A system applied to in-vehicle road noise partition active control and a control method thereof

Abstract

The application relates to the field of automobile technology, in particular to a system applied to in-vehicle road noise partition active control and a control method thereof. Mainly, noise data collected by an automobile chassis vehicle-mounted sensor is taken as a reference signal, an in-vehicle sensor is used to identify a control area, and the control area is acquired; based on maximum noise reduction of a target control area, a maximum weighted multiple coherence analysis method is used to screen out an optimal reference signal combination with the best coherence with the target control area, and a multi-channel active road noise control algorithm with low calculation complexity and low delay is used to generate a secondary noise with the same amplitude and opposite phase as a noise signal (in-vehicle primary noise) to be cancelled, so that interference cancellation of noise in the control area is realized.

Description

Technical Field

[0001] This invention relates to the field of automotive technology, and in particular to a system and control method for active control of road noise zones within a vehicle. Background Technology

[0002] As automotive powertrains become increasingly electrified, the masking effect of internal combustion engine noise is lost, making road noise the primary source of interior noise in electric vehicles. Effectively reducing in-vehicle road noise is a major challenge for electric vehicles. Traditional passive noise reduction (PNC) technologies achieve this through vehicle structure design or by adding vibration absorbers, but these suffer from poor low-frequency noise control and complex optimization schemes. Active noise control (ANC), on the other hand, is more effective at reducing road noise with broadband random low-frequency characteristics without affecting the vehicle's hardware structure and performance.

[0003] Active Road Noise Control (ARNC) technology is based on the principle of acoustic wave interference cancellation. It introduces a loudspeaker (secondary sound source) into the sound field and controls it to emit cancellation noise with the same amplitude but opposite phase to the noise to be canceled (primary noise), thereby creating a quiet zone in a specific area. The ARNC system can not only effectively reduce low-frequency noise that is difficult to control using PNC methods, but also adaptively control the system by tracking changes in the road noise spectrum inside the vehicle.

[0004] ARNC technology faces two key challenges in real-world vehicle applications: accurate acquisition of reference signals with good correlation to in-vehicle road noise, and implementation of adaptive filtering algorithms with low computational complexity and high noise reduction. The reference signal provides prior information for active control of in-vehicle road noise, and the correlation between the two is directly related to the noise reduction amount. Currently, most ARNC systems employ the traditional filter-x least mean square (FxLMS) algorithm. However, the FxLMS algorithm suffers from high computational complexity, slow convergence speed, and poor stability in ARNC applications.

[0005] Meanwhile, theoretical research has revealed that zoned control of in-vehicle road noise based on the number and location of passengers can, to some extent, improve noise reduction in specific areas and enhance in-vehicle sound quality. Since multiple passenger seats in a car require noise reduction, and each seat has a different noise transmission path, balanced control of the entire vehicle area will negatively impact noise reduction in the target area compared to zoned control. Controlling based on passenger seating arrangements not only improves noise reduction in localized areas but also efficiently utilizes in-vehicle resources. Therefore, in-vehicle road noise control should be zoned based on passenger seating arrangements. However, using a full-area (including target and non-target areas) balanced control approach for in-vehicle road noise control can prevent the controller's reference signal combination selection from achieving optimal control in the target area. Summary of the Invention

[0006] The technical problem to be solved by this invention is to provide a system and control method for active control of road noise zones in vehicles, which addresses the shortcomings of existing technologies. This system can use active noise control technology to control the low- and mid-frequency noise of a vehicle in zones, thereby achieving noise reduction in the vehicle and solving the problem of prominent low- and mid-frequency road noise in electric vehicles after the loss of the masking effect of engine noise.

[0007] To solve the above-mentioned technical problems, the technical solution adopted by the present invention is as follows:

[0008] An active control system for road noise zoning in a vehicle includes multiple sets of sound pressure sensors 1 installed inside the vehicle to collect road noise signals to be eliminated inside the vehicle and multiple sets of vibration acceleration sensors 2 installed outside the vehicle to collect reference signals of external noise. The vehicle is also equipped with a PCR radar sensor 3 for identifying the seating status of each occupant and determining the target control area. The sound pressure sensors 1, vibration acceleration sensors 2 and PCR radar sensors 3 are all electrically connected to a DSP controller 4.

[0009] The DSP controller 4 adopts an in-vehicle road noise zoning active control method, which controls the headrest speaker 5 to generate a secondary cancellation signal with the same amplitude but opposite phase as the in-vehicle road noise signal to be canceled in the target control area, thereby performing noise reduction control in the in-vehicle target control area.

