Transmission path grading detection and reference sensor selection method based on in-vehicle active noise control

Abstract

The invention discloses a transmission path grading detection and reference sensor selection method based on in-vehicle active noise control, and belongs to the field of active noise control. The method comprises the steps that sensors are arranged at suspension rod pieces of a to-be-tested automobile, and microphones are arranged at seat headrests of the to-be-tested automobile; obtaining an error noise signal collected by the microphone and a reference vibration signal collected by the sensor when the to-be-tested automobile runs under the test working condition; a main control area and a reference vibration signal group are selected, and sensors corresponding to reference vibration signals which are not selected into the reference vibration signal group are removed; according to the method, sensors are arranged at wheel centers, a front auxiliary frame and a rear auxiliary frame of a to-be-detected automobile, error noise signals and reference vibration signals are obtained, L reference vibration signals with the highest multi-coherence are obtained through traversal optimization, and finally active noise control is conducted through a road noise active control system on the basis of the L reference vibration signals. According to the invention, the integrity and the feasibility of transmission path detection are improved, and the test cost is also saved.

Description

Technical Field

[0001] The present invention belongs to the field of active noise control, and particularly relates to a method for hierarchical detection of transmission paths and selection of reference sensors based on in-vehicle active noise control. Background Art

[0002] With the rapid development of the automotive industry, automotive NVH performance has become an important indicator for evaluating automotive comfort, and automotive road noise active control technology has become the current trend in the automotive NVH industry. Road surface noise is one of the main sources of global noise in the vehicle. Many studies and applications at home and abroad have proved that automotive road surface noise is controllable. However, the reference vibration signal arranged on the chassis must be strongly correlated with the controlled noise signal in the vehicle, so as to obtain an accurate prior signal in advance to achieve an effective control effect. However, there are multiple transmission paths from the excitation at the wheel center of the wheel to the human ear to generate noise. Therefore, by detecting the transmission paths and calculating the contribution of the reference vibration signal, and optimizing the best combination from the numerous reference vibration signals at the chassis, is a key step in the development process of the automotive road noise active control system.

[0003] Currently, the method for selecting reference vibration signals is mainly to select a group of reference vibration signals with high coherence through coherence analysis, so as to improve the control effect of the active road noise control system on road noise in the vehicle. However, the general coherence analysis method does not consider the weights between different control seats, nor does it consider the main peak frequencies of the in-vehicle noise.

[0004] In coherence analysis, it is necessary to collect vibration signals at the automotive chassis. The general transmission path detection method requires arranging sensors at all components of the automotive chassis for detection to ensure the integrity of multiple coherence analysis. However, this requires a certain number of vibration sensors and data acquisition with corresponding channel numbers, and the test cost is relatively high. Summary of the Invention

[0005] To solve the problems in the prior art, the present invention provides a method for hierarchical detection of transmission paths and selection of reference sensors based on in-vehicle active noise control.

[0006] The technical solution of the present invention is as follows:

[0007] The present invention discloses a method for hierarchical detection of transmission paths and selection of reference sensors based on in-vehicle active noise control, including the following steps:

[0008] 1) Arrange three-axis vibration acceleration sensors at each suspension rod of the vehicle to be tested, and arrange microphones at the headrests of each vehicle seat; obtain the error noise signal collected by the microphone and the reference vibration signal collected by the three-axis vibration acceleration sensor when the vehicle to be tested is driving under the test conditions;

[0009] 2) Calculate the total A-weighted sound pressure level of each error noise signal within the control frequency band, and set the position corresponding to the error noise signal with the highest total A-weighted sound pressure level as the main control area;

[0010] 3) Based on the error noise signals and reference vibration signals in step 1) and the main control area in step 2), calculate and optimize the weighted average multiple coherence coefficient to obtain a group of reference vibration signals;

[0011] 4) Remove the sensors corresponding to the reference vibration signals not selected into the group of reference vibration signals; then arrange three-axis vibration acceleration sensors at the wheel centers and the front and rear subframes of the vehicle to be tested; obtain all the error noise signals and reference vibration signals of the vehicle to be tested under the test conditions in the current situation;

[0012] 5) Based on the error noise signals, reference vibration signals in step 4) and the main control area in step 2), calculate and optimize the weighted average multiple coherence coefficient again to obtain the L reference vibration signals with the highest multiple coherence. Finally, perform active noise reduction based on the L reference vibration signals using the road noise active control system.

