A near-ultrasonic high-precision time of arrival estimation method based on velocity compensation

Abstract

The application discloses a near-ultrasonic high-precision time of arrival estimation method based on speed compensation. The method adopts a stationary loudspeaker to emit a customized signal, and a microphone of a smart phone in a moving state receives a return signal; the return signal is filtered to obtain a rising / falling linear frequency modulation signal; the rising / falling linear frequency modulation signal is respectively correlated with the customized signal to output respective cross-correlation results, and corresponding time offsets are extracted; reference signals for generating the rising / falling linear frequency modulation signal are adjusted according to the two time offsets; according to the rising / falling linear frequency modulation signal and the corresponding reference signals, mixing, filtering and fast Fourier transform processing are sequentially performed, and then joint analysis and calculation processing is performed to obtain time of arrival estimation. The application does not depend on additional sensors, can accurately estimate the time of arrival of a moving target, and can realize accurate time of arrival estimation of a moving target in a complex indoor scene.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of indoor positioning, and particularly relates to a near-ultrasonic high-precision time of arrival estimation method based on speed compensation. BACKGROUND

[0002] In recent years, with the continuous expansion of application fields, indoor positioning technology has attracted much attention. It can not only provide services based on location information in public space navigation and other fields, but also be used for controlling indoor unmanned aerial vehicles and industrial robots. Acoustic positioning technology transmits ultrasonic waves by base stations, and the received signals are received by receiving devices for positioning. The technology has strong anti-interference ability, and high-precision positioning can be achieved in a noisy environment by using filtering and cross-correlation methods. Acoustic positioning technology relies on high-precision time of arrival estimation. However, when the time of arrival is estimated for a moving target, the Doppler effect will cause a certain deviation in the estimated value. The positioning result calculated according to this time of arrival has a large error. Therefore, it is of great significance to eliminate the influence of the Doppler effect and estimate the time of arrival of the moving target with high precision.

[0003] Some algorithms estimate the speed of the moving target in real time by adding sensors or transmitting additional signals, and correct the estimated time of arrival using the measured speed. However, additional sensors often require strict data calibration and complex processing algorithms. In addition, it is difficult to estimate the speed of the moving target in real time using additional signals, and the speed estimation deviation is large, resulting in poor estimation accuracy of the time of arrival.

[0004] Some other algorithms eliminate the influence of the Doppler effect through unique signal design, but existing algorithms are limited in speed and time of arrival estimation accuracy, and it is difficult to achieve high-precision time of arrival estimation for high-speed moving objects.

[0005] In summary, there is an urgent need for a high-precision time of arrival estimation algorithm that can effectively compensate for the Doppler shift. SUMMARY

[0006] To solve the above problems in the prior art, the application provides a near-ultrasonic high-precision time of arrival estimation algorithm based on speed compensation, which does not rely on additional sensors and accurately estimates the time of arrival of the moving target.

[0007] The application adopts the following technical solutions:

[0008] The near-ultrasonic high-precision time of arrival estimation method based on speed compensation comprises the following steps:

[0009] S1, a stationary loudspeaker emits a customized signal to the surrounding environment, and a smartphone microphone in a moving state receives the customized signal as a return signal;

[0010] S2, filtering and separating the return signal to separate the rising chirp signal and the falling chirp signal in the return signal;

[0011] S3, respectively correlating the rising chirp signal and the falling chirp signal with the customized signal to output respective correlation results, and extracting respective time offsets corresponding to maximum values in the two correlation results;

[0012] S4, time-shifting and frequency-expanding the customized signal according to the two time offsets to generate two different reference signals, respectively as reference signals of the rising chirp signal and the falling chirp signal;

[0013] S5, sequentially performing mixing, filtering, and fast Fourier transform processing on each chirp signal in the rising chirp signal and the falling chirp signal and its corresponding reference signal, and then performing joint analysis and calculation processing according to the two time offsets to obtain the time of arrival estimation.

