A wind speed measurement system and method fusing acoustic and inertial perception

Abstract

The application discloses a kind of fusion acoustic wave and inertial perception's wind speed measuring system and method, method is suitable for including inertial sensor and two groups of audio transceiver device mobile intelligent terminal, comprising the following steps: according to the initial pose of terminal obtained by inertial sensor, and the first component of wind speed is collected by audio device;Rotating terminal multiple times, the rotation angle of terminal compared with initial pose and the component corresponding to wind speed are calculated at the end of each rotation;Wind speed value is calculated and wind direction is determined by simultaneously solving each component and corresponding rotation angle.The application can effectively measure wind speed without the help of additional devices and without any modification to the hardware of the terminal.

Description

TECHNICAL FIELD

[0001] The present application belongs to the field of mobile sensing technology, and particularly relates to a wind speed measurement system and method fusing acoustic wave and inertial sensing. BACKGROUND

[0002] With the rapid development of intelligent terminal devices, more and more applications are developed, which greatly enrich people's life and provide a lot of convenience, including the measurement of wind speed. Wind speed sensor has a wide application prospect in weather monitoring, smart home system and other fields.

[0003] At present, the schemes for measuring wind speed based on intelligent terminal mainly include the following two kinds:

[0004] The first kind is to use additional hardware, such as using a commercial propeller anemometer to measure, and then sending the measured wind speed information to the intelligent terminal through wired or wireless communication. Or by modifying the mainboard circuit of the intelligent terminal, the wind speed detector module is integrated into the mainboard to realize the function of wind speed detection. This kind of scheme depends on external hardware, which not only increases the cost, but also is difficult to integrate into the intelligent terminal conveniently.

[0005] The second kind is a software-based method, which estimates the actual airflow speed through the difference between the output electrical signals of the two microphones at the bottom and top of the intelligent terminal, and collects multiple sets of double microphone signals during the rotation of the intelligent terminal. According to the maximum value of the difference and the corresponding direction of the intelligent terminal, the actual wind speed and direction are determined. This passive measurement method has obvious disadvantages in actual use. First, external sound will seriously affect the size of the microphone output electrical signal, and the sound intensity of the near sound source point and the sound intensity of the far sound source point usually also have differences, which may lead to the inability to effectively measure the wind speed in a noisy environment. Second, the rotation of the intelligent terminal will affect the wind field, and there is an unpredictable error in the collected signals during the rotation process, and this method also fails to effectively utilize the collected multiple signals for optimization.

[0006] Therefore, based on the above considerations, it is necessary to propose an innovative wind speed measurement system and method, which can directly measure the wind speed accurately and stably on the intelligent terminal without the aid of additional hardware. SUMMARY

[0007] In view of the above shortcomings of the prior art, the purpose of the present application is to provide a wind speed measurement system and method fusing acoustic wave and inertial sensing, so as to solve the problems that the existing wind speed measurement technology needs to rely on additional hardware, and cannot effectively adapt to the actual noisy environment and affect the wind field.

[0008] To achieve the above purpose, the technical scheme adopted by the present application is as follows:

[0009] The wind speed measurement system fusing acoustic wave and inertial sensing comprises an audio module, an inertial sensor, a processor and a memory.

[0010] The audio module is used for emitting and collecting acoustic wave signals, and comprises two sets of audio transceivers, i.e., a top earpiece / speaker and a microphone, and a bottom speaker and a microphone.

[0011] The inertial sensor is used for collecting the pose and rotation data of the intelligent terminal, and comprises an accelerometer, a gyroscope and a magnetometer.

[0012] The processor is used for controlling the audio module to emit and collect acoustic wave signals, and controlling the data collection of the inertial sensor, and processing the collected acoustic wave signal data and sensor data.

[0013] The memory is used for storing the collected acoustic wave signal data and sensor data, and the calculation result data of the processor.

[0014] Further, the layout structure of the two sets of audio transceivers should satisfy that the included angle between the connection line from the top earpiece / speaker to the bottom microphone and the connection line from the bottom speaker to the top microphone is within 30°.