[0013] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0014] 1. The present invention adopts a transfer path hierarchical detection method, which overcomes the disadvantage of the existing detection method that sensors need to be arranged at all key chassis components of the vehicle. While ensuring the integrity of the transfer path, through the pre-coherence analysis of non-critical components (such as multiple links of the multi-link suspension), high-coherence components are selected and added to the second-round complete transfer path detection, significantly reducing the number of sensors in the detection method while ensuring that the selected components contain a complete transfer path and have high coherence, and also reducing the test cost.

[0015] 2. The present invention adopts a weighted average multiple coherence calculation method, which overcomes the disadvantage of the current reference sensor selection method that does not consider the amplitude size of the error noise signals at different seats in the vehicle. When selecting the reference vibration signals, weighting is performed according to the amplitude size of the error noise signals at different seats, so that the selected reference vibration signals can strengthen the control effect in the seat area where the amplitude of the error noise signal is larger after being input into the active noise control system.

[0016] 3. The present invention adopts a weighted average multiple coherence calculation method, which overcomes the disadvantage of the current reference sensor selection method that does not consider the spectral characteristics of the error noise signals in the vehicle. When selecting the reference vibration signals, weighting is performed on the frequency peaks of the error noise signals, so that after the selected reference vibration signals are input into the active noise control system, the system can perform targeted control on the high-peak frequency in the error noise signals. Description of the Drawings

[0017] Figure 1 This is a schematic diagram of the global noise transmission path of the vehicle of the present invention;

[0018] Figure 2 This is a schematic diagram of the process of the method for optimizing the reference vibration signal of the vehicle of the present invention;

[0019] Figure 3 This is the active noise reduction effect diagram of the left front seat after selecting the reference vibration signal of the present invention.

[0020] Figure 4 This is the active noise reduction effect diagram of the right front seat after selecting the reference vibration signal of the present invention.

[0021] Figure 5 This is the active noise reduction effect diagram of the left rear seat after selecting the reference vibration signal of the present invention.

[0022] Figure 6 This is the active noise reduction effect diagram of the right rear seat after selecting the reference vibration signal of the present invention. Detailed implementation manners

[0023] The present invention will be further described and explained below in conjunction with the detailed implementation manners. The described embodiments are only examples of the disclosed content and do not delimit the scope of limitation. Without conflict, the technical features of each embodiment of the present invention can be combined accordingly.

[0024] The purpose of the present invention is to analyze the global noise transmission path of the vehicle, select a combination of reference vibration signals for active road noise control, reduce costs, and ensure that the input combination of reference vibration signals for active road noise control can have a good control effect while ensuring the integrity and accuracy of the method.

[0025] Therefore, the present invention first provides a method for hierarchical detection of the transmission path and selection of reference sensors based on in-vehicle active noise control. In order to reduce the test cost and improve the feasibility, the transmission paths corresponding to the vehicle seats and different components of the vehicle chassis are hierarchically detected multiple times by region, and the contribution amount is calculated and the combination of reference vibration signals is optimized to obtain a three-axis vibration acceleration sensor layout scheme with a complete global noise transmission path of the vehicle and a combination of reference vibration signals for active road noise control of the vehicle.