[0010] Furthermore, the specific arrangement positions of the multiple sets of sound pressure sensors 1 are as follows: one on the right side of the driver's seat, one on the left ear of the passenger seat, one on the right ear of the left rear seat, and one on the left ear of the right rear seat.

[0011] Furthermore, the specific arrangement positions of the multiple sets of vibration acceleration sensors 2 are as follows: one is arranged at the passive end of the connection point between the front end of the left and right triangular arms of the front suspension and the shock absorber; one is arranged at the passive end of the connection point between the rear end of the left and right triangular arms of the front suspension and the shock absorber; one is arranged at the passive end of the upper end of the left and right shock absorber dampers of the front suspension; one is arranged at the passive end of the left and right front wheel centers; one is arranged at the passive end of the middle of the front axle subframe; one is arranged at the passive end of the middle of the left and right connecting beams between the body and the chassis; one is arranged at the passive end of the front end of the left and right torsion beams of the rear suspension; one is arranged at the passive end of the rear end of the left and right torsion beams of the rear suspension; one is arranged at the passive end of the upper end of the left and right shock absorber dampers of the rear suspension; one is arranged at the passive end of the left and right rear wheel centers; and one is arranged at the passive end of the middle of the rear torsion beam.

[0012] An active road noise zoning control method for in-vehicle environments, based on the active road noise zoning control system described above, specifically includes the following steps:

[0013] S1, the PCR radar sensor identifies and senses the seating position of the occupants inside the vehicle, and determines the target control area and non-target control area for noise reduction inside the vehicle.

[0014] S2, assign corresponding weighting factors to the target control region and non-target control region;

[0015] S3, under multi-condition driving conditions, multiple sets of vibration acceleration sensors and multiple sets of sound pressure sensors are used to collect noise spectrum signals. The noise spectrum signal collected by the vibration acceleration sensor is used as the reference signal x(n), and the noise spectrum signal collected by the sound pressure sensor is used as the signal to be eliminated d(n).

[0016] S4, the reference signal x(n) and the signal to be canceled d(n) are stored and read through the DSP controller, and the optimal combination of reference signals is selected by combining the weighting factor and using the maximum weighted multicoherence analysis method.

[0017] S5, based on the selected optimal reference signal combination, a secondary cancellation signal is generated using the time-frequency domain multi-channel FxLMS algorithm;

[0018] S6, the headrest speakers in the target control area generate corresponding secondary sound sources according to the secondary cancellation signal to achieve noise reduction in the target area.

[0019] Furthermore, in step S1, the specific process of determining the target control area and non-target control area for in-vehicle noise reduction is as follows:

[0020] Determine if anyone is seated in the driver's or passenger's seat. If so, the seat is the target control area; otherwise, the seat is a non-target control area.

[0021] Furthermore, in step S2, the allocation of the corresponding weighting factor is specifically carried out as follows:

[0022] The coherence coefficient weighting value between the signal to be cancelled and the reference signal in the target control region is set to 1, and the coherence coefficient weighting value between the signal to be cancelled and the reference signal in the non-target control region is set to 0.

[0023] Furthermore, in step S3, the multi-condition driving conditions specifically refer to:

[0024] The new energy test vehicle, which adopts a MacPherson independent suspension structure at the front and a torsion beam non-independent suspension structure at the rear, serves as the controller to test the controlled object.

[0025] The vehicle was driven on rough asphalt roads and smooth asphalt roads. The vehicle speed was set to cruise control mode. Each 10 km / h increase from 30 to 120 km / h was considered a steady-state test condition. The acceleration method was full-throttle acceleration, which was used as a time-varying unsteady-state test condition.

[0026] Furthermore, in step S4, the optimal reference signal combination is selected using the maximum weighted multicoherence analysis method, and the specific calculation formula is as follows:

[0027]

[0028] In the formula: N is the number of in-vehicle control areas. The highest frequency point, The lowest frequency point, This is the weighted value of the coherence coefficient for each region; The multicoherence coefficient is calculated using the following formula:

[0029]

[0030] In the formula: This represents the theoretical maximum noise reduction. Reference signal The resulting filtered output signal The self-power spectrum; Signal to be cancelled Self-power spectrum, for and mutual spectrum, Reference signal Self-power spectrum.