[0015] The wind speed measurement method fusing acoustic wave and inertial sensing comprises the following steps based on the above system.

[0016] 1) After calibrating the magnetometer, the intelligent terminal is placed on a horizontal plane, and the initial pose of the intelligent terminal is determined based on the inertial sensor.

[0017] 2) The top earpiece / speaker and the bottom speaker periodically and alternately emit modulated acoustic wave signals, and the top microphone and the bottom microphone collect signals.

[0018] 3) The first component of the wind speed at the initial pose is calculated according to the two-channel signals collected by the two microphones.

[0019] 4) The intelligent terminal is rotated by a certain angle on the horizontal plane, and the emission and collection of acoustic waves are closed during the rotation.

[0020] 5) After detecting that the intelligent terminal is stationary, the rotation angle of the intelligent terminal compared with the initial pose is calculated based on the data of the inertial sensor, and the emission and collection of acoustic waves are started.

[0021] 6) The second component of the wind speed is calculated according to the two-channel signals collected by the two microphones.

[0022] 7) Steps 4) to 6) are repeated to obtain the remaining N-2 wind speed components, and N≥2.

[0023] 8) Calculate the horizontal wind speed value according to the N wind speed components and the corresponding rotation angles, and determine the wind direction according to the initial pose of the intelligent terminal.

[0024] Further, the specific method for determining the initial pose of the intelligent terminal based on the inertial sensor in step 1) is as follows:

[0025] 11) Define the three axes X g ,Y g ,Z g of the world coordinate system as the east, north, and sky directions, respectively, and the three axes X l ,Y l ,Z l of the intelligent terminal coordinate system as the right along the screen, up along the screen, and out of the screen, respectively; the vector measured by the accelerometer of the intelligent terminal is a=[a x ,a y ,a z ], and the vector measured by the magnetometer is m=[m x ,m y ,m z ], then the unit vector h along the eastward X g axis and the unit vector l along the northward Y g axis are obtained:

[0026]

[0027] where h x ,h y ,h z are the component values of the vector h along the three axes of the intelligent terminal coordinate system, respectively, and l x ,l y ,l z are the component values of the vector l along the three axes of the intelligent terminal coordinate system, respectively.

[0028] Further, the rotation matrix R from the world coordinate system to the intelligent terminal coordinate system is obtained:

[0029] R=[h,l,r] T

[0030] where r=[r x ,r y ,r z ]=a / |a|, and |a| is equal to the magnitude of the gravitational acceleration in the stationary state;

[0031] 12) Assume that the initial pose of the intelligent terminal in the world coordinate system is: rotated by an angle θ0 around the Z g axis, rotated by an angle around the Y g axis, and rotated by an angle gThe axis is rotated by an angle γ0; then, based on the rotation matrix, the following can be obtained:

[0032]

[0033] Furthermore, the acoustic signal emitted in step 2) is a frequency-modulated continuous wave with a signal period length of T. s The signal collected by the microphone is a mixed signal, including the direct signal from the earpiece / speaker, the signal reflected by nearby objects, and ambient noise.

[0034] Further, in steps 3), 6), and 7), the calculation of the i-th component of the wind speed, i∈[1,N], specifically involves:

[0035] S1) Process the dual-channel signal acquired by the microphone and calculate the propagation delay T of the sound wave from the bottom speaker to the top microphone. i1 And the propagation delay T of sound waves from the top earpiece / speaker to the bottom microphone. i2 ;

[0036] S2) Let V s W is the speed at which sound waves travel in air. i Let D be the i-th component of the wind speed, D1 be the distance from the bottom speaker to the top microphone, and D2 be the distance from the top earpiece / speaker to the bottom microphone. Then:

[0037]

[0038] The wind speed component W can then be obtained. i :

[0039]

[0040] Further, the calculation of the propagation delay of the sound wave from the bottom speaker to the top microphone in step S1) specifically involves:

[0041] S11) Bandpass filtering is performed on the dual-channel signal acquired by the microphone to remove ambient noise from the signal;