[0026] The method for hierarchical detection of the transmission path and selection of reference sensors based on in-vehicle active noise control of the present invention includes the following steps:

[0027] 1) Arrange triaxial vibration acceleration sensors at each suspension rod of the vehicle to be tested, and arrange microphones at the headrests of each vehicle seat; obtain the error noise signals collected by the microphones and the reference vibration signals collected by the triaxial vibration acceleration sensors when the vehicle to be tested is driving under test conditions;

[0028] 2) Calculate the total A-weighted sound pressure level of each error noise signal within the control frequency band, and set the position corresponding to the error noise signal with the highest total A-weighted sound pressure level as the main control area;

[0029] 3) Based on the error noise signals and reference vibration signals in step 1) and the main control area in step 2), calculate and optimize the weighted average multiple coherence coefficient to obtain a group of reference vibration signals;

[0030] 4) Remove the sensors corresponding to the reference vibration signals not selected into the group of reference vibration signals; then arrange triaxial vibration acceleration sensors at the wheel centers and the front and rear subframes of the vehicle to be tested; obtain all the error noise signals and reference vibration signals of the vehicle to be tested under test conditions in the current situation;

[0031] 5) Based on the error noise signals, reference vibration signals in step 4) and the main control area in step 2), calculate and optimize the weighted average multiple coherence coefficient again to obtain L reference vibration signals with the highest multiple coherence, and finally perform active noise reduction using the road noise active control system based on the L reference vibration signals.

[0032] In a specific embodiment of the present invention, step 1) is specifically:

[0033] Arrange triaxial vibration acceleration sensors N1, N2, N3, N4 respectively at all suspension rods at the left front, right front, left rear, and right rear of the vehicle chassis. N1 represents that there are N1 suspension rods at the left front of the vehicle chassis, N2 represents that there are N2 suspension rods at the right front of the vehicle chassis, N3 represents that there are N3 suspension rods at the left rear of the vehicle chassis, and N4 represents that there are N4 suspension rods at the right rear of the vehicle chassis; that is, arrange a triaxial vibration acceleration sensor at each suspension rod of the vehicle chassis. Arrange a microphone at the middle position of the headrest of the vehicle seat to be controlled, and arrange one for each headrest of the vehicle seat, with a total of M W pieces.

[0034] As a preferred solution of the present invention, step 2) includes:

[0035] The low-frequency cut-off frequency of the control frequency band is f u and the high-frequency cut-off frequency is f d , calculate the error noise signals collected at different positions respectively at the frequency f u to f dThe total A-weighted sound pressure level within the control frequency band. According to the calculation results, set the position corresponding to the error noise signal with a higher total A-weighted sound pressure level as the main control area, that is, the main control area, and the serial number of this error noise signal is m σ .

[0036] Calculate the total A-weighted sound pressure level of each error noise signal within the control frequency band, that is, within the frequency band from frequency f u to f d . Calculate the A-weighted sound pressure level of each error noise signal at each frequency respectively, and then sum up all the A-weighted sound pressure levels of this error noise signal to obtain the total A-weighted sound pressure level of this error noise signal.

[0037] In a preferred embodiment of the present invention, step 3) includes:

[0038] 3.1) First, calculate the weighted average multiple coherence coefficient between all the error noise signals in step 1) and each reference vibration signal in step 1) within the control frequency band, and use the reference vibration signal corresponding to the maximum weighted average multiple coherence coefficient as the selected reference vibration signal and add it to the currently empty reference vibration signal group;

[0039] 3.2) Respectively form several first signal groups with the unselected reference vibration signals and the reference vibration signals in the current reference vibration signal group, then calculate the weighted average multiple coherence coefficient between all the error noise signals and each first signal group within the control frequency band, and obtain the first signal group corresponding to the maximum weighted average multiple coherence coefficient, and use this first signal group as the updated reference vibration signal group;

[0040] 3.3) Repeat step 3.2) until the number of reference vibration signals in the obtained reference vibration signal group reaches the preset value.