[0031] Furthermore, in step S5, the specific process of the time-frequency domain multi-channel FxLMS algorithm is as follows:

[0032] Step 1:

[0033] Reference signal Treat it as an extended main block, appending it at its beginning and end. Then divide the extended main block into zeros. There are several overlapping sub-blocks, which are windowed by a half-period sinusoidal window function. Then, all sub-blocks are converted from the time domain to the frequency domain. The calculation formula is as follows:

[0034]

[0035] In the formula: Represented as the first in the frequency domain The first block A windowed master sub-block reference signal group Elements of the extended block, This is represented by the Fast Fourier Transform, which converts a time-domain signal into a frequency-domain signal. Represented as a window function, windowing is applied during signal conversion to prevent signal leakage and distortion.

[0036] Step Two:

[0037] The formula for calculating the filtered frequency domain output signal is:

[0038]

[0039] In the formula: For the first in the frequency domain The first block The filtered output signal of the windowed block For the adaptive filter in the frequency domain The coefficient of each block;

[0040] Step 3:

[0041] The frequency domain filtered output signal needs to be reconstructed into the time domain through an inverse fast Fourier transform, and the calculation formula is as follows:

[0042]

[0043] In the formula: Represented as Inverse Fast Fourier Transform;

[0044] The formula for calculating the desired reconstructed filtered output signal is:

[0045]

[0046] In the formula: superscript " " indicates transpose;

[0047] Step Four:

[0048] The residual error signal is represented as:

[0049]

[0050] In the formula: The signal to be canceled. The impulse response of the secondary path in the time domain. For the filtered output signal to be reconstructed, This is represented as a convolution calculation;

[0051] Step 5:

[0052] The residual error signal obtained in step four is divided into The signal is divided into several overlapping sub-blocks. Each sub-block is then windowed and converted from a time-domain signal to a frequency-domain signal. The calculation formula is as follows:

[0053]

[0054] In the formula: It is the first in the frequency domain The first block A windowed residual sub-block, It is a windowed residual sub-block. It is a block element;

[0055] Step Six:

[0056] The adaptive filter weight coefficients are updated as follows:

[0057]

[0058] In the formula: It is expressed as the normalized convergence factor of the frequency domain with a windowed block. For filtering reference signal, This is represented as an error signal;

[0059] The filtered reference signal is represented as:

[0060]

[0061] In the formula: It is represented as the impulse response of the secondary path in the frequency domain.

[0062] Compared with the prior art, the present invention has the following main advantages:

[0063] 1. This application aims to achieve the best noise reduction effect in the target control area. It selects the optimal combination of reference signals based on the target areas inside the vehicle under different driving and riding area scenarios and uses active noise control technology to control the low and medium frequency noise of the car in a zone to achieve the purpose of noise reduction inside the vehicle. This can solve the problem of the prominence of low and medium frequency road noise in electric vehicles after the loss of the masking effect of engine order noise.

[0064] 2. This application adopts an active noise control algorithm based on the time-frequency domain multi-channel FxLMS algorithm. It calculates gradient estimation and filter reference signal in the frequency domain to reduce computational complexity, and updates control filter coefficients in the time domain to minimize delay. The control algorithm has fast convergence speed, low delay and high stability. Attached Figure Description

[0065] Figure 1 This is a schematic diagram of the in-vehicle road noise zoning active control system in an embodiment of the present invention;

[0066] Figure 2 This is a flowchart of the in-vehicle road noise zoning active control method in an embodiment of the present invention;

[0067] Figure 3 This is a flowchart of the maximum weighted multiple coherence analysis method in an embodiment of the present invention;

[0068] Figure 4 This is a flowchart of the time-frequency domain multi-channel FxLMS algorithm in an embodiment of the present invention;

[0069] Figure 5This is a diagram showing the centralized region theoretical noise reduction calculation results in an embodiment of the present invention;

[0070] Figure 6 This is a diagram showing the distributed region theoretical noise reduction calculation results in an embodiment of the present invention.

[0071] In the diagram: 1. Sound pressure sensor (microphone); 2. Vibration acceleration sensor; 3. PCR radar sensor; 4. DSP controller; 5. Headrest speaker. Detailed Implementation

[0072] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention. Furthermore, the technical features involved in the various embodiments of this invention described below can be combined with each other as long as they do not conflict with each other.

[0073] It should be noted that, depending on the implementation needs, the various steps / components described in this application can be broken down into more steps / components, or two or more steps / components or parts of the operation of steps / components can be combined into new steps / components to achieve the purpose of this invention.

[0074] This application uses vehicle-mounted sensors on the vehicle chassis to collect noise data as a reference signal, and uses in-vehicle sensors to identify the control area to obtain the control area. Based on maximizing the noise reduction of the target control area, the maximum weighted multicoherence analysis method is used to select the optimal reference signal combination with the best coherence to the target control area. Then, a multi-channel active road noise control algorithm with low computational complexity and low latency is used to generate secondary noise with the same amplitude and opposite phase as the noise signal to be canceled (in-vehicle primary noise) to achieve interference cancellation of noise in the target control area.