[0042] S12) The filtered signal is cross-correlated with the transmitted signal to obtain the processing result Xcorr from the bottom microphone. b And the processing results of the top microphone Xcorr t ;

[0043] S13) The sequence Xcorr b With the period length T of the transmitted signal s Peak detection was performed at intervals to obtain peak values ​​PV1, PV2, ..., PV. K , and then take the former Corresponding subscript of large peak value

[0044] S14) Extracting subsequence with subscript in [P b -μT t ·F j , P s +3μT s ·F j ] from sequence Xcorr s and Xcorr s respectively, extracting upper envelope of subsequence and normalizing, then X times up-interpolation is performed on normalized upper envelope curve, sequence and are obtained respectively, where μ∈(0,0.2), F s is system sampling rate,

[0045] S15) Obtaining maximum peak subscript in sequence and peak subscript closest to in sequence then:

[0046]

[0047] Wherein, the method for calculating the propagation time delay of the sound wave from the top microphone / loudspeaker to the bottom microphone in the step S1) is consistent with the method for calculating the propagation time delay of the sound wave from the bottom loudspeaker to the top microphone, which is not described here.

[0048] Further, the basic principle of the rotation of the intelligent terminal in the step 4) is to rotate clockwise or counterclockwise in the horizontal plane, and the rotation angle σ ∈ (-π, π).

[0049] Further, the rotation angle of the intelligent terminal compared with the initial pose in the step 5) is specifically:

[0050] 51) Calculate the rotation angle σ around the Z g axis according to the gyroscope data in the current rotation process;

[0051] 52) Calculate the current pose of the intelligent terminal through the accelerometer data and the magnetometer data when the rotation is over and the intelligent terminal is stationary, and obtain the rotation angle θ around the Z g axis of the intelligent terminal;

[0052] 53) Record the rotation angle β i-1 of the intelligent terminal compared with the initial pose at the last time when the intelligent terminal is stationary, and the rotation angle β i of the intelligent terminal compared with the initial pose at the current time when the intelligent terminal is stationary, i ≥ 1, and let ω = θ0+ β i-1+σ, then:

[0053]

[0054] wherein, initially, β0=0; ρ1 and ρ2 are weight factors, ρ1+ρ2=1.

[0055] Further, the method for calculating the wind speed value and determining the wind direction in the step 8) is specifically:

[0056] 81) record the i-th component of the wind speed and the rotation angle of the corresponding intelligent terminal compared with the initial pose as (W i ,β i ), i∈[1,N];

[0057] 82) assuming that the horizontal wind speed value is W, and the initial pose of the intelligent terminal and the wind direction on the horizontal plane is α, obtain:

[0058] W i = W·cos(β i -α);

[0059] Considering the interference in the actual situation and the measurement error, the wind speed value W and the angle α are further optimized based on the least square method:

[0060]

[0061] 83) further considering the initial pose θ0 of the intelligent terminal in the world coordinate system based on the angle α, i.e., the horizontal rotation angle around the Z g axis of the world coordinate system, to obtain the wind direction (θ0+α) in the world coordinate system.

[0062] Advantages of the present application:

[0063] 1. High robustness: the present application actively measures the wind speed by using the modulated acoustic wave signal, which can effectively reduce the interference of external noise and adapt to various actual environments.

[0064] 2. High-precision wind speed measurement: the method of the present application does not affect the wind field; the interpolation method can effectively improve the measurement accuracy, and the least square fitting method based on multiple wind speed components can effectively remove abnormal points, so that the calculation result is more accurate.