[0041] As a preferred solution of the present invention, the calculation method of the weighted average multiple coherence coefficient includes:

[0042] 1.1.1) At frequency f, calculate the cross-power spectral density matrix S xm (f) between each error noise signal and the reference vibration signal / reference vibration signal group, and obtain the conjugate transpose of each cross-power spectral density matrix S xm (f) is the cross-power spectral density matrix between the mth error noise signal and the reference vibration signal group x at frequency f. Among them, when calculating the weighted average multiple coherence coefficient in step 3.1), x is the reference vibration signal; when calculating the weighted average multiple coherence coefficient in step 3.2), x is the reference vibration signal group.

[0043] 1.1.2) Calculate the generalized inverse matrix of the auto-power spectral density matrix of the reference vibration signal x or the reference vibration signal group x itself at the frequency f And calculate the auto-power spectral density S of each error noise signal at the frequency f mm (f).

[0044] 1.1.3) Calculate the matrix multiplication To obtain the multiple coherence coefficient

[0045] 1.1.4) Confirm the control region weight coefficient ε(m). When m = m σ Then ε(m) = 1. When m ≠ m σ Then ε(m) = 0.8.

[0046] 1.1.5) Calculate the frequency weighting coefficient λ(f). Calculate That is, the auto-power spectral density of the m-th error noise signal under A-weighting. Calculate That is, the maximum peak frequency f θ At the sound pressure level of the m-th error noise signal.

[0047] 1.1.6) Calculate the weighted average multiple coherence coefficient That is, the weighted average multiple coherence coefficient between all M W Error noise signals and the reference vibration signal group x within the control frequency band.

[0048] In a specific embodiment of the present invention, in step 4), a three-axis vibration acceleration sensor is arranged at each wheel center, and two three-axis vibration acceleration sensors are arranged at the front and rear subframes of the vehicle, and the two three-axis vibration acceleration sensors on the same subframe are evenly arranged.

[0049] As a preferred solution of the present invention, in step 5), obtain the L reference vibration signals with the highest multiple coherence, including:

[0050] 5.1) First, calculate the weighted average multiple coherence coefficient between all the error noise signals in step 4) and each reference vibration signal in step 4) within the control frequency band, and use the reference vibration signal corresponding to the maximum weighted average multiple coherence coefficient as the selected reference vibration signal, construct an initially empty reference group, and add the selected reference vibration signal to the reference group;

[0051] 5.2) The unselected reference vibration signals are respectively combined with the reference vibration signals in the current reference group to form a number of second signal groups. Then, the weighted average multiple coherence coefficient between all the error noise signals and each second signal group within the control frequency band is calculated, and the second signal group corresponding to the maximum weighted average multiple coherence coefficient is obtained. Then, this second signal group is used as the updated reference group.

[0052] 5.3) Repeat step 5.2) until the number of reference vibration signals in the obtained reference group reaches the preset value L. The L reference vibration signals in the finally obtained reference group are the L reference vibration signals with the highest multiple coherence.

[0053] As a preferred solution of the present invention, when calculating the weighted average multiple coherence coefficient in step 5), the reference vibration signal group x in steps 1.1.1)-1.1.6) is replaced with the reference group in step 5) for calculation. Similarly, when calculating the weighted average multiple coherence coefficient in step 5.1), x is the reference vibration signal; when calculating the weighted average multiple coherence coefficient in step 5.2), x is the reference group.

[0054] Embodiment 1

[0055] To make the present invention more specific, this embodiment provides a method for hierarchical detection of transmission paths and selection of reference sensors based on in-vehicle active noise control.

[0056] In this embodiment, the chassis transmission path diagram corresponding to the vehicle's global noise is as shown in the appendix Figure 1 As shown. By analyzing the vehicle's global noise transmission path, due to the damping vibration of the shock absorber and the spring itself along the spring direction, the coherence between the global noise transmitted from this transmission path and the transmission path is not high. Therefore, the wheel center - suspension link - subframe is selected as the detected transmission path and the installation positions of the three-axis vibration acceleration sensors.