[0014] The customized signal in S1 is set according to the following formula:

[0015] s(t) = s up (t) + s down (t)

[0016]

[0017] The customized signal mainly consists of the rising chirp signal and the falling chirp signal, wherein and respectively represent the starting frequency of the rising chirp signal and the falling chirp signal, B up and B down respectively represent the bandwidth of the rising chirp signal and the falling chirp signal, T represents the duration of the signal, and t represents time, t ∈ [0, T].

[0018] The correlation in step S3 uses generalized cross-correlation.

[0019] In S4, the customized signal is time-shifted and frequency-expanded according to the following settings:

[0020] The starting time of the reference signal corresponding to the rising chirp signal is A represents a time-shifting parameter, and the frequency is expanded in a manner that the ratio of is unchanged;

[0021] The starting time of the reference signal corresponding to the falling chirp signal is The frequency is expanded in a manner that the ratio of is unchanged.

[0022] The S5 specific steps are:

[0023] S5.1, for each linear frequency modulation signal in the rising linear frequency modulation signal and the falling linear frequency modulation signal and its corresponding reference signal, sequentially mixing, filtering, fast Fourier transform processing to obtain the respective intermediate frequency signal spectrum;

[0024] S5.2, jointly analyzing the intermediate frequency signal spectrum of the rising linear frequency modulation signal and the falling linear frequency modulation signal to obtain the direct signal frequency mean;

[0025] S5.3, according to the customized signal, the direct signal frequency mean and the time offset And Calculate the arrival time estimate.

[0026] The S5.1 specific steps are:

[0027] S5.1.1, mix the rising linear frequency modulation signal and the falling linear frequency modulation signal with their corresponding reference signals respectively to obtain respective mixed signals, and then filter to obtain respective intermediate frequency signals;

[0028] S5.1.2, fast Fourier transform the respective intermediate frequency signals to obtain the respective intermediate frequency signal spectrum of the linear frequency modulation signal.

[0029] The S5.2 specific steps are:

[0030] S5.2.1, respectively peak detection on the intermediate frequency signal spectrum of the rising linear frequency modulation signal and the falling linear frequency modulation signal;

[0031] S5.2.2, in the intermediate frequency signal spectrum of the rising linear frequency modulation signal / falling linear frequency modulation signal, extract the frequency corresponding to the peak with the maximum power after satisfying the following formula as the direct signal peak frequency

[0032]

[0033] Where And Respectively represent the starting frequency of the rising linear frequency modulation signal and the falling linear frequency modulation signal, v max Indicates the maximum speed of the smartphone microphone motion.

[0034] S5.3.3, weighted average the direct signal peak frequencies of the rising linear frequency modulation signal and the falling linear frequency modulation signal to obtain the direct signal frequency mean.

[0035] The S5.3 arrival time estimate is obtained by the following formula:

[0036]

[0037] wherein, τ d denotes the time of arrival estimation, denotes the time offset corresponding to the maximum value in the result of the up chirp signal and the customized signal cross-correlation, denotes the time offset corresponding to the maximum value in the result of the down chirp signal and the customized signal cross-correlation, A denotes a time shift adjustment parameter, B up denotes the bandwidth of the up chirp signal, T denotes the duration of the signal.

[0038] The beneficial effects of the present application are:

[0039] The mixed signal of 16-19kHz and 22-19kHz is used as the transmitting signal; the up chirp signal and the down chirp signal in the received signal are separated by a band-pass filter; the GCC is used to roughly estimate the time of arrival, and the FMCW algorithm is used to jointly estimate the peak frequency corresponding to the direct signal, so that the accurate time of arrival estimation of the moving target in a complex indoor scene is realized.

[0040] The near-ultrasonic high-precision time of arrival estimation algorithm based on speed compensation has high precision and good robustness, and can accurately estimate the time of arrival of a moving target. BRIEF DESCRIPTION OF DRAWINGS

[0041] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the drawings needed to be used in the embodiments or the prior art description will be briefly introduced. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative labor.