[0065] 3. Low cost and easy to expand: the present application does not need to rely on additional hardware, and most of the current common mobile intelligent terminals meet the device and layout conditions required by the system; the present application can be easily expanded to measure three-dimensional wind speed; in addition, the propagation speed of the acoustic wave in the air can be directly calculated, and the temperature evaluation function can be further expanded based on the relationship between the temperature and the acoustic speed. BRIEF DESCRIPTION OF DRAWINGS

[0066] Figure 1 System architecture diagram of the present application;

[0067] Figure 2 Flow chart of the method of the present application;

[0068] Figure 3 Schematic diagram of measuring each component of the rotating terminal;

[0069] Figure 4 Schematic diagram of the key propagation path of the sound wave;

[0070] Figure 5 Schematic diagram of calculating the sound wave propagation time delay;

[0071] Figure 6 Schematic diagram of optimizing the calculation result by using the least square method;

[0072] Figure 7 Schematic diagram of calculating the actual wind direction. DETAILED DESCRIPTION

[0073] In order to facilitate the understanding of those skilled in the art, the present application will be further described below in conjunction with the embodiments and the accompanying drawings. The content mentioned in the embodiments is not a limitation of the present application.

[0074] Referring to Figure 1 The wind speed measurement system of the present application fuses sound wave and inertial perception, comprising: an audio module, an inertial sensor, a processor and a memory; wherein,

[0075] The audio module is used for emitting and collecting sound wave signals, which contains two groups of audio transceivers, i.e. the top earpiece / speaker and microphone, and the bottom speaker and microphone.

[0076] The inertial sensor is used for collecting the pose and rotation data of the intelligent terminal (such as a smart phone), which contains an accelerometer, a gyroscope and a magnetometer.

[0077] The processor is used for controlling the audio module to emit and collect sound wave signals, and controlling the data collection of the inertial sensor, and processing the collected sound wave signal data and sensor data.

[0078] The memory is used for storing the collected sound wave signal data and sensor data, and the calculation result data of the processor.

[0079] The layout structure of the two groups of audio transceivers should meet the following requirements: the included angle between the connection line from the top earpiece / speaker to the bottom microphone and the connection line from the bottom speaker to the top microphone is within 30°, and the smaller the included angle, the more accurate the calculated wind speed component.

[0080] Referring toFigure 2 、 Figure 3 The wind speed measurement method of the fusion of acoustic waves and inertial sensing according to the application is based on the system and comprises the following steps:

[0081] 1) After calibrating the magnetometer, the intelligent terminal is placed on a horizontal plane, and the initial pose of the intelligent terminal is determined based on the inertial sensor;

[0082] The specific method for determining the initial pose of the intelligent terminal based on the inertial sensor is as follows:

[0083] 11) Define the three axes X g ,Y g ,Z g of the world coordinate system as the east, north and sky directions respectively, and the three axes X l ,Y l ,Z l of the intelligent terminal coordinate system as right along the screen, up along the screen and outward perpendicular to the screen respectively; the vector measured by the accelerometer of the intelligent terminal is a = [a x ,a y ,a z ], and the vector measured by the magnetometer is m = [m x ,m y ,m z ], so as to obtain the unit vector h along the eastward X g axis and the unit vector l along the northward Y g axis:

[0084]

[0085] wherein h x ,h y ,h z are the component values of the vector h along the three axes of the intelligent terminal coordinate system respectively; l x ,l y ,l z are the component values of the vector l along the three axes of the intelligent terminal coordinate system respectively;

[0086] Further, the rotation matrix R from the world coordinate system to the intelligent terminal coordinate system is obtained:

[0087] R = [h, l, r] T

[0088] wherein r = [r x ,r y ,r z ] = a / |a|, and |a| is equal to the magnitude of the gravitational acceleration in the stationary state;

[0089] 12) Set the initial pose of the intelligent terminal in the world coordinate system as: around Z gThe rotation angle of the axis θ0, around Y g The rotation angle of the axis around X g The rotation angle of the axis γ0; then according to the rotation matrix, we get:

[0090]

[0091] 2) The top earpiece / speaker and the bottom speaker periodically emit modulated acoustic signals alternately, while the top and bottom microphones collect signals;

[0092] Wherein, the emitted acoustic signal is a frequency-modulated continuous wave, and the signal period length is T s ; The signal collected by the microphone is a mixed signal, including the direct signal of the earpiece / speaker, the reflected signal via the nearby object, and the noise in the environment.