[0057] Step S1: A total of 12 three-axis vibration acceleration sensors are arranged in four link areas of the vehicle's left front, right front, left rear, and right rear. Their installation positions and the corresponding directions of the three-axis vibration acceleration sensors are shown in Table 1. Microphones are respectively arranged in the middle of the four seat headrests of the vehicle to collect the reference vibration signals at the chassis and the error noise signals

[0058] Table 1

[0059]

[0060]

[0061] Step S2: Calculate the auto-power spectral density of the error noise signals at four positions and the A-weighted sound pressure level at each frequency within the frequency bands of f d = 30 Hz, f u = 400 Hz, obtaining the spectrum of each seat and the total A-weighted sound pressure level from 30 - 400 Hz. The results of the total A-weighted sound pressure levels from 30 - 400 Hz at the four seats are shown in Table 2. Therefore, the right rear seat is selected as the main control area, m σ = 4.

[0062] Table 2

[0063] Seat Left front (1) Right front (2) Left rear (3) Right rear (4) A-weighted sound pressure level (dBA) 54.95 55.57 55.43 56.74

[0064] Step S3: Calculate the multiple coherence coefficients at the four seats

[0065]

[0066] where S xm (f) is the cross-power spectral density matrix between the m-th error noise signal and the reference vibration signal group x at frequency f, is the generalized inverse matrix of the auto-power spectral density matrix of the reference vibration signal group x itself at frequency f, is the conjugate transpose of the cross-power spectral density matrix S xm (f), and S mm (f) is the auto-power spectral density of the m-th error noise signal at frequency f.

[0067] Then calculate the average multiple coherence coefficient

[0068]

[0069] where is the control area weight coefficient, m σ is the serial number of the error noise signal in the specified main control area, is the frequency weighting coefficient, is the auto-power spectral density of the m-th error noise signal under A-weighting, is the sound pressure level of the m-th error noise signal at the maximum peak frequency f θ . And perform sensor optimization according to Appendix Figure 2 to obtain the reference vibration signal group Select the four best reference vibration signals as the X direction of the right front suspension traction arm, the X direction of the left front suspension rear control arm, the Y direction of the left rear suspension rear control arm, and the X direction of the right rear suspension traction arm.

[0070] Step S4: retain the four three-way vibration acceleration sensors of the right front suspension traction arm, the left front suspension rear control arm, the left rear suspension rear control arm, and the right rear suspension traction arm, and remove the other sensors and attach them to the wheel center and the front and rear subframes respectively. The specific layout positions and sensor channel directions are shown in Table 3. Collect the reference vibration signal and error noise signal

[0071] Table 3

[0072]

[0073]

[0074] Step S5: Repeat step S3 to optimize all reference vibration signals and finally obtain the ranking of all reference vibration signals with the complete transfer path of the car as shown in Table 4.

[0075] Table 4 Sorting table of reference vibration signals with complete transfer path

[0076] Sensor arrangement position Sensor direction Sensor arrangement position Sensor direction Right end of front subframe X Left rear wheel center Z Right front suspension trailing arm X Right front suspension trailing arm Y Right end of rear subframe X Right front suspension trailing arm Y Left end of rear subframe X Left front wheel center Z Left end of front subframe Z Left front suspension rear control arm X Left end of rear subframe Y Right end of front subframe Y Right end of rear subframe Y Right rear wheel center X Right front wheel center Y Left rear wheel center X Left end of front subframe Y Right rear wheel center Z Right front suspension trailing arm X Right front wheel center X Left rear wheel center Y Right rear wheel center Y Left front wheel center Y Left rear suspension rear control arm Z Right end of front subframe Z Right front wheel center Z Right front suspension trailing arm Z Left rear suspension rear control arm Y Left end of rear subframe Z Left rear suspension rear control arm X Left front wheel center X Left end of front subframe X Right front suspension trailing arm Z Left front suspension rear control arm Y Right end of rear subframe Z Left front suspension rear control arm Z

[0077] Step S6: The first 12 channels (i.e., L=12 here) of the reference vibration signal are introduced into the road noise active control system for simulation, and the following is obtained: Figures 3 - 6 The road noise active control effect on a rough road surface (30km / h working condition) is shown. In the actual application of active noise reduction using the road noise active control system, the input of the road noise active control system is L randomly selected reference vibration signals, so the L reference vibration signals with the highest multi-coherence obtained by the present invention can be directly input into the road noise active control system for active noise reduction, and the control effect of the L reference vibration signals with the highest multi-coherence of the present invention is better than that of the L randomly selected reference vibration signals.

