors can be set, while if the requirement for high noise reduction efficiency is high, fewer error sensors can be set.

[0111] A loudspeaker, also known as a horn, is a device that converts electrical signals into sound signals. Loudspeakers are hardware input / output units in noise control systems, primarily responsible for playing noise-reducing signals within the cabin. These noise-reducing signals, also called inverse noise signals, are calculated by the noise control device based on a reference signal collected by a reference sensor and an error signal collected by an error sensor. This inverse noise signal is superimposed on the noise signal within the cabin, effectively purifying the acoustic environment. In some scenarios, loudspeakers can also perform other functions, such as immersive sound playback, navigation playback, or real-time traffic alerts.

[0112] Optionally, the number of speakers can be one or more, which can be integrated into any location within the cabin, including but not limited to: the inside of the doors, the interior glass of the windows, the interior trim film, the center console, ambient lighting, the steering wheel, the seat back, and the headrests. For example, as shown in Figure 3c(A), an example of installing speakers in a headrest is illustrated, with one speaker on the left and right sides of the seat back corresponding to the head position. Another example, as shown in Figure 3c(B), is an example of installing speakers in a floating headrest. When the speakers and error sensors are installed together in a floating headrest, as shown in Figure 3c(A) or (B), a certain positional offset is required between them to reduce the impact of the anti-phase noise signal played by the speakers on the error signal collected by the error sensor.

[0113] Visual sensors, an optional hardware input unit in noise control systems, include, but are not limited to, in-cabin cameras. In-cabin cameras are typically mounted centrally above the windshield, as shown in Figure 3d. In some examples, they can also be mounted in other locations within the cabin, such as the centrally located above the rear windshield or the inner door handle. The in-cabin camera captures images of the cabin interior, and based on the image states of error sensors and speakers within these images, it can determine whether secondary paths in the current noise reduction scenario have changed. In other scenarios, visual sensors can also be radar (e.g., lidar, millimeter-wave radar, or ultrasonic radar), which can be used to detect point cloud data of the in-vehicle environment. This point cloud data can assist the noise control device in more accurately identifying changes in secondary paths.

[0114] It should be understood that the architecture of the noise control system described above is only an example. In other examples, the noise control system may include more, fewer, or different components, and each component may include more, fewer, or different elements. Furthermore, the components shown or not shown may be combined or divided in any way. The division of the components shown is merely a logical functional division. In actual implementation, they may be fully or partially integrated into a single physical entity or physically separated. This application does not impose any specific limitations on this.

[0115] Based on the noise control system shown in Figure 2, please refer to Figure 4, which shows a flowchart of a noise control method provided in this application. This method is applicable to noise control devices, such as the noise control device shown in Figure 2. As shown in Figure 4, the method includes the following steps:

[0116] Step 401: Control the speaker to play the noise reduction signal.

[0117] Optionally, when a user needs noise reduction, they can issue a noise reduction command to the noise control device, which instructs the device to activate the noise reduction operation. This noise reduction command can be triggered by speaking, clicking the human-machine interface, pressing a vehicle button (such as double-clicking the cigarette lighter), making a set gesture, sending a text message, or sending a notification.

[0118] For example, taking the noise reduction command triggered by clicking the human-machine interface as an example, the noise control device can also be connected to the vehicle's infotainment screen. If the noise control device is the vehicle's infotainment system, the controller is directly connected to the screen. If the noise control device is not the infotainment system, it can be connected to the system, indirectly connecting to the screen. After the vehicle is powered on, if the user starts the infotainment system, it can control the screen to display the initial interface. As an example, the initial interface is shown in Figure 5, which includes multiple function buttons (controls), including an "Active Noise Reduction" button. If the user needs noise reduction, they can click the "Active Noise Reduction" button on the initial interface. The infotainment screen detects this click, generates a noise reduction command, and sends it to the system. If the noise control device is the infotainment system, it executes the following steps in response to the command. Conversely, if the noise control device is not the infotainment system, it can forward the command to the device, which then executes the following steps.

[0119] For example, taking voice-triggered noise reduction commands as an example, the noise control device can also be connected to an in-vehicle voice device, such as a car audio system or car speakers. After the vehicle is powered on, the in-vehicle voice device automatically starts and waits for the user's voice messages. If the user utters a command such as "Turn on active noise cancellation," "Cockpit noise reduction," or similar phrases, the in-vehicle voice device will collect and perform semantic analysis. Based on the semantic analysis results, the in-vehicle voice device will generate a noise reduction command and send it to the noise control device. The noise control device will then respond to the noise reduction command by executing the following steps.

[0120] For example, taking a noise reduction command triggered by sending a message as an example, the noise control device can also connect to a user terminal via a cloud server. The user terminal could be a mobile phone, laptop, or smart glasses. When a user needs noise reduction, they can open a pre-installed noise reduction application (APP) on their terminal and select the "Active Noise Reduction" button on the APP interface. The noise reduction APP generates a noise reduction command based on this operation and calls the user terminal's communication function to send the command to the cloud server. The cloud server verifies the noise reduction command; if the verification is successful, it forwards the command to the noise control device. The noise control device responds to the command by executing the following steps.

[0121] Of course, there are other triggering methods, which will not be listed here.

[0122] Optionally, after receiving a noise reduction command, the noise control device can first check whether the current vehicle status meets the preset noise reduction conditions before starting the noise reduction operation. If it does, the noise reduction operation can be started. If it does not, the noise reduction operation will not be started. If the current noise control system is already on, the noise control system can also be turned off, such as turning off the speakers or error sensors used for noise reduction. Optionally, the user can also be returned with "Noise reduction conditions not met, noise reduction failed" or similar content. The preset noise reduction conditions can be configured according to actual noise reduction needs. For example, in one example, if the noise reduction is for road noise during vehicle movement, the preset noise reduction conditions can be configured as follows: vehicle speed greater than or equal to a first vehicle speed, and all body switch components are in the off state. Body switch components refer to body components used to isolate the external space and the internal space of the vehicle. Most vehicles include doors and windows, and in vehicles with sunroofs, the sunroof is also included.

[0123] In the above description, the first vehicle speed can be understood as the minimum speed at which the vehicle is in motion, such as 10 km / h. Optionally, there can also be a second vehicle speed to represent the maximum speed at which the vehicle is in motion, such as 150 km / h. When the vehicle speed exceeds 150 km / h, wind noise becomes very loud, making noise reduction less effective. Therefore, when the vehicle speed is greater than 150 km / h, although the vehicle is in motion, there is no need to activate noise reduction to save unnecessary resources. Based on this, in another example, the preset noise reduction conditions can also be configured as follows: the vehicle speed is greater than or equal to the first vehicle speed and less than or equal to the second vehicle speed, and all body switch components are in the off state.

[0124] Taking a vehicle speed of 10 km / h and a speed of 150 km / h as an example, and considering vehicle body switches including doors, windows, and sunroof, the noise control device, upon receiving a noise reduction command, can interact with the vehicle speed sensor to obtain the current vehicle speed, and with the door Hall effect sensors, window Hall effect sensors, and sunroof Hall effect sensors to obtain the door, window, and sunroof statuses. If the current vehicle speed is between 10 km / h and 150 km / h (inclusive), and all doors, windows, and sunroof are closed, the noise reduction command takes effect, confirming the activation of noise reduction operation. Conversely, if at least one condition is not met, the noise reduction command is not activated, meaning noise reduction operation is not initiated.

[0125] Optionally, after determining that noise reduction operation should be initiated, the noise control device can send control signals to the reference sensor and the error microphone, instructing them to start operation. Once activated, the reference sensor acquires a reference signal at a set sampling frequency and sends it to the noise control device. Similarly, once activated, the error microphone acquires an error signal at a set sampling frequency and sends it to the noise control device. The noise control device performs a noise reduction operation each time it receives a reference signal and an error signal. Optionally, to ensure consistent signal acquisition, the sampling frequencies of the reference sensor and the error sensor can be set to the same, and at relatively high values, such as thousands or even tens of thousands of times per second. This results in very short sampling periods, effectively achieving real-time noise reduction, accurately responding to constantly changing noise environments, and achieving better noise reduction results.

[0126] Understandably, the first noise reduction operation is special in a series of noise reduction operations because the speaker has not yet played a noise-reduced signal when noise reduction is first initiated. The first noise reduction operation can determine a noise-reduced signal based solely on the reference signal and the error signal and drive the speaker to play it.

[0127] As an example, please refer to Figure 6, which shows a schematic diagram of the noise reduction process for the first noise reduction provided in this application, mainly including the following steps:

[0128] Step 601: Obtain the second reference signal acquired by the reference sensor and the second error signal acquired by the error sensor.

[0129] Understandably, when the sampling frequencies of the two sensors are set to the same, after receiving the control signal sent by the noise control device, the two sensors will start signal acquisition and transmission operations at the same sampling frequency. Therefore, the noise control device can receive the second reference signal and the second error signal simultaneously or almost simultaneously.

[0130] Furthermore, during the initial signal acquisition, since the speaker has not yet played the noise-reduced signal, the second reference signal acquired by the reference sensor is the vehicle vibration signal generated under the influence of noise sources such as road noise, while the second error signal acquired by the error sensor is the noise signal that the human ear can hear under the influence of noise sources such as road noise. When the reference sensor is a vehicle accelerometer, the second reference signal is the acceleration signal picked up by that accelerometer, which can be used to measure the amplitude of vehicle vibration. When the error sensor is an error microphone, the second error signal is the audio signal picked up by that error microphone, which can be used to characterize the noise that the human ear can hear.

[0131] Step 602: Determine the initial filter coefficients based on the second reference signal, the second error signal, and the path parameters of the default secondary path, and adjust the initial filter coefficients using the first step length.

[0132] Here, the default secondary path can be understood as the standard secondary path, which refers to the secondary path under ideal driving conditions. Ideal driving conditions refer to driving environments where the in-vehicle speakers and error sensors are not obstructed. The path parameters of the default secondary path are obtained through test modeling before the vehicle leaves the factory and are pre-stored in the vehicle. As an example, the specific steps of test modeling can be: First, park the vehicle in a quiet environment, such as a anechoic chamber; Second, configure the in-vehicle speakers and error sensors in an ideal driving scenario, i.e., a scenario where all error sensors and speakers are unobstructed and functioning normally. For example, when both the speakers and error sensors are located in the headrest area on the seat back, the scenario shown in Figure 7(A) can be configured; when the speakers are located on the seat back and the error sensors are located on the floating headrest, the scenario shown in Figure 7(G) can be configured; Third, control the in-vehicle speakers to play test signals and acquire the error signals collected by the error sensors; Fourth, determine the path parameters of the default secondary path based on the test signals and error signals. For example, in one example, the path parameter is configured as a transfer function. In this case, the test signal and the error signal can be deconvolved to calculate the transfer function, and this transfer function can be used as the path parameter of the default secondary path.