[0093] 3) Calculate the first component of the wind speed in the initial pose according to the two-channel signals collected by the two microphones;

[0094] 4) Rotate the smart terminal by a certain angle on the horizontal plane, and turn off the emission and collection of acoustic waves during the rotation process;

[0095] The basic principle of the rotation of the smart terminal is: rotate clockwise or counterclockwise on the horizontal plane, and the rotation angle σ ∈ (-π, π).

[0096] 5) After detecting that the smart terminal is stationary, calculate the rotation angle of the smart terminal compared with the initial pose based on the data of the inertial sensor, and simultaneously start the emission and collection process of the acoustic wave;

[0097] Wherein, the calculation of the rotation angle of the smart terminal compared with the initial pose is specifically:

[0098] 51) Calculate the rotation angle σ of the smart terminal around the Z g axis according to the gyroscope data in the current rotation process;

[0099] 52) Calculate the current pose of the smart terminal through the accelerometer data and the magnetometer data when stationary after rotation, and obtain the rotation angle θ of the smart terminal around the Z g axis;

[0100] 53) Record the rotation angle of the smart terminal compared with the initial pose when stationary last time as β i-1 , and the rotation angle of the smart terminal compared with the initial pose when stationary this time as β i , i ≥ 1, let ω = θ0+ β i-1 + σ, then:

[0101]

[0102] Wherein, β0=0 at the initial time; ρ1 and ρ2 are weight factors, ρ1+ρ2=1.

[0103] 6) calculating a second component of the wind speed according to the two-channel signal collected by the two microphones;

[0104] 7) repeating steps 4) to 6) to obtain the remaining N-2 wind speed components, N≥2;

[0105] Wherein, the i-th component of the wind speed calculated in steps 3), 6), and 7), i∈[1,N], is specifically:

[0106] S1) processing the two-channel signal collected by the microphone to calculate the propagation time delay T i1 of the sound wave from the bottom loudspeaker to the top microphone i2 ;

[0107] S2) letting V s be the propagation speed of the sound wave in the air, W i be the i-th component of the wind speed, D1 be the distance from the bottom loudspeaker to the top microphone, and D2 be the distance from the top earpiece / loudspeaker to the bottom microphone, then:

[0108]

[0109] Further, the wind speed component W i can be obtained:

[0110]

[0111] Specifically, as shown in the accompanying drawings, Figure 4 , Figure 5 the calculation of the propagation time delay of the sound wave from the bottom loudspeaker to the top microphone in step S1) is specifically:

[0112] S11) band-pass filtering the two-channel signal collected by the microphone to remove environmental noise in the signal;

[0113] S12) cross-correlating the filtered signal with the transmitted signal respectively to obtain the processing result Xcorr b of the bottom microphone t and the processing result Xcorr b of the top microphone s ;

[0114] S13) performing peak detection on the sequence Xcorr b with the period length T s of the transmitted signal as the interval to obtain the peak values PV1, PV2,..., PV K , and then taking the corresponding subscripts of the first large peak values

[0115] S14) Extracting subsequence with index in [P b - μT t · F j , P s + 3 μT s · F j ] from sequence Xcorr s and Xcorr s respectively, extracting upper envelope of subsequence and normalizing, then X times up-interpolation to normalized upper envelope curve, respectively obtaining sequence and where μ ∈ (0, 0.2), F s is system sampling rate,

[0116] S15) Obtaining maximum peak index in sequence and peak index closest to in sequence then:

[0117]

[0118] where, μ = 0.1, T s = 0.01, F s = 48000, X = 64.

[0119] Wherein, the step S1) in the calculation of the sound wave propagation delay from the top microphone / loudspeaker to the bottom microphone The method of calculating the propagation delay from the bottom loudspeaker to the top microphone is consistent, which will not be repeated here.