[0075] I. Active Road Noise Zoning Control System

[0076] like Figure 1 As shown, an in-vehicle road noise zoning active control system according to the present invention includes at least:

[0077] 1) Multiple sound pressure sensors 1 (microphones) are installed inside the vehicle to collect road noise signals to be canceled;

[0078] 2) Multiple sets of vibration acceleration sensors 2 are installed outside the vehicle to collect external noise reference signals;

[0079] 3) PCR radar sensor 3, installed inside the vehicle, is used to identify the seating arrangement of each passenger and determine the target control area;

[0080] The sound pressure sensor 1, vibration acceleration sensor 2, and PCR radar sensor 3 are all electrically connected to the DSP controller 4.

[0081] 4) The DSP controller 4 adopts an in-vehicle road noise zoning active control method to control the headrest speaker 5 to generate a secondary cancellation signal with the same amplitude and opposite phase as the in-vehicle road noise to be canceled signal in the target control area, so as to perform noise reduction control in the in-vehicle target control area.

[0082] Specifically, the arrangement of the multiple sets of sound pressure sensors 1 (microphones) is as follows:

[0083] One earpiece is placed on the right side of the driver's seat (DR), one on the left side of the passenger seat (PL), one on the right side of the rear left seat (RLR), and one on the left side of the rear right seat (RRL).

[0084] The arrangement positions of the multiple sets of vibration acceleration sensors 2 are as follows:

[0085] One vibration acceleration sensor is placed at the passive end of the connection point between the front end of the left and right triangular arms and the shock absorber on each side of the front suspension; one is placed at the passive end of the connection point between the rear end of the left and right triangular arms and the shock absorber on each side of the front suspension; one is placed at the passive end of the upper end of the left and right shock absorber dampers on each side of the front suspension; one is placed at the passive end of the center of the left and right front wheel hubs; one is placed at the passive end of the middle of the front axle subframe; one is placed at the passive end of the middle of the left and right connecting beams between the body and the chassis on each side of the front suspension; one is placed at the passive end of the front end of the left and right torsion beams on each side of the rear suspension; one is placed at the passive end of the upper end of the left and right shock absorber dampers on each side of the rear suspension; one is placed at the passive end of the center of the left and right rear wheel hubs on each side of the rear suspension; and one is placed at the passive end of the middle of the rear suspension torsion beam. A total of 20 sampling points are set up with 20 vibration acceleration sensors.

[0086] Furthermore, the vehicle driving conditions and operating conditions in this embodiment are as follows:

[0087] A new energy test vehicle with a MacPherson independent front suspension and a torsion beam non-independent rear suspension was used as the controller to test the controlled object, and vibration and noise were tested on it on a rough asphalt road surface.

[0088] The vehicle speed was set to cruise control mode, with each 10 km / h increase from 30 to 120 km / h serving as a steady-state test condition; the acceleration method was full-throttle acceleration, which was used as a time-varying non-steady-state test condition.

[0089] The spectral noise signal acquired by the accelerometer is the reference signal, and the spectral noise signal from the microphone next to the ear is the signal to be eliminated, i.e., the primary signal.

[0090] The test conditions are shown in the table below:

[0091]

[0092] The required instruments and equipment are shown in the table below:

[0093]

[0094] II. Active Road Noise Zone Control Methods

[0095] Based on the same inventive concept, this application also provides an in-vehicle road noise zoning active control method, based on the in-vehicle road noise zoning active control system described above, such as... Figure 2 As shown, the specific steps include the following:

[0096] S1, the PCR radar sensor identifies and senses the seating position of the occupants inside the vehicle, and determines the target control area and non-target control area for noise reduction inside the vehicle.

[0097] S2, assign corresponding weighting factors to the target control region and non-target control region;

[0098] S3, under multi-condition driving conditions, multiple sets of vibration acceleration sensors and multiple sets of sound pressure sensors are used to collect noise spectrum signals. The noise spectrum signal collected by the vibration acceleration sensor is used as the reference signal x(n), and the noise spectrum signal collected by the sound pressure sensor is used as the signal to be eliminated d(n).

[0099] S4, the reference signal x(n) and the signal to be canceled d(n) are stored and read through the DSP controller, and the optimal combination of reference signals is selected by combining the weighting factor and using the maximum weighted multicoherence analysis method.