[0133] Optionally, the path parameters of the default secondary path can be pre-stored in the noise control device or in a relevant memory. After acquiring the second reference signal and the second error signal, the noise control device can read the path parameters from local or relevant memory, and then use the second reference signal, the second error signal, and the path parameters to obtain the initial filter coefficients. For example, in one example, an adaptive filtering algorithm can be used to process the second reference signal, the second error signal, and the path parameters to obtain the initial filter coefficients. The adaptive filtering algorithm can be any type of filtering algorithm, such as the filtered-X least-mean-square (FXLMS) algorithm. The FXLMS algorithm can adaptively adjust the filter coefficients based on the error between the input signal and the desired output signal, according to the second reference signal, the second error signal, and the path parameters, thereby achieving noise suppression.

[0134] Furthermore, the noise control device can also use the first step length to adaptively update the initial filter coefficients. For example, first multiply the first step length by the initial filter coefficients, and then determine whether the product diverges (see step 406 below for details, which will not be explained here). If it diverges, the step length is reduced, and the reduced step length is multiplied by the initial filter coefficients. Then it is determined whether it diverges again. This operation is continuously performed until it is determined that there is no divergence. The filter coefficients at this time are then used as the adjusted initial filter coefficients.

[0135] The first step size can be understood as the initial step size. The purpose of the first step size is to avoid problems caused by excessively large noise reduction in a single denoising operation. For example, if the calculated filter coefficients are inaccurate, the first step size can reduce the actual number of filter coefficients used for denoising, thus mitigating the negative impact of inaccurate filter coefficients on the actual denoising operation. Optionally, the value of the first step size can be a number less than 1 or greater than 1, depending on the algorithm used. For ease of understanding, as an example, the first step size can be chosen as a value greater than 0 and less than 1, relatively close to 1, such as 0.8. Because the first step size is relatively close to 1, it ensures a certain denoising speed, and if the filter coefficients are correct, it can converge to the target noise reduction amount relatively quickly.

[0136] Step 603: Determine the noise reduction signal based on the adjusted initial filter coefficients and the second reference signal.

[0137] For example, the noise control device can convolve the adjusted initial filter coefficients onto the second reference signal to obtain the noise-reduced signal. The noise control device can also send the calculated noise-reduced signal to the loudspeaker, which then emits an inverted sound wave based on the noise-reduced signal to reduce noise at the target point (such as the location of the human ear). At this point, the first noise reduction operation is complete.

[0138] Subsequently, upon receiving the reference signal and error signal again, the noise control device initiates a second noise reduction operation. In this second operation, the noise reduction signal is recalculated based on the received reference signal, error signal, and the noise reduction signal calculated in the first operation. This recalculated signal is then sent to the speaker, enabling the speaker to reduce noise at the target point. Upon receiving the reference signal and error signal again, the noise control device initiates a third noise reduction operation. This process continues until the noise reduction effect reaches the desired level or a preset exit condition is met, at which point the noise reduction operation terminates.

[0139] It should be noted that steps 402 to 406 below are described using the second noise reduction operation as an example. Subsequent third noise reduction operations, fourth noise reduction operations, and so on are all similar, and this application will not repeat them.

[0140] Step 402: Obtain the first reference signal, the first error signal, and the first state information of the current secondary path.

[0141] Optionally, in the second noise reduction operation, since the speaker has already started playing the noise-reduced signal calculated in the first noise reduction operation (i.e., noise reduction has begun), the noise perceived by the human ear changes. Therefore, the first error signal collected by the error sensor differs from the second error signal in the first noise reduction operation. The first reference signal collected by the reference sensor may be the same as or different from the second reference signal in the first noise reduction operation, depending on whether the road noise has changed. If the road noise changes significantly, the first reference signal and the second reference signal will differ considerably; if the road noise changes little, the first reference signal and the second reference signal will be relatively close.

[0142] In addition, the noise control device can also acquire the first state information of the current secondary path. This first state information can be understood as any type of state information capable of characterizing the current secondary path scene. Optionally, in one example, the first state information can be obtained using acoustic elements in the noise control system, such as by determining the correlation between the detection signal played by the speaker and the error signal collected by the error sensor. This correlation can characterize whether the speaker and / or error sensor is obstructed in the current scene. See Implementation Scheme 1 below for a related implementation method. Alternatively, in another example, it can also be obtained using non-acoustic elements in the noise control system, such as by determining the image information collected by a visual sensor. This image information contains the current image of the area where the speaker and error sensor are located, which can intuitively show whether the speaker and / or error sensor is obstructed in the current scene. See Implementation Scheme 2 below for a related implementation method.

[0143] Understandably, in other examples, the first state information can also be determined based on components other than the noise control system. For example, it can be determined by asking the user "Is there any speaker or microphone obstruction?" and waiting for the user's response, or by displaying "Is there any speaker or microphone obstruction?" on the interface and waiting for the user's input, etc. No specific limitation is made here.

[0144] Step 403: By matching the first state information with the state information of multiple pre-stored secondary paths, determine whether there is a pre-stored secondary path that matches the first state information among the multiple pre-stored secondary paths. If yes, proceed to step 404; otherwise, proceed to step 405.

[0145] Optionally, multiple pre-stored secondary paths are obtained through test modeling before the vehicle leaves the factory. These multiple pre-stored secondary paths include the default secondary paths described above, as well as at least one other secondary path, where each other secondary path can be understood as a secondary path under a non-ideal driving environment.

[0146] For example, when both the speaker and the error sensor are located in the headrest area above the seat back, other secondary paths may include, but are not limited to: as shown in Figure 7(B), the user's head obscures the secondary path of the speaker on the left side of the headrest area; as shown in Figure 7(C), the user's head obscures the secondary path of the speaker on the right side of the headrest area; as shown in Figure 7(D), the user's head obscures the secondary path of the error sensor on the left side of the headrest area; as shown in Figure 7(E), the user's head obscures the secondary path of the error sensor on the right side of the headrest area; and as shown in Figure 7(F), the user places a soft headrest in front of the headrest area, which simultaneously obscures the secondary paths of the error sensors and speakers on both sides of the headrest area. For example, when the speaker is located in the headrest area above the seat back and the error sensor is located in the floating headrest, other secondary paths may include, but are not limited to: as shown in Figure 7 (H), a secondary path of an error sensor on the left side of the floating headrest in the user's head-obstruction diagram; and as shown in Figure 7 (I), a secondary path of an error sensor on the right side of the floating headrest in the user's head-obstruction diagram.

[0147] Understandably, other secondary paths also exist, such as the secondary path where the user's head blocks the two error sensors on the left side of the floating headrest, the secondary path where the user's head blocks the two error sensors on the right side of the floating headrest, the secondary path where the user's head directly blocks the speaker and error microphone on the left side of the headrest area when there is no soft headrest or floating headrest, the secondary path where the user's head directly blocks the speaker and error microphone on the right side of the headrest area, the secondary path where all error sensors on one side are blocked, the secondary path where all speakers on one side are blocked, the secondary path where all error sensors are blocked, the secondary path where all speakers are blocked, the secondary path where both all error sensors and all speakers are blocked, and so on, which will not be listed here.

[0148] It should be noted that "left" and "right" in this application refer to the left and right in the figure, not to the left and right in the actual scene. The left in the figure can be understood as the right in the actual scene, and the right in the figure can be understood as the left in the actual scene.

[0149] Optionally, multiple pre-stored secondary paths can be obtained through test modeling. Specifically, this can involve obtaining relevant information about multiple pre-stored secondary paths through test modeling. This relevant information may include path parameters and state information, or other information. For each pre-stored secondary path, its path parameters and state information have a one-to-one correspondence, with each state information corresponding to a set of path parameters. For example, for the six secondary path scenarios shown in Figures 7(A) to (F), there will be six sets of pre-stored secondary paths: a default secondary path and five sets of non-default secondary paths. The default secondary path includes state information that can be used to characterize the state where neither the speaker nor the error sensor is obstructed, and a set of path parameters that characterize the secondary path corresponding to the scenario where neither the speaker nor the error sensor is obstructed. Each non-default secondary path includes state information that can be used to characterize the state where the speaker and / or error sensor is obstructed in the corresponding secondary path scenario, and a set of path parameters that characterize the secondary path corresponding to the non-default secondary path scenario.

[0150] As an example, the path parameters of each pre-stored secondary path can be a transfer function. The specific test modeling method can be as follows: First, park the vehicle in a quiet environment, such as a anechoic chamber; second, configure the in-vehicle speakers and error sensors in the scenario corresponding to each pre-stored secondary path, such as a scenario where some or all error sensors are blocked, and / or a scenario where some or all speakers are blocked; third, control the in-vehicle speakers to play test signals and acquire error signals collected by the error sensors; fourth, calculate the transfer function based on the test signal and the error signal, and determine the transfer function as the path parameter for each pre-stored secondary path.

[0151] Unlike the path parameters of pre-stored secondary paths, state information can be represented in various ways, and different representation methods lead to different test modeling approaches. For example, if the state information is determined based on the test signal played by the speaker and the error signal collected by the error sensor, the test modeling approach could be to analyze the correlation between the test signal played by the speaker and the error signal collected by the error sensor in each pre-stored secondary path scenario, and extract the features in the correlation as the state information for each pre-stored secondary path. See Implementation Scheme 1 below for a relevant implementation method. Alternatively, if the state information is determined based on image information collected by a vision sensor, the test modeling approach could be to identify the image information collected by the vision sensor in each pre-stored secondary path scenario, obtaining image features that characterize each pre-stored secondary path scenario as the state information for each pre-stored secondary path. See Implementation Scheme 2 below for a relevant implementation method. Other state information can be deduced by analogy, and will not be listed here.

[0152] Optionally, the relevant information for multiple pre-stored secondary paths can be pre-stored in the vehicle, such as locally in the noise control device or in relevant memory. After obtaining the first state information of the current secondary path, the noise control device can retrieve the state information of multiple pre-stored secondary paths from its local storage or relevant memory, and then compare the first state information with the state information of each of the multiple pre-stored secondary paths. If the state information of a pre-stored secondary path is the same as or similar to the first state information, for example, the deviation between the two state information is less than or equal to the allowable deviation fluctuation range, then this pre-stored secondary path can be used as the pre-stored secondary path matching the current secondary path. Conversely, if the state information of all pre-stored secondary paths differs significantly from the first state information, for example, the deviation between the state information of each pre-stored secondary path and the first state information is greater than the allowable deviation fluctuation range, then it is determined that there is no matching pre-stored secondary path, and the current secondary path belongs to a scene that has not been previously modeled.