[0078] Using the general method to select the reference vibration signal requires at least 24 three-axis vibration acceleration sensors and 76-channel data acquisition, while in this embodiment, only 12 three-axis vibration acceleration sensors and 40-channel data acquisition are ultimately required.

[0079] In summary, the transfer path hierarchical detection and reference sensor selection method based on active noise control in the vehicle provided by the present invention has the following beneficial effects: the transfer path hierarchical detection is adopted to improve the integrity and feasibility of the transfer path detection, and at the same time, the test cost is significantly saved. Considering the weight of the seats in the vehicle and the weight of the spectrum peak, the accuracy and pertinence of the contribution calculation method are significantly improved.

[0080] The above-described embodiments merely represent several implementation manners of the present invention. The description thereof is relatively specific and detailed, but it should not be construed as a limitation on the scope of the patent of the present invention. For those of ordinary skill in the art, without departing from the concept of the present invention, several modifications and improvements can still be made, and these all fall within the protection scope of the present invention.

Claims

1. A method for hierarchical detection of transfer paths and selection of reference sensors based on active noise control in a vehicle, characterized in that: The following steps are involved: 1) A three-dimensional vibration acceleration sensor is arranged at each suspension rod of the vehicle to be tested, and a microphone is arranged at each vehicle seat headrest; an error noise signal collected by the microphone and a reference vibration signal collected by the three-dimensional vibration acceleration sensor are obtained when the vehicle to be tested is running under the test condition; 2) Calculate the total A-weighted sound pressure level of each error noise signal in the control frequency band, and set the position corresponding to the error noise signal with the highest total A-weighted sound pressure level as the main control area; 3) Based on the error noise signal and the reference vibration signal of step 1) and the main control area of ​​step 2), weighted average multiple coherence coefficients are calculated and optimized to obtain a reference vibration signal group; 4) removing the sensor corresponding to the reference vibration signal that is not selected into the reference vibration signal group; Then, three-way vibration acceleration sensors are arranged at the wheel centers and front and rear subframes of the vehicle to be tested; all error noise signals and reference vibration signals of the vehicle to be tested under the test conditions are obtained; 5) Based on the error noise signal of step 4), the reference vibration signal and the main control area of ​​step 2), weighted average multiple coherence coefficients are calculated and optimized to obtain L reference vibration signals with the highest multiple coherence. Finally, active noise reduction is performed in the road noise active control system based on the L reference vibration signals.

2. The method for transfer path hierarchical detection and reference sensor selection based on in-vehicle active noise control according to claim 1, characterized in that: In step 1), a three-dimensional vibration acceleration sensor is arranged at each suspension rod; a microphone is arranged at each car seat headrest, and each microphone is arranged at the center of the headrest.

3. The method for transfer path hierarchical detection and reference sensor selection based on in-vehicle active noise control according to claim 1, characterized in that: In step 2), the low frequency cutoff frequency of the control frequency band is f u , the high frequency cutoff frequency is f d .