[0100] S5, based on the selected optimal reference signal combination, a secondary cancellation signal is generated using the time-frequency domain multi-channel FxLMS algorithm;

[0101] S6, the headrest speakers in the target control area generate corresponding secondary sound sources according to the secondary cancellation signal to achieve noise reduction in the target area.

[0102] Specifically:

[0103] 1) In step S1, the PCR radar sensor is used to identify and sense the seating situation of the people in the vehicle. It should be noted that the PCR radar sensing technology is very mature. This patent only borrows from the existing technology. It is mainly used to identify and sense the seating situation of the occupants in each seat of the vehicle, that is, to distinguish how many occupants are in the vehicle and their respective seating positions, thereby forming an active control architecture (centralized and distributed).

[0104] Among them, centralized means that the vehicle is fully occupied, while distributed means that the vehicle is occupied separately, i.e., single area, two areas, and three areas.

[0105] The specific process for determining the target control area for noise reduction inside the vehicle is as follows: determine whether anyone is sitting in the driver's or passenger's seat. If so, the seat is the target control area; otherwise, the seat is a non-target control area.

[0106] 2) In step S2, the specific process of allocating the corresponding weighting factors is as follows:

[0107] When the PCR radar sensor system detects that there is only one occupant in the vehicle, i.e., the driver's seat, and road noise control is required, the weighted value of the coherence coefficient between the primary signal (signal to be cancelled) at the response point of the target control area and the reference signal at the chassis is set to 1, and the weighted values ​​at other non-target control areas are all set to 0. Similarly, when the PCR radar sensor system detects two areas, three areas, or full occupancy, the weighted value of the coherence coefficient between the primary signal (signal to be cancelled) at the response point of the corresponding target control area and the reference signal at the chassis is set to 1, while the weighted values ​​of the coherence coefficient in non-target control areas without passengers are all set to 0.

[0108] The acceleration sensor (reference signal acquisition device) located on the vehicle body, the microphone (signal acquisition device to be canceled) at the seat headrest, and the speaker (secondary signal generator) at the seat headrest are all in working condition and do not require separate control or operation, which helps to reduce the control complexity of the system.

[0109] 3) In step S3, the multi-condition driving conditions are specifically as follows:

[0110] The new energy test vehicle, which adopts a MacPherson independent suspension structure at the front and a torsion beam non-independent suspension structure at the rear, serves as the controller to test the controlled object.

[0111] The vehicle was driven on both rough and smooth asphalt surfaces. The speed was set to cruise control mode, with each 10 km / h increase from 30 to 120 km / h serving as a steady-state test condition. Acceleration was performed using full throttle, which was then used as a time-varying, unsteady-state test condition.

[0112] The spectral noise signal acquired by the accelerometer is the reference signal, and the spectral noise signal from the microphone next to the ear is the signal to be eliminated, i.e., the primary signal.

[0113] 4) In step S4, the maximum weighted multiple coherence analysis method is an improvement on the multiple coherence analysis method. Its principle is to calculate the coherence coefficient between the combined subset formed by multiple reference signals and the signal to be canceled at the response point, and to perform weighted calculation on the coherence coefficients calculated between the reference signals and the signal to be canceled in each region to obtain the reference signal combination with the largest noise reduction in the target control area. Based on the theoretical maximum noise reduction (the reference signal with the largest coherence), different combination signals are sorted and screened. Based on the screened reference points, the signals are recombined until the target effect is achieved.

[0114] like Figure 3 As shown, the specific process of selecting the optimal reference signal combination using the maximum weighted multicoherence analysis method is as follows:

[0115] In this application, the input signal is the vibration signal collected by 20 acceleration sensors at the passive end of the suspension-body connection, and the output signal is the sound pressure signal collected by a microphone placed at the passenger headrest inside the vehicle. The coherence coefficient is used as the basis for this. The target noise reduction amount is obtained using the following formula:

[0116] (1)

[0117] In equation (1), the multicoherence coefficient The calculation formula is as follows:

[0118] (2)

[0119] In the formula: This represents the theoretical maximum noise reduction. For multiple coherence coefficients, Input signal Caused output The self-power spectrum; Primary noise signal Self-power spectrum. for and The cross spectrum, with the superscript "T" indicating transpose, for From the power spectrum, the superscript "-1" indicates that the matrix is ​​being calculated as a generalized inverse.

[0120] First, calculate the multicoherence coefficients of all N reference signals and the signal to be canceled in each region. Then, weight the calculated coherence coefficients of each region according to the weighting rules, take the reference signal corresponding to the maximum value of the multicoherence coefficient, sort it as 1, and store the record.