[0153] It should be emphasized that the above description only uses the example of multiple pre-stored secondary paths including a default secondary path. However, in another example, multiple pre-stored secondary paths may not include a default secondary path. In this case, there may be only one pre-stored secondary path. However, when matching the first state information, the first state information and multiple pre-stored secondary paths as well as the default secondary path need to be matched. If the default secondary path is matched, step 404 is also executed, except that the pre-stored secondary path in step 404 is replaced with the default secondary path. This application will not repeat the description of this.

[0154] Step 404: Determine the filter coefficients based on the path parameters of the matched pre-stored secondary path, the first reference signal, and the first error signal, and adjust the filter coefficients using the first step length adjustment.

[0155] Optionally, if a matching pre-stored secondary path exists, it indicates that the current secondary path is highly likely to belong to a previously modeled pre-stored secondary path. The path parameters of this pre-stored secondary path are the applicable path parameters for the current secondary path, and these path parameters have a good noise reduction effect on the current secondary path. Based on this, the noise control device can call the path parameters of the matching pre-stored secondary path to determine the filter coefficients for the second noise reduction operation. For example, an adaptive filtering algorithm can be used to process the path parameters of the matching pre-stored secondary path, the first reference signal, and the first error signal to obtain the filter coefficients for the second noise reduction operation. Furthermore, the filter coefficients for the second noise reduction operation can be adjusted using the first step length. For example, the first step length can be multiplied by the calculated filter coefficients to obtain the adjusted filter coefficients. The first step length is also called the initial step size, and its value is relatively large, such as 0.8. Of course, other algorithms can also use numbers greater than 1, without limitation.

[0156] In the above discussion, using pre-modeled matching path parameters to determine filter coefficients ensures the accuracy of filter coefficient calculation and maintains good noise reduction performance. Furthermore, while ensuring noise reduction accuracy based on matched path parameters, using a large step size to adjust the filter coefficients ensures that the adjusted coefficients still maintain large values. Larger filter coefficients contribute to faster convergence speed and greater noise reduction depth. Therefore, combining matched path parameters and a large step size can ensure noise reduction accuracy while maintaining a relatively fast noise reduction speed in most scenarios, where "most scenarios" refers to pre-modeled secondary path scenarios.

[0157] Step 405: Determine the filter coefficients based on the path parameters of the default secondary path, the first reference signal, and the first error signal, and adjust the filter coefficients using a second step size, which is smaller than the first step size.

[0158] Optionally, if no matching pre-stored secondary path exists, it means that the current secondary path likely belongs to a scene not previously modeled, and therefore no matching path parameters exist. In this case, to implement the second noise reduction operation, the path parameters of the default secondary path can be used to calculate the filter coefficients. For example, an adaptive filtering algorithm can be used to process the path parameters of the default secondary path, the first reference signal, and the first error signal to obtain the filter coefficients for the second noise reduction operation. Simultaneously, to reduce the impact of the default path parameters on the inaccuracy of the current secondary path noise reduction, a small step size can be used to adjust the filter coefficients. For example, the second step size can be multiplied by the calculated filter coefficients to obtain the adjusted filter coefficients. The second step size is less than the first step size, for example, it can be half of the first step size. For instance, when the first step size is 0.8, the second step size can be 0.4.

[0159] In the above content, by using default path parameters to determine the filter coefficients when no pre-stored secondary path is matched, it is ensured that the filter coefficients can be calculated within the current cycle, thus ensuring that the noise reduction operation is continuous and will not be interrupted in any cycle. However, when the default path parameters cannot guarantee the accuracy of noise reduction, a small step size is used to adjust the filter coefficients, resulting in smaller values. Smaller filter coefficients reduce the probability of divergence or abnormal sounds caused by inaccurate path parameter selection, ensuring the noise reduction stability of the noise control system. Based on this, combining the default path parameters and the second step size, it is possible to ensure the continuity of noise reduction while maintaining its stability in a small number of scenarios, referring to those secondary paths that could not be pre-modeled.

[0160] Step 406: Update the noise reduction signal using the adjusted filter coefficients.

[0161] In one example, after calculating the adjusted filter coefficients, the noise control device can directly use the adjusted filter coefficients to update the noise reduction signal. For example, it can convolve the adjusted filter coefficients onto the first reference signal as the output signal and send the output signal as the updated noise reduction signal to the speaker so that the speaker can emit an anti-phase sound wave to reduce noise at the target point based on the noise reduction signal.

[0162] Alternatively, in another example, to ensure the accuracy of the noise reduction signal update, the noise control device can perform divergence detection based on the adjusted filter coefficients before updating the noise reduction signal. If divergence is detected when using the adjusted filter coefficients for noise reduction, the step size used in the previous adjustment of the filter coefficients is reduced, and the filter coefficients are readjusted using the reduced step size. Divergence detection is then repeated using the readjusted filter coefficients, and this process is continued until it is determined that the adjusted filter coefficients no longer have a divergence trend. At this point, the adjusted filter coefficients are used to update the noise reduction signal. Alternatively, the noise reduction operation can be partially or completely terminated when a set condition is met at a certain adjustment.

[0163] Optionally, divergence in divergence detection can be understood as amplitude divergence, such as the amplitude divergence of the signal to be output. Amplitude can also be characterized by signal strength, so it can also be understood as the intensity divergence of the signal to be output. When the intensity of the signal to be output is too high, it may not only fail to reduce noise but may even amplify it, causing the system to exhibit divergence or abnormal sounds. Therefore, a preset intensity can be configured. If the intensity of the signal to be output is greater than the preset intensity, it is considered to be diverging; otherwise, it is considered not to be diverging.

[0164] Based on this, after calculating the adjusted filter coefficients, the noise control device can first convolve the first reference signal with the adjusted filter coefficients to obtain the output signal. Then, it compares the strength of the output signal with the set strength. If the strength of the output signal is less than or equal to the set strength, it indicates that the output signal does not have a divergence trend. Therefore, the output signal can be used to update the noise reduction signal, that is, the output signal is used as the updated noise reduction signal and output to the speaker for playback. Conversely, if the strength of the output signal is greater than the set strength, it indicates that the output signal has a divergence trend. In this case, the noise reduction signal is not updated. Instead, the step size used to adjust the filter coefficients is reduced, for example, to half of the previously used step size. The reduced step size is then used to readjust the filter coefficients. The above divergence detection operation is repeated using the readjusted filter coefficients until it is determined that there is no divergence. Then, the latest output signal is used to update the noise reduction signal, or the noise reduction operation is partially or completely exited if the set conditions are met.

[0165] Optionally, the setting conditions can be configured according to actual needs. For example, it can be one or more of the following: the number of times the step size is reduced reaches a set number; the reduced step size is less than or equal to the set step size; the time for reducing the step size reaches a set time length; the adjusted filter coefficient is less than or equal to the set filter coefficient, or other conditions. The parameters such as the set number of times, set step size, set time length, and set filter coefficient can all be determined based on actual vehicle testing. For example, taking the set step size as an example, during actual vehicle testing, by analyzing the noise reduction effect corresponding to different step sizes, the critical step size that allows the noise reduction effect to just meet expectations can be found. This critical step size is the minimum threshold for the allowed step size of the filter coefficient adjustment, i.e., the set step size. Based on this set step size, when the noise reduction effect just begins to fail to meet expectations, some or all noise reduction operations can be decisively turned off, saving unnecessary computational overhead.

[0166] For ease of understanding, let's assume the first step size is 0.8 and the second step size is 0.4, with the condition that the reduced step size is less than or equal to 10. -5 Here are two specific examples to illustrate this:

[0167] Example 1: Assuming the first state information of the current secondary path matches the state information of a pre-stored secondary path, the filter coefficient W is first calculated based on the path parameters of the pre-stored secondary path, the first reference signal, and the first error signal. The filter coefficient W is then adjusted using a step size of 0.8, resulting in a filter coefficient of 0.8W. The adjusted filter coefficient 0.8W and the first reference signal X are convolved to obtain the output signal 0.8W*X (where * is the convolution symbol). It is then determined whether the signal strength of the output signal 0.8W*X exceeds a set strength. If it does not exceed the set strength, it is determined to have no divergence trend, and the output signal is sent to the speaker as the updated noise-reduced signal. If it exceeds the set strength, it is determined to have a divergence trend, and the noise-reduced signal is not updated; instead, the step size is reduced to the previously used step size of 0. Half of 8, i.e., 0.4, is used to readjust the filter coefficient W with a reduced step size of 0.4, resulting in a filter coefficient of 0.4W. The adjusted filter coefficient 0.4W is then convolved with the first reference signal X to obtain the output signal 0.4W*X. It is then determined whether the signal strength of the output signal 0.4W*X exceeds a set strength. If it does not exceed the set strength, it is considered to have no divergence trend, and the output signal is sent to the speaker as the updated noise-reduced signal. If it exceeds the set strength, it is considered to have a divergence trend, and the noise-reduced signal is not updated. Instead, the step size is reduced to half of the previously used step size of 0.4, i.e., 0.2, and the filter coefficient W is adjusted with a reduced step size of 0.2, resulting in a filter coefficient of 0.2W. This process is repeated until the reduced step size is less than or equal to 10. -5If the signal strength of the signal to be output still exceeds the set strength, the noise reduction operation will be partially or completely exited. For example, some or all speakers will be turned off so that some or all speakers will no longer play the noise reduction signal, or the signal played by some or all speakers will be set to zero.

[0168] Example 2: Assuming that none of the pre-stored secondary paths has state information matching the first state information of the current secondary path, the default secondary path parameters are obtained from the path parameters of the multiple pre-stored secondary paths. Using these path parameters, the first reference signal, and the first error signal, the filter coefficient W is calculated, and the filter coefficient W is adjusted using a second step size of 0.4, resulting in an adjusted filter coefficient of 0.4W. The adjusted filter coefficient 0.4W and the first reference signal X are convolved to obtain the output signal 0.4W*X. It is then determined whether the signal strength of the output signal 0.4W*X exceeds a set strength. If it does not exceed the set strength, it is determined to have no divergence trend, and the output signal is sent to the speaker as the updated noise-reduced signal for playback. If it exceeds the set strength, it is determined to have a divergence trend, and the noise-reduced signal is not updated; instead, the step size is adjusted. The step size is reduced to half of the previously used step size of 0.4, i.e., 0.2. The filter coefficient W is adjusted using the reduced step size of 0.2, resulting in a filter coefficient of 0.2W. The adjusted filter coefficient 0.2W and the first reference signal X are convolved to obtain the output signal 0.2W*X. It is then determined whether the signal strength of the output signal 0.4W*X exceeds the set strength. If it does not exceed the set strength, it is determined that there is no divergence trend, and the output signal is sent to the speaker as the updated noise-reduced signal for playback. If it exceeds the set strength, it is determined that there is a divergence trend, and the noise-reduced signal is not updated. Instead, the step size is reduced to half of the previously used step size of 0.2, i.e., 0.1, and the filter coefficient W is adjusted using the reduced step size of 0.1, resulting in a filter coefficient of 0.1W. This process is repeated until the reduced step size is less than or equal to 10. -5 If the strength of the adjusted output signal still exceeds the set strength, the noise reduction operation will be partially or completely exited. For example, some or all speakers will be turned off so that some or all speakers will no longer play the noise reduction signal, or the speakers will not be turned off but the signal played by some or all speakers will be set to zero, which is equivalent to playing an empty noise reduction signal.