4. The method for transfer path hierarchical detection and reference sensor selection based on in-vehicle active noise control according to claim 1, characterized in that: Step 3) includes: 3.1) First, calculate the weighted average multi-coherence coefficient between all error noise signals in step 1) in the control frequency band and each reference vibration signal in step 1), and use the reference vibration signal corresponding to the largest weighted average multi-coherence coefficient as the selected reference vibration signal and add it to the currently empty reference vibration signal group; 3.2) The unselected reference vibration signals are respectively combined with the reference vibration signals in the current reference vibration signal group to form a plurality of first signal groups, and then the weighted average multi-correlation coefficients between all error noise signals and each first signal group in the control frequency band are calculated, and the first signal group corresponding to the largest weighted average multi-correlation coefficient is obtained, and the first signal group is used as the updated reference vibration signal group; 3.3) Repeat step 3.2) until the number of reference vibration signals in the obtained reference vibration signal group reaches a preset value.

5. The method for transfer path hierarchical detection and reference sensor selection based on in-vehicle active noise control according to claim 4, characterized in that: The calculation method of the weighted average multi-coherence coefficient includes: a) At frequency f, calculate the cross-power spectral density matrix S between each error noise signal and the reference vibration signal group xm (f) and obtain the conjugate transpose of each cross power spectral density matrix Among them, S xm (f) represents the cross power spectral density matrix between the mth error noise signal and the reference vibration signal group x at frequency f; b) Calculate the generalized inverse matrix of the power spectrum density matrix of the reference vibration signal group x itself at frequency f And calculate the autopower spectral density S of each error noise signal at frequency f mm (f); c) Calculate the multi-coherence coefficient Then set the control area weight coefficient ε(m), that is, when m=m σ When ε(m)=1; when m≠m σ When ε(m)=0.8, m σ is the serial number of the error noise signal with the highest total value of A-weighted sound pressure level in step 2); d) Calculate the autopower spectral density of the mth error noise signal under A weighting and the maximum peak frequency f θ The sound pressure level of the mth error noise signal at Then calculate the frequency weighting coefficient λ(f); e) Calculate the weighted multi-correlation coefficients between all error noise signals in the control frequency band and the reference vibration signal group x 6. The method for transfer path hierarchical detection and reference sensor selection based on in-vehicle active noise control according to claim 5, characterized in that: In step c), the multiple coherence coefficient The calculation formula is:

7. The method for transfer path hierarchical detection and reference sensor selection based on in-vehicle active noise control according to claim 5, characterized in that: In step d), the frequency weighting coefficient λ(f) is calculated as:

8. The method for transfer path hierarchical detection and reference sensor selection based on in-vehicle active noise control according to claim 5, characterized in that: In step e), weighted multi-coherence coefficients The calculation formula is: Among them, M W is the total number of error noise signals; f u is the low frequency cutoff frequency of the control band, f d To control the high frequency cutoff frequency of the frequency band.

9. The method for transfer path hierarchical detection and reference sensor selection based on in-vehicle active noise control according to claim 1, characterized in that: In step 4), a three-axis vibration acceleration sensor is arranged at each wheel center, two three-axis vibration acceleration sensors are arranged at the front and rear subframes of the vehicle, and the two three-axis vibration acceleration sensors on the front subframe or the rear subframe are evenly arranged.

10. The method for transfer path hierarchical detection and reference sensor selection based on in-vehicle active noise control according to claim 1, characterized in that: In step 5), obtaining L reference vibration signals with the highest multi-coherence includes: 5.1) First, calculate the weighted average multiple coherence coefficients between all error noise signals in step 4) within the control frequency band and each reference vibration signal in step 4), and use the reference vibration signal corresponding to the largest weighted average multiple coherence coefficient as the selected reference vibration signal, construct a currently empty reference group, and add the selected reference vibration signal to the reference group; 5.2) The unselected reference vibration signals are respectively combined with the reference vibration signals in the current reference group to form several second signal groups, and then the weighted average multi-correlation coefficients between all error noise signals in the control frequency band and each second signal group are calculated, and the second signal group corresponding to the largest weighted average multi-correlation coefficient is obtained, and the second signal group is used as the updated reference group; 5.3) Repeat step 5.2) until the number of reference vibration signals in the reference group obtained reaches a preset value L, and the L reference vibration signals in the reference group finally obtained are the L reference vibration signals with the highest multi-coherence.

Images