[0121] In the second iteration loop, the reference signal ranked 1 is removed, the multicoherence coefficients of the remaining N-1 reference signals and the signal to be canceled are calculated, and the reference signal corresponding to the maximum value of the multicoherence coefficient is selected and ranked as 2.

[0122] This process is repeated until the last reference signal is reached. The location of the reference signal that appears earlier in the sequence can be selected as the placement location of the vibration acceleration sensor in the road noise control system, and this location can be used as the reference signal for system control.

[0123] The formula for selecting the reference signal using the maximum weighted multicoherence coefficient method is as follows:

[0124] (3)

[0125] In the formula: N is the number of in-vehicle control areas. The highest frequency point, The lowest frequency point, This is the weighted value of the coherence coefficient for each region.

[0126] 5) In step S5, in order to solve the problem of high computational burden and slow convergence speed of FxLMS algorithm in broadband noise control, this application adopts an active noise control algorithm of time-frequency domain FxLMS algorithm, which calculates gradient estimation and filter reference signal in frequency domain to reduce computational complexity, and updates control filter coefficients in time domain to minimize delay.

[0127] like Figure 4 As shown, the Short-Time Fourier Transform (STFT) converts a time-domain signal into a frequency-domain signal for analyzing non-stationary signals. In active control algorithms, a perfect reconstruction method is needed to reconstruct the output signal using STFT. Furthermore, a half-cycle sinusoidal window function, with a 50% overlap, can reconstruct the signal relatively accurately. This half-cycle sinusoidal window function can be expressed as:

[0128] (4)

[0129] In the formula: The length of the window function. For elements of the window function.

[0130] The time-frequency domain FxLMS algorithm is implemented as follows:

[0131] Step 1:

[0132] Reference signal Treat it as an extended main block, appending it at its beginning and end. Then divide the extended main block into zeros. There are several overlapping sub-blocks, which are windowed by a half-period sinusoidal window function. Then, all sub-blocks are converted from the time domain to the frequency domain. The calculation formula is as follows:

[0133] (5)

[0134] In the formula: Represented as the first in the frequency domain The first block A windowed master sub-block reference signal group Elements of the extended block, This is represented by the Fast Fourier Transform, which converts a time-domain signal into a frequency-domain signal. Represented as a window function, windowing is applied during signal conversion to prevent signal leakage and distortion.

[0135] Step Two:

[0136] The formula for calculating the filtered frequency domain output signal is:

[0137] (6)

[0138] In the formula: For the first in the frequency domain The first block The filtered output signal of the windowed block For the adaptive filter in the frequency domain The coefficient of each block.

[0139] Step 3:

[0140] The frequency domain filtered output signal needs to be reconstructed into the time domain through an inverse fast Fourier transform, and the calculation formula is as follows:

[0141] (7)

[0142] In the formula: This is represented as the inverse fast Fourier transform.

[0143] The formula for calculating the desired reconstructed filtered output signal is:

[0144] (8)

[0145] In the formula: superscript " " indicates transpose.

[0146] Step Four:

[0147] The residual error signal is represented as:

[0148] (9)

[0149] In the formula: The signal to be canceled. The impulse response of the secondary path in the time domain. For the filtered output signal to be reconstructed, This is represented as a convolution calculation.

[0150] Step 5:

[0151] The residual error signal obtained from the above equation is divided into... The signal is divided into several overlapping sub-blocks. Each sub-block is then windowed and converted from a time-domain signal to a frequency-domain signal. The calculation formula is as follows:

[0152] (10)

[0153] In the formula: It is the first in the frequency domain The first block A windowed residual sub-block, It is a windowed residual sub-block. It is a block element.

[0154] Step Six:

[0155] The adaptive filter weight coefficients are updated as follows:

[0156] (11)

[0157] In the formula: It is expressed as the normalized convergence factor of the frequency domain with a windowed block. For filtering reference signal, This is represented as an error signal.

[0158] The filtered reference signal is represented as:

[0159] (12)

[0160] In the formula: It is represented as the impulse response of the secondary path in the frequency domain.

[0161] III. Method Validation

[0162] To simplify the verification process, the control areas in this embodiment are the driver's seat and the right rear passenger position. The microphone, accelerometer, speaker and other equipment are arranged as described above. Therefore, the target control area in this example is divided into centralized and distributed. In the centralized control method, noise reduction is required for both the driver's seat and the right rear passenger position. Both positions are target control areas, and the coherence coefficients of the target control areas need to be weighted with a weight of 1. In the distributed control method, noise reduction is required for the driver's seat, which is a target control area, while noise reduction is not required for the right rear passenger position, which is a non-target control area. Therefore, the coherence coefficient weight at the driver's position is set to 1, and the coherence coefficient weight at the right rear passenger position is set to 0.