[0120] 8) Calculate the horizontal wind speed value by simultaneously considering N wind speed components and corresponding rotation angles, and determine the wind direction according to the initial pose of the intelligent terminal, specifically:

[0121] 81) Record the i-th component of the wind speed and the rotation angle of the intelligent terminal compared to the initial pose as (W i , β i ), i ∈ [1, N];

[0122] 82) Let the horizontal wind speed value be W, and the angle between the initial pose of the intelligent terminal and the wind direction on the horizontal plane be α, then:

[0123] W i = W · cos (β i - α);

[0124] Considering the interference and measurement error in actual situation, refer to Figure 6As shown, the wind speed value W and the included angle a are further optimized based on the least square method:

[0125]

[0126] 83) On the basis of the included angle a, the initial pose θ0 of the intelligent terminal in the world coordinate system is further considered, that is, the horizontal rotation angle around the world coordinate system Z g axis, to obtain the wind direction (θ0+a) in the world coordinate system, as shown in Figure 7 .

[0127] The present application has many specific application approaches, and the above description is only the preferred embodiment of the present application. It should be pointed out that for ordinary skilled persons in the technical field, some improvements can be made without departing from the principles of the present application, and these improvements should also be considered as the protection scope of the present application.

Claims

1. A method of wind speed measurement based on fusion of acoustic and inertial sensing, based on a system of fusion of acoustic and inertial sensing, the system comprising: An audio module, an inertial sensor, a processor, and a memory; The audio module is used for emitting and collecting sound wave signals, and includes two sets of audio transceivers, i.e., a top earpiece / speaker and a microphone, and a bottom speaker and a microphone; The inertial sensor is used for collecting the position and rotation data of the intelligent terminal, and includes an accelerometer, a gyroscope, and a magnetometer; The processor is used for controlling the audio module to emit and collect sound wave signals, and controlling the data collection of the inertial sensor, and processing the collected sound wave signal data and sensor data; The memory is used for storing the collected sound wave signal data and sensor data, and the calculation result data of the processor; The method comprises the following steps: 1) After calibrating the magnetometer, the intelligent terminal is placed on a horizontal plane, and the initial position of the intelligent terminal is determined based on the inertial sensor; 2) The top earpiece / speaker and the bottom speaker periodically and alternately emit modulated sound wave signals, while the top microphone and the bottom microphone collect signals; 3) The first component of the wind speed at the initial position is calculated according to the two-channel signals collected by the two microphones; 4) The intelligent terminal is rotated by a certain angle on the horizontal plane, and the emission and collection of sound waves are turned off during the rotation; 5) After detecting that the intelligent terminal is stationary, the rotation angle of the intelligent terminal compared with the initial position is calculated based on the data of the inertial sensor, and the emission and collection of sound waves are turned on; 6) The second component of the wind speed is calculated according to the two-channel signals collected by the two microphones; 7) Steps 4) to 6) are repeated to obtain the remaining N-2 wind speed components, N≥2; 8) The horizontal wind speed value is calculated based on the N wind speed components and the corresponding rotation angles, and the wind direction is determined according to the initial position of the intelligent terminal; The method for calculating the wind speed value and determining the wind direction in step 8) is as follows: 81) the i-th component of the wind speed and the rotation angle of the corresponding intelligent terminal compared to the initial pose are (W i ,β i ), i ∈ [1, N]; 82) Let the horizontal wind speed value be W, and the angle between the initial position of the intelligent terminal and the wind direction on the horizontal plane be α, then we have: W i = W cos(β - α) i -α); Considering the interference and measurement error in the actual situation, the wind speed value W and the angle α are further optimized based on the least squares method: 83) Further consider the initial pose θ0of the intelligent terminal in the world coordinate system on the basis of the included angle α, that is, the horizontal rotation angle of the intelligent terminal around the world coordinate system Z g axis, to obtain the wind direction (θ0+α) in the world coordinate system.

2. The method of claim 1, wherein, The layout structure of the two sets of audio transceivers should satisfy that the angle between the line connecting the top earpiece / speaker to the bottom microphone and the line connecting the bottom speaker to the top microphone is within 30°.