[0042] Figure 1 is a flow chart of the near-ultrasonic high-precision time of arrival estimation algorithm based on speed compensation according to the present application;

[0043] Figure 2 is a waveform diagram and a spectrogram of the customized signal;

[0044] Figure 3 is a design scheme diagram of the reference signal;

[0045] Figure 4 is a spectrum diagram of the intermediate frequency signal of the up chirp signal and the down chirp signal; DETAILED DESCRIPTION

[0046] Following, the advantages and effects of the present application can be easily understood by those skilled in the art from the content disclosed in the specification. The present application can also be implemented or applied by other different specific embodiments, and the details in the specification can be modified or changed based on different views and applications without departing from the spirit of the present application. It should be noted that the following embodiments and features in the embodiments can be combined with each other without conflict.

[0047] As shown in Figure 1 , the specific embodiment of the present application near-ultrasonic high-precision time of arrival estimation method comprises the following steps:

[0048] S1, a stationary loudspeaker emits a customized signal to the surrounding environment, and the customized signal is received by a smartphone microphone in a moving state as a return signal; the return signal includes a direct signal;

[0049] As shown in Figure 2 and Figure 3 , the customized signal in S1 is set according to the following formula:

[0050] s(t)=s up (t)+s down (t)

[0051]

[0052] The customized signal is mainly composed of an uplink linear frequency modulation signal s up (t) and a downlink linear frequency modulation signal s down (t), wherein and represent the starting frequencies of the uplink linear frequency modulation signal and the downlink linear frequency modulation signal, respectively, B up and B down represent the bandwidths of the uplink linear frequency modulation signal and the downlink linear frequency modulation signal, respectively, T represents the duration of the signal, and t is time, t∈[0,T].

[0053] As shown in Figure 2 , the customized signal is mainly composed of an uplink linear frequency modulation signal and a downlink linear frequency modulation signal;

[0054] In this embodiment, the duration of the customized signal is 40ms, the frequency of the uplink linear frequency modulation signal is 16-19kHz, the frequency spectrum of the downlink linear frequency modulation signal is 22-19kHz, the sampling frequency is 48kHz, and the signal refresh rate is 5Hz.

[0055] The TDOA algorithm is often used in the near-ultrasound indoor positioning technology, and the sound signal needs to select an inaudible frequency band and a signal with good correlation. The linear sweep signal meets the above requirements, and the frequency of the signal changes linearly with time. The signal has good correlation.

[0056] In this embodiment, the upper and lower cutoff frequencies of the bandpass filter separating the up chirp signal are 15.9 kHz and 19.1 kHz respectively, and the order is 20; the upper and lower cutoff frequencies of the bandpass filter separating the down chirp signal are 18.9 kHz and 22.1 kHz respectively, and the order is 20.

[0057] S2, the filter is used for bandpass filtering and separating the return signal, and the up chirp signal and the down chirp signal in the return signal are separated;

[0058] Under the condition of ignoring multipath propagation, the return signal s r (t) is:

[0059]

[0060] The moving microphone separates the received signal into up chirp signal and down chirp signal

[0061]

[0062] Where t d is the delay time of signal propagation, and a is the attenuation coefficient.

[0063] In the formula, h up (t) and h down (t) represent the bandpass filter separating the up chirp signal and the down chirp signal respectively, and * represents convolution operation. The above formula represents filtering of the received signal to separate the received up chirp signal and down chirp signal.

[0064] Since the transmitted signal is a mixed signal of up chirp signal and down chirp signal, in order to avoid mutual interference, a bandpass filter must be designed to effectively separate them.

[0065] S3, the up chirp signal and the down chirp signal are respectively correlated with the customized signal to output respective cross-correlation results, and the time offset corresponding to the respective maximum values of the two cross-correlation results is extracted and

[0066] The cross-correlation uses Generalized Cross-Correlation (GCC).