[0163] Figure 5 The results are from a centralized regional theoretical noise reduction calculation, which is based on a centralized equilibrium control mode. Figure 6 The result is the noise reduction calculation result of the distributed region theory, which is the partition target control mode;

[0164] contrast Figure 5 and Figure 6 It can be clearly observed that in the zoned target control mode, the target control area is the driver's seat, and its theoretical noise reduction is relatively high. Figure 4 The centralized equilibrium control mode has been significantly improved.

[0165] In summary:

[0166] 1. This application aims to achieve the best noise reduction effect in the target control area. It selects the optimal combination of reference signals based on the target area inside the vehicle under different driving and riding area scenarios and uses active noise control technology to control the low and medium frequency noise of the car in a zone to achieve the purpose of noise reduction inside the vehicle. This can solve the problem of the high prominence of low and medium frequency road noise in electric vehicles after the loss of the masking effect of engine order noise.

[0167] 2. This application adopts an active noise control algorithm based on the time-frequency domain multi-channel FxLMS algorithm. It calculates gradient estimation and filter reference signal in the frequency domain to reduce computational complexity, and updates control filter coefficients in the time domain to minimize delay. The control algorithm has fast convergence speed, low delay and high stability.

[0168] Those skilled in the art will readily understand that the above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for active control of in-vehicle road noise zoning, characterized by, It comprises the following steps: S1, the PCR radar sensor identifies the seating position of the passenger in the vehicle and determines the target control area and the non-target control area for noise reduction in the vehicle; S2, the target control area and the non-target control area are assigned corresponding weighting factors; S3, under multi-working condition driving conditions, a plurality of vibration acceleration sensors and a plurality of sound pressure sensors are used to collect noise frequency spectrum signals, the noise frequency spectrum signals collected by the vibration acceleration sensors are used as reference signals x(n), and the noise frequency spectrum signals collected by the sound pressure sensors are used as signals to be eliminated d(n); S4, the reference signals x(n) and the signals to be eliminated d(n) are stored and read by a DSP controller, and the best reference signal combination is screened out by using the maximum weighted multiple coherence analysis method in combination with the weighting factors; The best reference signal combination screened out by using the maximum weighted multiple coherence analysis method is calculated according to the following formula: Wherein: N is the number of in-vehicle control areas, is the highest frequency point, is the lowest frequency point, is the weighted value of the coherence coefficient of each area; is the multiple coherence coefficient, and the calculation formula is as follows: wherein: is the theoretical maximum noise reduction, is the reference signal causing the filtered output signal is the auto-power spectrum of the filtered output signal; is the signal to be cancelled is the auto-power spectrum of the signal to be cancelled, is the is the cross-spectrum of the signal to be cancelled and the reference signal is the cross-spectrum of the filtered output signal and the reference signal, is the reference signal is the auto-power spectrum of the reference signal; S5, based on the best reference signal combination screened out, a secondary cancellation signal is generated by using a time-frequency domain multi-channel FxLMS algorithm; S6, the headrest loudspeaker in the target control area generates a corresponding secondary sound source according to the secondary cancellation signal, so as to realize noise reduction in the target area.

2. The in-vehicle road noise zoned active control method according to claim 1, characterized in that In step S1, the target control area and the non-target control area for noise reduction in the vehicle are determined, and the specific process is as follows: It is judged whether there is a person seated at the driving position, if yes, the position is the target control area, and if not, the position is the non-target control area.

3. The in-vehicle road noise zoned active control method of claim 1, wherein In step S2, the corresponding weighting factors are assigned, and the specific process is as follows: The weighting value of the coherence coefficient of the reference signal and the signal to be eliminated in the target control area is set to 1, and the weighting value of the coherence coefficient of the reference signal and the signal to be eliminated in the non-target control area is set to 0.

4. The in-vehicle road noise zoned active control method of claim 1, wherein In step S3, the multi-working condition driving conditions are as follows: A new energy test vehicle with a front suspension of a McPherson independent suspension structure and a rear suspension of a torsion beam non-independent suspension structure is used as a controller test object; The road conditions for vehicle driving are rough asphalt pavement and smooth asphalt pavement, the vehicle speed is adjusted and set to a constant speed cruise mode, and each increase of 10 km / h from 30 to 120 km / h is used as a steady-state test condition; the acceleration mode is full throttle acceleration, which is used as a time-varying non-steady-state test condition.