[0169] It should be noted that the above example only provides one possible way to adjust the filter coefficients based on the step size. Different filtering algorithms may have different step size adjustment methods. For example, in some other examples, the step size may be a number greater than 1. When adjusting the filter coefficients according to this step size, it is no longer a simple multiplication, but some other addition, subtraction, multiplication, division operations, or other formulas. The above content is only an example given to facilitate understanding of the scheme. This application does not limit the specific step size adjustment method.

[0170] Based on the above noise control method, by using a larger step size to adjust the filter coefficients when matching the secondary path state with the state of a pre-stored secondary path, a relatively fast convergence speed and a large noise reduction depth can be guaranteed in most scenarios with pre-modeling. Conversely, by using a smaller step size to adjust the filter coefficients when no matching pre-stored secondary path exists, noise reduction stability can be guaranteed in a small number of scenarios without pre-modeling, avoiding abnormal noise and divergence. Furthermore, by continuously detecting divergence trends during the noise control process, the adjustment step size can be reduced in a timely manner when divergence is detected, or even the noise control system can be shut down to prevent excessive noise reduction and the generation of obvious abnormal noise.

[0171] Based on the above noise control methods, the noise control process corresponding to different types of state information will be described in detail below, based on Implementation Scheme 1 and Implementation Scheme 2.

[0172] Implementation Plan 1

[0173] Please refer to Figure 8, which shows a flowchart of a noise control method provided in Implementation Scheme 1. This method is applicable to noise control devices, such as the noise control device shown in Figure 2. In Implementation Scheme 1, the state information of the secondary path is calculated based on the signals of acoustic elements in the noise control system. These acoustic elements include, but are not limited to, loudspeakers and error sensors. As shown in Figure 8, the method includes the following steps:

[0174] Step 801: Before the vehicle leaves the factory, based on the test signal played by the speaker and the error signal collected by the error sensor, determine the status information and path parameters of multiple pre-stored secondary paths, and establish the correspondence between the status information and path parameters of multiple secondary paths.

[0175] For each pre-stored secondary path, due to the distance between the speaker and the error sensor, the test signal emitted by the speaker will exhibit certain amplitude and phase responses during propagation. These responses can be characterized using one or more of the transfer function, cross-spectral function, and cross-correlation coefficient. Based on this, the state information of the pre-stored secondary path can be configured to include one or more of the transfer function, cross-spectral function, and cross-correlation coefficient. The transfer function refers to the function related to the transmission characteristics during the transmission of the test signal emitted by the speaker to the error sensor. Assuming the transfer function of the secondary path is h(t), the test signal output by the speaker is f(t), and the error signal received by the error sensor is g(t), then: g(t) = f(t) * h(t), where * represents convolution. Besides the transfer function, the cross-spectral function and cross-correlation coefficient can also be used to characterize the correlation between the test signal emitted by the speaker and the error signal received by the error sensor. The relevant calculation methods can be found in existing technologies, and this application will not describe them in detail.

[0176] Based on this, as an example, the state information and path parameters of each pre-stored secondary path are obtained through testing as follows: First, the vehicle is parked in a quiet environment, such as a soundproof room; second, the in-vehicle speakers and error sensors are configured in the scenarios corresponding to each pre-stored secondary path, as shown in Figures 7(A) to (I); third, the in-vehicle speakers are controlled to play test signals, and the error signals collected by the error sensors are acquired; fourth, the state information and path parameters of each pre-stored secondary path are determined based on the test signals and error signals. For example, the transfer function can be calculated based on the test signals and error signals, and the transfer function can be used as both the state information and the path parameters. Alternatively, the transfer function can be calculated based on the test signals and error signals, and the cross-spectral function and / or cross-correlation coefficient can also be calculated, with only the transfer function used as the path parameter, and the cross-spectral function and / or cross-correlation coefficient used as the state information. Alternatively, the transfer function can be used as the path parameter, and the transfer function plus the cross-spectral function and / or cross-correlation coefficient can be used as the state information.

[0177] Step 802: After the vehicle activates noise reduction, control the speaker to play a noise reduction signal.

[0178] For example, after a vehicle leaves the factory, if the noise control device receives a noise reduction command, it acquires the vehicle's current state and determines whether the current state meets the preset noise reduction conditions. If not, the noise reduction operation is not initiated. Conversely, if it does meet the conditions, it acquires the second reference signal picked up by the reference sensor and the second error signal picked up by the error sensor. Simultaneously, it finds the default secondary path from multiple pre-stored secondary paths stored locally or in relevant memory and acquires the path parameters of the default secondary path. Based on the second reference signal, the second error signal, and the path parameters of the default secondary path, it calculates the initial filter coefficients and adaptively updates and iterates the initial filter coefficients using the first step length. Then, it convolves the initial filter coefficients obtained from the reference signal to obtain the noise reduction signal. This noise reduction signal is input to the speaker to control the speaker to emit an anti-phase sound wave of the ambient noise to reduce noise at the target point.

[0179] Step 803: Control the speaker to play the detection signal and acquire the first error signal collected by the error sensor. Determine the first state information of the current secondary path based on the detection signal and the first error signal.

[0180] Optionally, while the speaker is playing the noise-reduced signal, the noise control device simultaneously starts generating a detection signal and outputs it to the speaker. The detection signal can be a randomly generated signal or a signal generated according to a set rule. For example, it can be a randomly generated set of white noise signals, a signal obtained by processing a randomly generated white noise signal, or a manually set fixed noise signal; there are no limitations.

[0181] In one example, to reduce the probability that the detection signal is perceived by the human ear, the spectrum of the detection signal can be designed with reference to the signals that the human ear can normally hear, thus masking the detection signal. Since the error sensor is located at the ear, the detection signal is designed with reference to the error signal collected by the error sensor. Based on the masking effect of the human ear, the detection signal can be designed to have a spectrum shape similar to the error signal but with a lower signal strength.

[0182] Based on this, in one example, the detection signal can be generated by designing the steps in steps one through three as follows:

[0183] Step 1: Generate an initial signal randomly or according to a set rule.

[0184] For example, a noise control device can randomly generate a set of white noise signals and use that set of white noise signals as the initial signal.

[0185] Step 2: Obtain the third error signal collected by the error sensor.

[0186] Understandably, in step one, the noise control device only generates a white noise signal, which has not yet been sent to the speaker. Therefore, the speaker currently only plays the noise-reduced signal. The third error signal is the signal after the actual noise signal and the noise-reduced signal are superimposed at the human ear.

[0187] Step 3: Based on the third error signal, the initial signal is subjected to spectrum shaping and gain control to obtain the detection signal.

[0188] Optionally, spectrum shaping involves making the spectral shape of the detected signal approximate the spectral shape of the third error signal. For example, the deviation between the spectral shape of the detected signal and the spectral shape of the third error signal is kept within a certain threshold, such as within 5 dB. Gain control, on the other hand, ensures that the noise intensity generated at the error sensor after the detected signal is played through the speaker is lower than the noise intensity of the third error signal, and the difference between the noise intensity of the two signals is within a set range. For example, the noise intensity of the detected signal is 10–30 dB lower than the noise intensity of the third error signal.

[0189] Taking a spectral shape deviation within 5dB and a noise intensity low of 15-25dB as an example, assuming the spectral curve of a randomly generated set of white noise signals is the dashed line shown in Figure 9(A), and the spectral curve of the third error signal is the solid line shown in Figure 9(A), then:

[0190] During spectrum shaping, the white noise spectrum curve shown in Figure 9(A) can be superimposed on the spectrum curve of the third error signal shown in Figure 9(A), and the two spectrum curves can be discretized in frequency to obtain multiple discrete points. Each set of discrete points is compared point by point: if the noise intensity difference between the white noise discrete point and the corresponding discrete point of the third error signal is greater than 5dB, the noise intensity value of the white noise discrete point is adjusted to bring it closer to the corresponding discrete point of the third error signal, ensuring that the noise intensity difference between the two is less than or equal to 5dB. For white noise discrete points with a noise intensity difference less than or equal to 5dB, no adjustment is made. In this way, after adjusting all white noise discrete points with a noise intensity difference greater than 5dB, the noise intensity difference between all white noise discrete points and the corresponding discrete points of the third error signal is within 5dB. The spectrum shape of the white noise signal approaches the spectrum shape of the third error signal, as shown in Figure 9(B). The white noise signal can be masked by the third error signal, ensuring that the human ear cannot subjectively perceive it, or hardly perceives it, based on the masking effect of the human ear.

[0191] When performing gain control, the spectrum-shaped white noise signal can be shifted as a whole towards a lower noise intensity direction, for example, shifted downwards by 20dB, as shown in Figure 9(C). This ensures that the noise intensity of all discrete points of white noise after the shift is lower than the noise intensity of the corresponding discrete point of the third error signal. Furthermore, since the noise intensity values ​​of the spectrum-shaped white noise discrete points are at most 5dB higher and at least 5dB lower than the noise intensity of the corresponding discrete point of the third error signal, after shifting downwards by 20dB, the noise intensity difference between all discrete points of white noise and the corresponding discrete point of the third error signal is within the range of 15dB to 25dB. Through gain control, it can be ensured that the noise intensity of the adjusted white noise signal is not too high to affect human hearing, nor too low to result in an excessively low signal-to-noise ratio. In other words, the detected signal-to-noise ratio can be increased as much as possible while ensuring that it is imperceptible to the human ear, which can be used to more accurately determine the secondary path state.

[0192] Optionally, after designing a detection signal similar to the third error signal collected by the error sensor, the noise control device can send the detection signal and the noise reduction signal to the speaker for playback separately, or it can first superimpose the detection signal and the noise reduction signal, and then send the superimposed signal to the speaker for playback; there is no limitation. It should be noted that the detection signal can be played for only a period of time, or it can be played continuously throughout the noise reduction process; there is no specific limitation.