3. The method of claim 1, wherein, The specific method for determining the initial position of the intelligent terminal based on the inertial sensor in step 1) is as follows: 11) three axes X g ,Y g ,Z g respectively east, north, sky direction, three axes X l ,Y l ,Z l of the intelligent terminal coordinate system are respectively right along the screen, up along the screen and outward perpendicular to the screen; the vector measured by the intelligent terminal accelerometer is a = [a x ,a y ,a z ], the vector measured by the magnetometer is m = [m x ,m y ,m z ], then the unit vector h along the eastward X g axis and the unit vector l along the northward Y g axis are obtained. wherein h x ,h y ,h z are the component values of the vector h along the three axes of the intelligent terminal coordinate system, respectively; l x ,l y ,l z are the component values of the vector l along the three axes of the intelligent terminal coordinate system, respectively; Further, the rotation matrix R from the world coordinate system to the intelligent terminal coordinate system is obtained: R = [h, l, r] T where r = [r x ,r y ,r z ] = a / |a| and |a| in the rest state is equal to the magnitude of the gravitational acceleration; 12) Set the initial pose of the intelligent terminal in the world coordinate system as: rotating angle θ0 around Z g axis, rotating angle γ0 around Y g axis axis g axis; then according to the rotation matrix, 4. The method of claim 1, wherein, The acoustic wave signal emitted in step 2) is a frequency-modulated continuous wave, and the signal period length is T s The signal collected by the microphone is a mixed signal, including the direct signal of the earpiece / speaker, the reflected signal via nearby objects, and the noise in the environment.

5. The method of claim 1, wherein, The i-th component of the wind speed is calculated in steps 3), 6), and 7), i∈[1,N], which is as follows: S1 ) processing the two-channel signal captured by the microphones, calculating the propagation time delay T of the sound wave from the bottom loudspeaker to the top microphone i1 and the propagation time delay T of the sound wave from the top earpiece / loudspeaker to the bottom microphone i2 ; S2) Let V s be the speed of sound in air, W i be the i-th component of the wind velocity, D1 be the distance from the bottom speaker to the top microphone, and D2 be the distance from the top earpiece / speaker to the bottom microphone, then: Further, the wind speed component W is obtained i :

6. The method of claim 5, wherein, The propagation time delay of the sound wave from the bottom speaker to the top microphone is calculated in step S1), which is as follows: S11) Band-pass filtering is performed on the two-channel signals collected by the microphone to remove environmental noise in the signals; S12) cross-correlating the filtered signals with the transmitted signals respectively, and then obtaining the processing result Xcorr of the bottom microphone b and the processing result Xcorr of the top microphone t ; S13) taking the sequence Xcorr b with the period length T s Peak detection is performed at intervals of the period length T K , and the peak values PV1, PV2,..., PV corresponding to the largest peak values S14) taking the subsequence with index in [P j - μT s · F s , P j + 3 μT s · F s ] from the sequence Xcorr b and Xcorr t respectively, extracting the upper envelope of the subsequence and normalizing it, and then performing an X-fold up-interpolation of the normalized upper envelope curve, to obtain the sequences and respectively, where μ ∈ (0, 0.2) and F s is the system sampling rate, S15) obtaining the sequence the index of the maximum peak in and the sequence the index of the peak closest to the index of the maximum peak in then:

7. The method of claim 1, wherein, The basic principle of the rotation of the intelligent terminal in step 4) is as follows: clockwise or counterclockwise rotation is performed on the horizontal plane, and the rotation angle σ∈(-π,π).

8. The method of claim 1, wherein, The rotation angle of the intelligent terminal compared with the initial position is calculated in step 5), which is as follows: 51) Calculate the rotation angle σ around the Z axis from the gyroscope data during the current rotation g cycle. 52) Calculate the current pose of the intelligent terminal by the accelerometer data and the magnetometer data when stationary after the rotation ends, obtain the rotation angle θ of the intelligent terminal around the Z g axis; 53) the rotation angle of the intelligent terminal compared to the initial position at the last time of stillness is β i-1 , the rotation angle of the intelligent terminal compared to the initial position at the current time of stillness is β i , i≥1, let ω=θ0+β i-1 +σ, then: wherein β0=0 at the initial time; ρ1 and ρ2 are weight factors, and ρ1+ρ2=1.

Images