5. The in-vehicle road noise zoned active control method of claim 1, wherein In step S5, the time-frequency domain multi-channel FxLMS algorithm is used, and the specific process is as follows: Step one: for the reference signal considering it as an extended primary block, appending zeroes at its beginning and end, then dividing the extended primary block into overlapping sub-blocks, windowing these overlapping sub-blocks by a half-cycle sine window function, then converting all sub-blocks from time domain to frequency domain, whose computational formula is as follows: wherein: a k-th windowed primary sub-block reference signal group of a n-th block in the frequency domain, a k-th windowed primary sub-block reference signal group of a n-th block in the frequency domain, a k-th windowed primary sub-block reference signal group of a n-th block in the frequency domain, an element of an extended block, denotes a fast Fourier transform, which converts a time domain signal into a frequency domain signal, denotes a window function, which distorts the signal conversion process in order to prevent signal leakage. Step two: The calculation formula of the filtered frequency domain output signal is as follows: wherein: is a filtered output signal of the th windowed sub-block of the th block in the frequency domain, is a coefficient of the adaptive filter for the th block in the frequency domain; Step three: The frequency domain filtered output signal needs to be reconstructed into the time domain through inverse fast Fourier transform, and the calculation formula is as follows: In the formula: denotes an inverse fast Fourier transform; The calculation formula of the expected reconstructed filtered output signal is as follows: wherein the superscript denotes transposition; Step four: The residual error signal is represented as: wherein: is the signal to be cancelled, is the impulse response of the secondary path in the time domain, is the filtered output signal to be reconstructed, is represented as a convolution calculation; Step five: The residual error signal obtained in step four is divided into overlapping sub-blocks, and then each sub-block is windowed and converted from a time domain signal to a frequency domain signal, with the following formula: wherein: is the element of the th windowed residual subblock of the th block in the frequency domain, is the windowed residual subblock, is the element of the block; Step six: The adaptive filter weight coefficient is updated as follows: wherein: is a normalized convergence factor expressed as a frequency domain windowed sub-block, is a filtered reference signal, is an error signal; The filtered reference signal is represented as: wherein: represents the impulse response of the secondary path in the frequency domain.

6. An in-vehicle road noise zoned active control system characterized by: It comprises a plurality of groups of sound pressure sensors (1) arranged in the vehicle for collecting the road noise signal to be eliminated in the vehicle and a plurality of groups of vibration acceleration sensors (2) arranged outside the vehicle for collecting the noise reference signal outside the vehicle, and the vehicle is also provided with a PCR radar sensor (3) for identifying the seating condition of each seat passenger and determining the target control area, and the sound pressure sensor (1), the vibration acceleration sensor (2) and the PCR radar sensor (3) are all electrically connected with the DSP controller (4); The DSP controller (4) adopts the in-vehicle road noise partition active control method according to any one of claims 1-5, controls the headrest loudspeaker (5) to generate a secondary cancellation signal with the same amplitude and opposite phase as the in-vehicle road noise signal to be eliminated in the target control area, and performs noise reduction control on the target control area in the vehicle.

7. The in-vehicle road noise zoned active control system of claim 6, wherein: The specific arrangement positions of the plurality of groups of sound pressure sensors (1) are: one on the right side of the main driver, one on the left ear of the co-driver, one on the right ear of the left rear seat, and one on the left ear of the right rear seat.

8. The in-vehicle road noise zoned active control system of claim 6, wherein: The specific arrangement positions of the plurality of groups of vibration acceleration sensors (2) are: one on each of the passive ends of the front suspension left and right triangular arm front ends and shock absorber connection points, one on each of the passive ends of the front suspension left and right triangular arm rear ends and shock absorber connection points, one on each of the passive ends of the front suspension left and right shock damper upper ends, one on each of the passive ends of the front and rear wheel hubs, one on the passive end of the front axle subframe middle part, one on each of the passive ends of the left and right body and chassis connecting beams, one on each of the passive ends of the front ends of the rear suspension left and right torsion beams, one on each of the passive ends of the rear ends of the rear suspension left and right torsion beams, one on each of the passive ends of the upper ends of the rear suspension left and right shock dampers, one on each of the passive ends of the rear and front wheel hubs, and one on the passive end of the middle part of the rear suspension torsion beam.

9. A new energy vehicle, characterized in that, An in-vehicle road noise partition active control system according to any one of claims 6-8 is adopted.

Images