[0193] Furthermore, after controlling the loudspeaker to play the detection signal or the superimposed signal, the noise control device acquires the first error signal collected by the error sensor. Since the loudspeaker is playing both the noise-reduced signal and the detection signal at this time, the first error signal is actually a signal obtained by superimposing the noise-reduced signal, the detection signal, and the actual noise signal at the human ear. The noise control device calculates the first state information of the current secondary path based on the detection signal and the first error signal. The first state information can be one or more of a transfer function, a cross-spectral function, and a cross-correlation coefficient. When the first state information is a transfer function, the deconvolution of the first error signal and the detection signal can be used as the first state information. When the first state information is a cross-spectral function, the time-domain characteristics of the correlation between the detection signal and the first error signal can be used as the first state information. When the first state information is a cross-correlation coefficient, the frequency-domain characteristics of the correlation between the detection signal and the first error signal can be used as the first state information. When the first state information is multiple of a transfer function, a cross-spectral function, and a cross-correlation coefficient, multiple of the following can be used as the first state information: the deconvolution of the first error signal and the detection signal, the time-domain characteristics of the correlation between the detection signal and the fourth error signal, and the frequency-domain characteristics.

[0194] Step 804: Match the first state information with the state information of multiple pre-stored secondary paths to determine if there is a matching pre-stored secondary path. If yes, proceed to step 805; otherwise, proceed to step 806.

[0195] Optionally, for each of the multiple pre-stored secondary paths, the noise control device can calculate the deviation between the first state information and the state information of that pre-stored secondary path, and then determine whether the deviation is within the allowable deviation fluctuation range. If so, it is determined that the pre-stored secondary path matches the first state information; otherwise, it is determined that the pre-stored secondary path does not match the first state information. The allowable deviation fluctuation range can be obtained from actual vehicle testing or configured by those skilled in the art based on experience. For example, in one example, it can be set to a symmetrical range, such as ±3dB, to accommodate fluctuations in both increasing and decreasing directions.

[0196] Furthermore, after analyzing all the pre-stored secondary paths, if all pre-stored secondary paths do not match the first state information, then it is determined that no matching pre-stored secondary path exists. If only one pre-stored secondary path matches the first state information, then that pre-stored secondary path is determined to be the matching pre-stored secondary path. If multiple pre-stored secondary paths match the first state information, one of them can be selected as the matching pre-stored secondary path. For example, the one with the smallest deviation can be selected, or the one with the middle deviation value can be selected, or a random one can be selected, etc., without limitation.

[0197] It should be noted that when the state information is a transfer function, cross-spectral function, or cross-correlation coefficient, this state information is not a single piece of information, but a set of information containing multiple sub-information pieces. In this case, the deviation between two pieces of state information can specifically be the average or weighted average of the deviations between the multiple sub-information pieces contained in the two pieces of state information. When the state information includes multiple transfer functions, cross-spectral functions, and cross-correlation coefficients, this state information is multiple sets of information, each set containing multiple pieces of information. In this case, the deviation between two pieces of state information can be the average or weighted average of the deviations of the multiple sets of information, and the deviation of each set of information is the average or weighted average of the deviations between the multiple sub-information pieces contained in that set of information.

[0198] Based on the above implementation, by using the cross-spectral function, transfer function, cross-correlation coefficient, and other combinations of the detection signal played by the speaker and the error signal collected by the error sensor to determine the current secondary path state, it is possible to accurately determine whether the current situation is in some extreme usage scenarios, such as soft headrest obstruction or error sensor blockage.

[0199] Step 805: Determine the filter coefficients based on the path parameters of the matched pre-stored secondary path, the first reference signal, and the first error signal, and adjust the filter coefficients using the first step length adjustment.

[0200] For example, a first reference signal acquired by a reference sensor and a first error signal acquired by an error sensor can be obtained. Then, a preset algorithm (such as an adaptive filtering algorithm) is used to process the first reference signal, the first error signal, and the path parameters of the matched pre-stored secondary path to determine the filter coefficients. The filter coefficients are then adjusted using a first-step size, for example, by multiplying the first-step size by the calculated filter coefficients. By calling the matched pre-stored secondary path and using an initial update step size parameter (i.e., a larger step size) to control the filter update, the best noise reduction performance and the fastest convergence speed can be maintained.

[0201] Step 806: Determine the filter coefficients based on the path parameters of the default secondary path, the first reference signal, and the first error signal, and adjust the filter coefficients using a second step size, which is smaller than the first step size.

[0202] For example, a first reference signal acquired by a reference sensor and a first error signal acquired by an error sensor can be obtained. Then, a preset algorithm (such as an adaptive filtering algorithm) is used to process the first reference signal, the first error signal, and the path parameters of the default secondary path to determine the filter coefficients. The filter coefficients are then adjusted using a second step size, for example, by multiplying the calculated filter coefficients by the second step size. In one example, the second step size could be half the first step size. Alternatively, it could be 1 / 3, 3 / 4, 1 / 4, etc., without limitation. By calling the default pre-stored secondary path and using a smaller step size to control the filter update, the stability of noise reduction can be ensured, avoiding abnormal sounds and divergence.

[0203] Step 807: Determine whether the adjusted filter coefficients have a divergence trend. If yes, proceed to step 808; otherwise, proceed to step 809.

[0204] For example, the adjusted filter coefficients and the first reference signal can be convolved to obtain the output signal. Then, it can be determined whether the signal strength of the output signal is greater than the set strength. If it is, it is determined that there is a divergence trend. If it is not greater, it is determined that there is no divergence trend.

[0205] Step 808: Reduce the step size and adjust the filter coefficients according to the reduced step size, then proceed to step 807.

[0206] Optionally, after each reduction in step size, it can be determined whether the reduced step size is less than the set minimum threshold (i.e., the set step size). If so, the noise reduction operation is partially or completely exited. If not, the filter coefficients are adjusted according to the reduced step size. For example, the reduced step size is multiplied by the previously calculated filter coefficients to obtain the adjusted filter coefficients. Then, it is determined whether the adjusted filter coefficients have a divergence trend. If so, the step size is reduced again, and the above process is repeated until it is determined that the adjusted filter coefficients do not have a divergence trend. Then, step 809 is executed, or the noise reduction operation is partially or completely exited when the step size is lower than the set minimum threshold.

[0207] Step 809: Update the noise reduction signal based on the adjusted filter coefficients.

[0208] For example, the adjusted filter coefficients and the first reference signal can be convolved to obtain the output signal, and the output signal can be sent to the speaker as the updated noise reduction signal. Then, wait to receive the new reference signal and error signal to start a new round of noise reduction operation.

[0209] To facilitate understanding, a specific example will be used below to illustrate the complete implementation process of Implementation Plan 1.

[0210] Assume the noise control system is an in-vehicle active noise cancellation system using headrest speakers and headrest microphones. Please refer to Figures 10a and 10b, which show possible flowcharts of the noise control method in this example. Figure 10a shows the execution flowchart, and Figure 10b shows the principle block diagram. Combining Figures 10a and 10b, the overall process includes the following steps one through six:

[0211] Step one: Before the car leaves the factory, several sets of secondary path state information and path parameters are pre-stored through test modeling, with each set of state information and path parameters having a one-to-one correspondence. For example, four sets of state information and path parameters are pre-stored: state information and path parameters for the standard state, state information and path parameters for the soft headrest covering the seat headrest, state information and path parameters for the head covering the left side of the seat headrest, and state information and path parameters for the head covering the right side of the seat headrest. These four sets of secondary path state information correspond one-to-one with the path parameters modeled under the four secondary path states.

[0212] Step two: After active noise cancellation is activated, the control filter coefficients W are calculated using the reference signal picked up by the vehicle's accelerometer, the error signal picked up by the microphone, and the pre-stored path parameters G of the secondary path (i.e., the path parameters in the standard state). The noise-canceling signal is then output to the headrest speaker. For example, the signal obtained by convolving the calculated filter coefficients W with the reference signal is output to the headrest speaker.

[0213] Optionally, the active noise cancellation operation can be activated when the noise cancellation command is received, the vehicle speed is between 10km / h and 150km / h, and all doors, windows and sunroof (if any) are closed.

[0214] Step 3: Simultaneously, a secondary path detection signal is generated. This detection signal is then spectrally shaped and gain controlled before being output to the headrest speaker, making the signal played by the headrest speaker a superposition of the noise reduction signal and the detection signal.

[0215] Specifically, the spectrum shaping method can be: acquiring the signal picked up by the microphone (referred to as the microphone signal), designing a filter with a similar spectrum shape based on the characteristics of the microphone signal, and making it almost imperceptible to the human ear due to the masking effect, such that the deviation between the spectrum shape of the detected signal and the noise spectrum shape picked up by the microphone is within 5dB.

[0216] Specifically, the gain control method can be: based on the signal amplitude collected by the microphone, make the amplitude of the detection signal emitted by the speaker 10 to 30 dB lower than the amplitude of the original noise signal at the microphone, and ensure that the human ear cannot subjectively perceive it.

[0217] Step 4: Use the detection signal and microphone signal to estimate the current secondary path state H, and determine whether the current secondary path state H is within the allowable fluctuation range of a certain set of pre-stored state information. The threshold of this fluctuation range can be preset to ±3dB.

[0218] For example, if the current secondary path state H is a transfer function, the transfer function is calculated based on the detection signal and the signal picked up by the microphone, and the deviation value between the transfer function and each set of pre-stored state information is calculated to determine whether there is a set of pre-stored state information whose deviation value is within the allowable fluctuation range.

[0219] Step 5: If the current secondary path state H is within the allowable fluctuation range of a set of pre-stored state information, then the corresponding pre-stored path parameters are called and the control filter coefficients W are updated using an initial step size (i.e., a larger step size) to maintain the best noise reduction performance and the fastest convergence speed. For example, if the current secondary path state H is detected to be within the allowable deviation range of the standard state's state information, then the path parameters corresponding to the standard state and a larger update step size (e.g., 0.8) are called to update the control filter coefficients, while simultaneously performing divergence detection. If no divergence trend is detected, the signal obtained by convolving the updated filter coefficients with the reference signal is used as the new noise reduction signal and sent to the speaker. If a divergence trend is detected, the update step size is reduced and divergence detection is repeated. For example, the step size reduction strategy is to reduce it to 1 / 2 of the original size each time, i.e., reduce the step size to 0.4 and repeat step 5 until the step size is reduced to less than a specified threshold (e.g., 10). -5 After that, turn off active noise cancellation.

[0220] Step six: If the current secondary path state H is outside the allowable fluctuation range of any set of pre-stored state information, reduce the update step size. For example, the step size reduction strategy is to reduce it to half of the original size each time, and use the reduced step size (i.e., the smaller step size) to update the control filter coefficients W. For instance, if the current secondary path state H is outside the allowable fluctuation range of any set of pre-stored state information, reduce the step size from 0.8 to 0.4 before calculating the control filter coefficients W, and simultaneously perform divergence detection. If no divergence trend is detected, the signal obtained by convolving the currently updated filter coefficients with the reference signal is used as the new noise-reduced signal and sent to the speaker. If a divergence trend is detected, continue to reduce the step size to 0.2 and repeat the divergence detection until the step size is reduced to less than a certain specified threshold (e.g., 1e-5), then turn off the active noise reduction operation.

[0221] Optionally, in step five or six, the specific method for divergence detection can be: calculating the next frame output signal based on the calculated control filter coefficients W that need to be updated and the acquired reference signal x. The calculation method is: next frame output signal y = W * x, where * represents convolution operation. It is then determined whether the energy (e.g., signal strength) of the next frame signal exceeds a specified threshold. If it does, it is determined that there is a divergence trend, and the control filter coefficients W are not updated. If it does not exceed the threshold, it is determined that there is no divergence trend, and the control filter coefficients are updated. This specified threshold needs to be determined through actual vehicle testing.

[0222] Based on the above implementation scheme one, the state information of the secondary path can be calculated by using a loudspeaker to play detection signals and an error sensor to collect error signals. In this way, the detection operation of the secondary path can be completed using only the acoustic components in the noise control system, without the need to introduce other components, which can reduce the structural complexity of the noise control system and reduce control costs.

[0223] Implementation Plan 2

[0224] Please refer to Figure 11, which shows a flowchart of a noise control method provided in Implementation Scheme 2. This method is applicable to noise control devices, such as the noise control device shown in Figure 2. In Implementation Scheme 2, the state information of the secondary path is calculated based on image information acquired by a visual sensor, such as an in-cabin camera. As shown in Figure 11, the method includes the following steps:

[0225] Step 1101: Before the vehicle leaves the factory, based on the cabin images collected by the cabin camera, determine the status information of multiple pre-stored secondary paths, and based on the test signals played by the speaker and the error signals collected by the error sensor, determine the path parameters of multiple pre-stored secondary paths, and establish the correspondence between the status information and path parameters of multiple secondary paths.

[0226] As an example, the state information of each pre-stored secondary path is obtained by testing as follows: First, the in-vehicle speakers and error sensors are configured in the scene corresponding to each pre-stored secondary path, as shown in (A) to (I) in Figure 7; Second, the in-vehicle camera is controlled to capture images of the cabin, and the area where the speakers and error sensors are located is identified from the cabin images, and the image features of the area are extracted; Third, the extracted image features are used as the state information of each pre-stored secondary path.

[0227] As an example, the path parameters for each pre-stored secondary path are obtained through testing as follows: First, the vehicle is parked in a quiet environment, such as a soundproof room; second, the in-vehicle speakers and error sensors are configured in the scenarios corresponding to each pre-stored secondary path, as shown in Figures 7(A) to (I); third, the in-vehicle speakers are controlled to play test signals, and the error signals collected by the error sensors are acquired; fourth, the path parameters for each pre-stored secondary path are determined based on the test signals and error signals. For example, the transfer function can be calculated based on the test signals and error signals, and the transfer function can be used as the path parameters.

[0228] Step 1102: After the vehicle activates noise reduction, control the speaker to play a noise reduction signal.

[0229] For example, after determining that noise reduction operation is enabled, the noise control device determines the noise reduction signal based on the second reference signal collected by the reference sensor, the second error signal collected by the error sensor, and the path parameters of the default secondary path, and controls the speaker to play the noise reduction signal.

[0230] Step 1103: Control the in-cabin camera to collect the current in-cabin image, and determine the first state information of the current secondary path based on the current in-cabin image.

[0231] Optionally, while the speaker is playing the noise reduction signal, the noise control device simultaneously starts controlling the cabin camera to capture the current cabin image, identifies the area image where the speaker and error sensor are located in the current cabin image, extracts the image features of the area image as the first state information of the current secondary path.

[0232] For example, the noise control device can control the cabin camera to acquire a frame of the current cabin image every 1 second, and analyze the current cabin image to determine whether there is a soft headrest blocking the view, or whether a person's head is shifted to the left or right to block the speaker or error sensor, etc. The analyzed state is used as the first state information of the current secondary path.

[0233] Step 1104: Match the first state information with the state information of multiple pre-stored secondary paths to determine if there is a matching pre-stored secondary path. If yes, proceed to step 1105; otherwise, proceed to step 1106.

[0234] Optionally, the noise control device can compare the state obtained from analyzing the current cabin image with the state of each pre-stored secondary path to determine whether a matching pre-stored secondary path exists. For example, if the image analysis result indicates that there is no soft headrest obstruction and no head shifting to the left or right obstructing the speakers or error sensors, then the matching pre-stored secondary path is determined to be the default secondary path, i.e., the secondary path corresponding to the scenario shown in Figure 7(A) or Figure 7(G). If the image analysis result indicates that a soft headrest obstructs all speakers and error sensors in the seat headrest, then the matching pre-stored secondary path is determined to be the secondary path corresponding to the scenario shown in Figure 7(I). If the image analysis result does not match the state of any of the pre-stored secondary paths, then it is determined that no matching pre-stored secondary path exists. Further examples are not listed here.

[0235] Step 1105: Obtain the first reference signal acquired by the reference sensor and the first error signal acquired by the error sensor; determine the filter coefficients based on the path parameters of the pre-stored secondary path, the first reference signal, and the first error signal; and adjust the filter coefficients using the first step length adjustment.

[0236] Step 1106: Obtain the first reference signal acquired by the reference sensor and the first error signal acquired by the error sensor. Determine the filter coefficients based on the path parameters of the default secondary path, the first reference signal, and the first error signal. Adjust the filter coefficients using a second step size, where the second step size is smaller than the first step size.

[0237] Step 1107: Determine whether the adjusted filter coefficients have a divergence trend. If yes, proceed to step 1108; otherwise, proceed to step 1109.

[0238] Step 1108: Reduce the step size and adjust the filter coefficients according to the reduced step size, then execute step 1107.

[0239] Step 1109: Update the noise reduction signal based on the adjusted filter coefficients.

[0240] It should be noted that the specific implementation of steps 1105 to 1109 above is as described in steps 805 to 809 of the above implementation scheme one, and will not be repeated here.

[0241] To facilitate understanding, a specific example will be used below to illustrate the complete implementation process of Implementation Scheme 2.

[0242] Assume the noise control system is an in-vehicle active noise cancellation system using headrest speakers and headrest microphones. Refer to Figures 12a and 12b, which show possible flowcharts of the noise control method in this example. Figure 12a shows the execution flowchart, and Figure 12b shows the principle block diagram. Combining Figures 12a and 12b, the overall process includes the following steps one through six:

[0243] Step one: Before the car leaves the factory, several sets of secondary path state information and path parameters are pre-stored through test modeling, with each set of state information and path parameters having a one-to-one correspondence. For example, four sets of state information and path parameters are pre-stored: state information and path parameters for the standard state, state information and path parameters for the soft headrest covering the seat headrest, state information and path parameters for the head covering the left side of the seat headrest, and state information and path parameters for the head covering the right side of the seat headrest. These four sets of secondary path state information correspond one-to-one with the path parameters modeled under the four secondary path states.

[0244] Step two: After active noise cancellation is activated, the control filter coefficients W are calculated using the reference signal picked up by the vehicle's accelerometer, the error signal picked up by the microphone, and the pre-stored path parameters G of the secondary path (i.e., the path parameters in the standard state). The noise-canceling signal is then output to the headrest speaker. For example, the signal obtained by convolving the calculated filter coefficients W with the reference signal is output to the headrest speaker.

[0245] Optionally, the active noise cancellation operation can be activated when the noise cancellation command is received, the vehicle speed is between 10km / h and 150km / h, and all doors, windows and sunroof (if any) are closed.

[0246] Step 3: The camera periodically acquires the current image and analyzes the current secondary path status based on the image information.

[0247] Compared to Implementation Scheme 1, Implementation Scheme 2 uses sensors other than the active noise reduction system, such as a camera, for secondary path state detection. For example, the camera acquires a frame of the current image every 1 second and analyzes the current image to determine whether there is a soft headrest obstructing the view, or whether the person's head is shifted to the left or right, blocking the speaker or microphone, etc.

[0248] Step 4: Determine whether the current secondary path is a pre-stored secondary path based on its current status.

[0249] For example, if image analysis detects soft headrest occlusion, the current secondary path state is set to soft headrest occlusion state, and the current secondary path is determined to be a pre-stored secondary path. If image analysis shows that the states of each pre-stored secondary path do not match the current secondary path state, the current secondary path is determined not to be a pre-stored secondary path.

[0250] Step 5: If the current secondary path is a pre-stored secondary path, the corresponding pre-stored path parameters are called, and an initial step size (i.e., a larger step size) is used to update the control filter coefficients W to maintain the best noise reduction performance and the fastest convergence speed. For example, if the current secondary path state is detected to be a soft headrest occlusion state, the path parameters corresponding to the soft headrest occlusion state are called, and a larger update step size (e.g., 0.8) is used to update the control filter coefficients, while simultaneously performing divergence detection. If no divergence trend is detected, the signal obtained by convolving the updated filter coefficients with the reference signal is used as the new noise reduction signal and sent to the speaker. If a divergence trend is detected, the update step size is reduced, and divergence detection is repeated. For example, the step size reduction strategy is to reduce it to 1 / 2 of the original value each time, i.e., reduce the step size to 0.4 and repeat step 5 until the step size is reduced to less than a certain specified threshold (e.g., 10). -5 After that, turn off active noise cancellation.

[0251] Step six: If the current secondary path is not a pre-stored secondary path, reduce the update step size. For example, the step size reduction strategy is to reduce it to half of the original size each time, and use the reduced step size (i.e., the smaller step size) to update the control filter coefficients W. For instance, if the current secondary path is not a pre-stored secondary path, reduce the step size from 0.8 to 0.4 before calculating the control filter coefficients W, and perform divergence detection simultaneously. If no divergence trend is detected, the signal obtained by convolving the currently updated filter coefficients with the reference signal is used as the new noise-reduced signal and sent to the speaker. If a divergence trend is detected, continue to reduce the step size to 0.2 and repeat the divergence detection until the step size is reduced to less than a certain specified threshold (e.g., 1e-5), then turn off the active noise reduction operation.

[0252] Optionally, in step five or six, the specific method for divergence detection can be: calculating the next frame output signal based on the calculated control filter coefficients W that need to be updated and the acquired reference signal x. The calculation method is: next frame output signal y = W * x, where * represents convolution operation. It is then determined whether the energy (e.g., signal strength) of the next frame signal exceeds a specified threshold. If it does, it is determined that there is a divergence trend, and the control filter coefficients W are not updated. If it does not exceed the threshold, it is determined that there is no divergence trend, and the control filter coefficients are updated. This specified threshold needs to be determined through actual vehicle testing.

[0253] Based on the above implementation scheme 2, the cabin images captured by the cabin camera can be used to calculate the status information of the secondary path. In this way, non-acoustic components in the cabin can be called to complete the detection operation of the secondary path, thereby improving the flexibility of noise control.

[0254] It should be noted that the above content only uses the noise control method applied to vehicles as an example. This noise control method can also be extended to any device or system that requires noise reduction stability. For example, it can be applied to any mobile device with a cockpit, including but not limited to ships, airplanes, high-speed trains, trains, helicopters, lawnmowers, etc. Alternatively, it can be applied to smart home scenarios, such as reducing external environmental noise indoors to improve the user's home experience. And so on.

[0255] Furthermore, as system architecture evolves and new scenarios emerge, the noise control method provided in this application is also applicable to similar technical problems, and this application does not impose any specific limitations on it.

[0256] Based on the noise control method described above, this application can also provide a noise control device that can be used to perform the above noise control method. The relevant features can be found in the above method embodiments, and will not be repeated here.

[0257] In one possible implementation, please refer to Figure 13, which shows a possible structural schematic diagram of a noise control device. The noise control device 1300 may be a cockpit or a module within the cockpit (such as a processor, chip, or chip system), or it may be an apparatus applied to or used in conjunction with a cockpit or its module to implement a noise control method performed by the cockpit or its module. The noise control device 1300 may include various units or modules for implementing the noise control methods in the embodiments shown in Figures 4, 6, 8, 10a, 10b, 11, 12a, and 12b.

[0258] As shown in Figure 13, the noise control device 1300 may include a control unit 1310, an acquisition unit 1320, a determination unit 1330, and an update unit 1340. The control unit 1310, acquisition unit 1320, determination unit 1330, and update unit 1340 can be used to implement the noise control methods in the embodiments shown in Figures 4, 6, 8, 10a, 10b, 11, 12a, and 12b. For example, when the noise control device 1300 executes the noise control method shown in Figure 4 above, the control unit 1310 is used to control the speaker to play a noise reduction signal; the acquisition unit 1320 is used to acquire the first reference signal, the first error signal, and the first state information of the current secondary path; the determination unit 1330 is used to match the first state information with the state information of multiple pre-stored secondary paths. If a matching pre-stored secondary path exists, the filter coefficients are determined according to the path parameters of the matching pre-stored secondary path, the first reference signal, and the first error signal, and the filter coefficients are adjusted using the first step length. If no matching pre-stored secondary path exists, the filter coefficients are determined according to the path parameters of the default secondary path, the first reference signal, and the first error signal, and the filter coefficients are adjusted using the second step length, which is smaller than the first step length; the update unit 1340 is used to update the noise reduction signal using the adjusted filter coefficients.

[0259] It should be noted that the aforementioned control unit 1310, acquisition unit 1320, determination unit 1330, and update unit 1340 can be implemented using virtual modules. For example, control unit 1310 can be implemented using a software function unit or virtual device, acquisition unit 1320 can be implemented using a software function or virtual device, determination unit 1330 can be implemented using a software function or virtual device, and update unit 1340 can be implemented using a software function or virtual device. Alternatively, control unit 1310, acquisition unit 1320, determination unit 1330, and update unit 1340 can also be implemented using physical devices. For example, if the noise control device 1300 is implemented using a chip / chip circuit, control unit 1310, acquisition unit 1320, determination unit 1330, and update unit 1340 can be integrated processors, microprocessors, or integrated circuits.

[0260] The unit division in this application embodiment is illustrative and only represents one logical functional division. In actual implementation, other division methods may be used. Furthermore, the functional units in each embodiment of this application can be integrated into a single processor, exist as separate physical units, or two or more units can be integrated into a single module. The integrated module can be implemented in hardware or as a software functional module.

[0261] In another possible implementation, please refer to Figure 14, which shows another possible structural schematic diagram of the noise control device. For example, the noise control device 1400 may be a chip or a chip system, used to implement the functions of the noise control device or its modules (such as processors, chips, or chip systems) described in the foregoing embodiments. Optionally, in the embodiments of this application, the chip system may be composed of chips or may include chips and other discrete devices.

[0262] As shown in FIG. 14, the noise control device 1400 may include at least one processor 1410, which is coupled to a memory. Optionally, the memory may be located within the noise control device 1400, integrated with the processor, or located outside the noise control device 1400. For example, the noise control device 1400 may also include at least one memory 1420. The at least one memory 1420 stores the necessary computer programs (or instructions) and / or data for implementing any of the above embodiments; the at least one processor 1410 may execute the computer programs (or instructions) and / or data stored in the at least one memory 1420 to complete the methods in any of the above embodiments.

[0263] The noise control device 1400 may also include a communication interface 1430, through which the noise control device 1400 can interact with other devices. For example, the communication interface 1430 may be a transceiver, circuit, bus, module, pin, or other type of communication interface. When the noise control device 1400 is a chip-type device or circuit, the communication interface 1430 may also be an input / output circuit, capable of inputting information (or receiving information) and outputting information (or sending information). The processor may be an integrated processor, microprocessor, integrated circuit, or logic circuit, and the processor can determine the output information based on the input information.

[0264] The coupling in this embodiment is an indirect coupling or communication connection between devices, units, or modules, which can be electrical, mechanical, or other forms, used for information exchange between devices, units, or modules. The processor 1410 may operate in conjunction with the memory 1420 and the communication interface 1430. This embodiment does not limit the specific connection medium between the processor 1410, the memory 1420, and the communication interface 1430.

[0265] Optionally, referring to Figure 14, the processor 1410, the memory 1420, and the communication interface 1430 are interconnected via a bus. The bus can be a peripheral component interconnect (PCI) bus or an extended industry standard architecture (EISA) bus, etc. The bus can be divided into address bus, data bus, control bus, etc. For ease of illustration, only one thick line is used in Figure 14, but this does not indicate that there is only one bus or one type of bus.

[0266] In the embodiments of this application, the processor 1410 may be a general-purpose processor, a digital signal processor, an application-specific integrated circuit, a field-programmable gate array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components, capable of implementing or executing the methods, steps, and logic block diagrams disclosed in the embodiments of this application. The general-purpose processor may be a microprocessor or any conventional processor. The steps of the methods disclosed in the embodiments of this application can be directly manifested as being executed by a hardware processor, or executed by a combination of hardware and software modules within the processor.

[0267] In this embodiment, the memory 1420 can be a non-volatile memory, such as a hard disk drive (HDD) or a solid-state drive (SSD), or it can be volatile memory, such as random-access memory (RAM). The memory 1420 can be any other medium capable of carrying or storing desired program code in the form of instructions or data structures, and accessible by a computer, but is not limited thereto. The memory 1420 in this embodiment can also be a circuit or any other device capable of implementing a storage function for storing program instructions and / or data.

[0268] Based on the above, this application also provides a noise control system, including the above noise control device. The noise control device may be the noise control device 1300 in FIG13 or the noise control device 1400 in FIG14, for executing the noise control method described in the above embodiments.

[0269] Optionally, as shown in Figure 2 above, the noise control system may further include a loudspeaker, a reference sensor, and an error sensor, with the noise control device connected to the loudspeaker, reference sensor, and error sensor respectively. When performing the noise control function, the loudspeaker plays a noise-reducing signal, the reference sensor acquires a reference signal and sends it to the noise control device, and the error sensor acquires an error signal and sends it to the noise control device. The noise control device then executes the noise control method described in the above embodiments based on the noise-reducing signal, the reference signal, and the error signal.

[0270] Alternatively, when the noise control system is installed in the cabin, the error sensor is located in the headrest, and the speaker is located in the headrest, seat, or inside the door.

[0271] Optionally, as shown in Figure 2 above, the noise control system may further include a visual sensor, such as the aforementioned cabin camera, and the noise control device is connected to the cabin camera. When performing the noise control function, the cabin camera is used to acquire cabin images and send them to the noise control device, which determines the first state information of the current secondary path based on the cabin images.

[0272] It should be noted that the functions of each device in the above noise control system, the concepts involved and related to the technical solutions provided in this application, explanations, detailed descriptions and other steps are described in the foregoing method embodiments, and will not be repeated here.

[0273] Based on the above, this application may also provide a means of transportation that includes the above noise control devices, such as the noise control device 1300 shown in FIG13 or the noise control device 1400 shown in FIG14, or includes the above noise control system, such as the noise control system shown in FIG2.

[0274] For example, the means of transportation can be a vehicle, such as a car, truck, motorcycle, bus, recreational vehicle, amusement park vehicle, construction equipment, tram, toy car, golf cart, train, etc., and this application does not impose any particular limitation. In addition, the vehicle can be a new energy vehicle, including electric vehicles, such as two-wheel drive electric vehicles or four-wheel drive electric vehicles, or a fuel-powered vehicle, and this application does not impose any limitation in this regard.

[0275] Based on the above, this application can also provide an electronic device connected to a noise control vehicle for communicating with the noise control vehicle to implement the above noise control method. For example, the electronic device can be a terminal device or another vehicle, or other devices capable of controlling the vehicle's loudspeakers. The electronic device includes units or modules for implementing the above noise control method, such as the noise control device 1300 shown in Figure 13 above, or the noise control device 1400 shown in Figure 14 above.

[0276] Based on the above, this application also provides a computer-readable storage medium storing instructions that, when executed, cause the method provided in any of the above-described method embodiments to be implemented. The computer-readable storage medium may include various media capable of storing program code, such as a USB flash drive, portable hard drive, read-only memory, random access memory, magnetic disk, or optical disk.

[0277] Based on the above, this application also provides a computer program product, which includes: a computer program (also referred to as code or instructions), which, when run on a computer, causes the computer to perform the method provided in any of the above method embodiments.

[0278] In this application, "at least one" means one or more, and "more than one" means two or more. "And / or" describes the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A alone, A and B simultaneously, or B alone, where A and B can be singular or plural. Furthermore, the various numbers involved in the embodiments of this application (such as the numerical numbers "first," "second," "third," etc.) are only for descriptive convenience and are not intended to limit the scope of the embodiments of this application. The order of the sequence numbers of the above processes does not imply the order of execution; the execution order of each process should be determined by its function and internal logic.

[0279] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product implemented on one or more computer-usable storage media (including, but not limited to, disk storage, compact disc read-only memory (CD-ROM), optical storage, etc.) containing computer-usable program code.

[0280] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to this application. It should be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in one or more flowchart illustrations and / or one or more block diagrams.

[0281] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means that implement the functions specified in one or more flowcharts and / or one or more block diagrams.

[0282] These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer-implemented process, such that the instructions, which execute on the computer or other programmable apparatus, provide steps for implementing the functions specified in one or more flowcharts and / or one or more block diagrams.

Claims

1. A noise control method characterized by, include: Control the speaker to play the noise-reduced signal; Acquire the first reference signal, the first error signal, and the first state information of the current secondary path; The first state information is matched with the state information of multiple pre-stored secondary paths. If a matching pre-stored secondary path exists, the filter coefficients are determined based on the path parameters of the matching pre-stored secondary path, the first reference signal, and the first error signal, and the filter coefficients are adjusted using the first step length. If no matching pre-stored secondary path exists, the filter coefficients are determined based on the path parameters of the default secondary path, the first reference signal, and the first error signal, and the filter coefficients are adjusted using a second step size, where the second step size is smaller than the first step size. The noise-reduced signal is updated using the adjusted filter coefficients.

2. The method as described in claim 1, characterized in that, The default secondary path is one of the plurality of pre-stored secondary paths; The default secondary path is: a secondary path in which neither the speaker nor the error sensor is obstructed; Other pre-stored secondary paths include at least one of the following: a secondary path in which at least one, but not all, of the multiple speakers is blocked; a secondary path in which all speakers are blocked; a secondary path in which at least one, but not all, of the multiple error sensors is blocked; and a secondary path in which all error sensors are blocked.

3. The method of claim 1 or 2, wherein, Before the speaker is controlled to play the noise-reducing signal, the following is also included: Confirm that noise cancellation is enabled; Acquire the second reference signal acquired by the reference sensor and the second error signal acquired by the error sensor; Based on the second reference signal, the second error signal, and the path parameters of the default secondary path, the initial filter coefficients are determined, and the initial filter coefficients are adjusted using the first step length. The noise reduction signal is determined based on the adjusted initial filter coefficients and the second reference signal.

4. The method of claim 3, wherein, The step of determining to enable noise reduction includes: In response to a noise reduction command, the vehicle speed and the status of the body switch components are obtained, and it is determined that the vehicle speed is greater than a first vehicle speed and that the status of the body switch components is closed. The vehicle body switch components include one or more of the following: doors, windows, and sunroofs.

5. The method of any one of claims 1 to 4, wherein, The step of obtaining the first state information of the current secondary path includes: Control the speaker to play the detection signal and acquire the first error signal collected by the error sensor; The first state information of the current secondary path is determined based on the detection signal and the first error signal. The first state information includes one or more of the transfer function, cross-spectral function, and cross-correlation coefficient.

6. The method of claim 5, wherein, The control of the loudspeaker to play the detection signal includes: Generate an initial signal randomly or according to a set rule; The third error signal collected by the error sensor is acquired, and the initial signal is subjected to spectrum shaping and gain control based on the third error signal to obtain the detection signal. The spectrum shape of the detection signal is close to the spectrum shape of the third error signal. The noise intensity generated at the error sensor after the detection signal is played by the speaker is lower than the noise intensity of the third error signal, and the difference between the noise intensity of the detection signal and the noise intensity of the third error signal is within a set range. The detection signal and the noise reduction signal are superimposed, and the speaker is controlled to play the superimposed signal.

7. The method of claim 5 or 6, wherein, The step of matching the first state information with the state information of multiple pre-stored secondary paths includes: For each of the plurality of pre-stored secondary paths, the deviation between the first state information and the state information of the pre-stored secondary path is determined. If the deviation is within the allowable deviation fluctuation range, then the first state information is determined to match the state information of the pre-stored secondary path.

8. The method of any one of claims 5 to 7, wherein, The status information of the multiple pre-stored secondary paths is obtained in the following way: In a silent environment, the speaker and the error sensor are configured in the scenario corresponding to each pre-stored secondary path. The speaker is controlled to play a test signal, and the error signal collected by the error sensor is acquired. The state information of each pre-stored secondary path is determined based on the test signal and the error signal.

9. The method of any one of claims 1 to 4, wherein, The step of obtaining the first state information of the current secondary path and matching the first state information with the state information of multiple pre-stored secondary paths includes: Acquire cabin images captured by the cabin camera, identify image features of the areas where the speakers and error sensors are located in the cabin images, and match the image features with the image features of the multiple pre-stored secondary paths.

10. The method of any one of claims 1 to 9, wherein, Updating the noise-reduced signal using the adjusted filter coefficients includes: Divergence detection is performed on the adjusted filter coefficients; If there is no divergence trend, the noise reduction signal is updated using the adjusted filter coefficients; If a divergence trend exists, the step size is reduced, and the filter coefficients are readjusted using the reduced step size. The noise-reduced signal is then updated using the readjusted filter coefficients.

11. The method of claim 10, wherein, The divergence detection of the adjusted filter coefficients includes: The adjusted filter coefficients and the first reference signal are convolved to obtain the output signal; If the signal strength of the signal to be output is less than or equal to the set strength, it is determined that the adjusted filter coefficients do not have a divergence trend; if the signal strength of the signal to be output is greater than the set strength, it is determined that the adjusted filter coefficients have a divergence trend.

12. The method of claim 10 or 11, wherein, The method further includes: When the reduced step size is less than or equal to the set step size, the noise reduction operation is completely or partially terminated.

13. A noise control device, characterized by include: The control unit is used to control the speaker to play the noise-reducing signal; The acquisition unit is used to acquire the first reference signal, the first error signal, and the first state information of the current secondary path; The determining unit is used to match the first state information with the state information of multiple pre-stored secondary paths. If a matching pre-stored secondary path exists, the filter coefficients are determined according to the path parameters of the matching pre-stored secondary path, the first reference signal, and the first error signal, and the filter coefficients are adjusted using the first step length. If no matching pre-stored secondary path exists, the filter coefficients are determined based on the path parameters of the default secondary path, the first reference signal, and the first error signal, and the filter coefficients are adjusted using a second step size, where the second step size is smaller than the first step size. An update unit is used to update the noise-reduced signal using the adjusted filter coefficients.

14. The apparatus as claimed in claim 13, characterized in that, The default secondary path is one of the plurality of pre-stored secondary paths; The default secondary path is: a secondary path in which neither the speaker nor the error sensor is obstructed; Other pre-stored secondary paths include at least one of the following: a secondary path in which at least one, but not all, of the multiple speakers is blocked; a secondary path in which all speakers are blocked; a secondary path in which at least one, but not all, of the multiple error sensors is blocked; and a secondary path in which all error sensors are blocked.

15. The apparatus of claim 13 or 14, wherein, Before the control unit controls the speaker to play the noise-reducing signal: The determining unit is also used to: determine whether to enable noise reduction operation; The acquisition unit is further configured to: acquire the second reference signal acquired by the reference sensor and the second error signal acquired by the error sensor; The determining unit is further configured to: determine initial filter coefficients based on the second reference signal, the second error signal, and the path parameters of the default secondary path; adjust the initial filter coefficients using the first step length; and determine the noise reduction signal based on the adjusted initial filter coefficients and the second reference signal.

16. The apparatus of claim 15, wherein, The determining unit is specifically used for: In response to a noise reduction command, the vehicle speed and the status of the body switch components are obtained, and it is determined that the vehicle speed is greater than a first vehicle speed and that the status of the body switch components is closed. The vehicle body switch components include one or more of the following: doors, windows, and sunroofs.

17. The apparatus of any one of claims 13 to 16, wherein, The acquisition unit is specifically used for: Control the speaker to play the detection signal and acquire the first error signal collected by the error sensor; The first state information of the current secondary path is determined based on the detection signal and the first error signal. The first state information includes one or more of the transfer function, cross-spectral function, and cross-correlation coefficient.

18. The apparatus of claim 17, wherein, The acquisition unit is specifically used for: Generate an initial signal randomly or according to a set rule; The third error signal collected by the error sensor is acquired, and the initial signal is subjected to spectrum shaping and gain control based on the third error signal to obtain the detection signal. The spectrum shape of the detection signal is close to the spectrum shape of the third error signal. The noise intensity generated at the error sensor after the detection signal is played by the speaker is lower than the noise intensity of the third error signal, and the difference between the noise intensity of the detection signal and the noise intensity of the third error signal is within a set range. The detection signal and the noise reduction signal are superimposed, and the speaker is controlled to play the superimposed signal.

19. The apparatus of claim 17 or 18, wherein, The determining unit is specifically used for: For each of the plurality of pre-stored secondary paths, the deviation between the first state information and the state information of the pre-stored secondary path is determined. If the deviation is within the allowable deviation fluctuation range, the first state information is determined to match the state information of the pre-stored secondary path.

20. The apparatus of any one of claims 17 to 19, wherein, The status information of the multiple pre-stored secondary paths is obtained in the following way: In a silent environment, the speaker and the error sensor are configured in the scenario corresponding to each pre-stored secondary path. The speaker is controlled to play a test signal, and the error signal collected by the error sensor is acquired. The state information of each pre-stored secondary path is determined based on the test signal and the error signal.

21. The apparatus as claimed in any one of claims 13 to 16, characterized in that, The acquisition unit is specifically used to: acquire cabin images captured by the cabin camera, and identify image features of the areas where the speakers and error sensors are located in the cabin images; The determining unit is specifically used to match the image features with the image features of the plurality of pre-stored secondary paths.

22. The apparatus of any one of claims 13 to 21, wherein, The update unit is specifically used for: Divergence detection is performed on the adjusted filter coefficients; If there is no divergence trend, the noise reduction signal is updated using the adjusted filter coefficients; If a divergence trend exists, the step size is reduced, and the filter coefficients are readjusted using the reduced step size. The noise-reduced signal is then updated using the readjusted filter coefficients.

23. The apparatus of claim 22, wherein, The update unit is specifically used for: The adjusted filter coefficients and the first reference signal are convolved to obtain the output signal; If the signal strength of the signal to be output is less than or equal to the set strength, then it is determined that the adjusted filter coefficients do not have a divergence trend. If the signal strength of the signal to be output is greater than the set strength, then it is determined that the adjusted filter coefficients have a divergence trend.

24. The apparatus of claim 22 or 23, wherein, The control unit is also used for: When the reduced step size is less than or equal to the set step size, the noise reduction operation is completely or partially terminated.

25. A noise control device, characterized by Includes a processor, which is coupled to memory: The processor is configured to execute a computer program or instructions stored in the memory to cause the noise control device to perform the method as described in any one of claims 1 to 12.

26. A noise control system characterized by, The noise control device, as described in any one of claims 13 to 25, further includes a loudspeaker, a reference sensor, and an error sensor, wherein the noise control device is connected to the loudspeaker, the reference sensor, and the error sensor, respectively. The speaker is used to play noise-reduced signals; The reference sensor is used to acquire the first reference signal; The error sensor is used to collect the first error signal; The noise control device is configured to perform the method as described in any one of claims 1 to 12 based on the noise reduction signal, the first reference signal, and the first error signal.

27. The system of claim 26, wherein, The noise control system is located in the cabin, the error sensor is located in the headrest, and the speaker is located in the headrest, seat, or inside the door.

28. A vehicle characterized by Including the noise control system as described in claim 26 or 27.

29. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a program or instructions that, when executed, implement the method as described in any one of claims 1 to 12.

30. A computer program product, characterised in that, The computer program product includes computer program code that, when run on a computer, causes the computer to perform the method as described in any one of claims 1 to 12.

Images