[0067] and are the time offsets corresponding to the respective maximums in the results of the two cross-correlations:

[0068]

[0069] where R up (τ) and R down (τ) represent the results of the upchirp and downchirp cross-correlations, respectively.

[0070] S4, according to the two time offsets and time-shift and frequency-spread the custom signal to generate two different reference signals, as the reference signals of the up-chirp and down-chirp linear frequency modulation signals, respectively;

[0071] where the time-shift and frequency-spread of the custom signal according to the following settings are performed, respectively:

[0072] The start time of the reference signal corresponding to the up-chirp linear frequency modulation signal is A represents a time-shift parameter, and the frequency is expanded in a manner that the ratio of the frequency to the time is unchanged; in this embodiment, the frequency is expanded to 16-20.5 kHz.

[0073] The start time of the reference signal corresponding to the down-chirp linear frequency modulation signal is The frequency is expanded in a manner that the ratio of the frequency to the time is unchanged; in this embodiment, the frequency is expanded to 22-17.5 kHz, and A is taken as 10 ms.

[0074] In this embodiment, the start time of the reference signal corresponding to the up-chirp linear frequency modulation signal is The duration is 60 ms, and the frequency is expanded to 16-20.5 kHz; the start time of the reference signal corresponding to the down-chirp linear frequency modulation signal is The duration is 60 ms, and the frequency is expanded to 22-17.5 kHz.

[0075] Reference signal design: By optimizing and adjusting the start time and bandwidth of the reference signal, the estimation accuracy of the algorithm for the time delay is enhanced.

[0076] ​​S5.1, for each linear frequency modulation signal in the ascending linear frequency modulation signal and the descending linear frequency modulation signal and its corresponding reference signal, sequentially perform mixing, filtering, fast Fourier transform processing to obtain respective intermediate frequency signal spectrum diagrams;

[0077] S5.1, for each linear frequency modulation signal in the ascending linear frequency modulation signal and the descending linear frequency modulation signal and its corresponding reference signal, sequentially perform mixing, filtering, fast Fourier transform processing to obtain respective intermediate frequency signal spectrum diagrams;

[0078] S5.1, for each linear frequency modulation signal in the ascending linear frequency modulation signal and the descending linear frequency modulation signal and its corresponding reference signal, sequentially perform mixing, filtering, fast Fourier transform processing to obtain respective intermediate frequency signal spectrum diagrams;

[0079] S5.1, for each linear frequency modulation signal in the ascending linear frequency modulation signal and the descending linear frequency modulation signal and its corresponding reference signal, sequentially perform mixing, filtering, fast Fourier transform processing to obtain respective intermediate frequency signal spectrum diagrams;

[0080] S5.1, for each linear frequency modulation signal in the ascending linear frequency modulation signal and the descending linear frequency modulation signal and its corresponding reference signal, sequentially perform mixing, filtering, fast Fourier transform processing to obtain respective intermediate frequency signal spectrum diagrams;

[0081] S5.1, for each linear frequency modulation signal in the ascending linear frequency modulation signal and the descending linear frequency modulation signal and its corresponding reference signal, sequentially perform mixing, filtering, fast Fourier transform processing to obtain respective intermediate frequency signal spectrum diagrams.

[0082] S5.1, for each linear frequency modulation signal in the ascending linear frequency modulation signal and the descending linear frequency modulation signal and its corresponding reference signal, sequentially perform mixing, filtering, fast Fourier transform processing to obtain respective intermediate frequency signal spectrum diagrams;

[0083] As shown in Figure 4 , mixing is performed without considering the Doppler effect. Using the formula and using a low-pass filter, the intermediate frequency signal of the ascending linear frequency modulation signal

[0084]

[0085] Considering that the signal speaker is stationary and the microphone moves relative to the speaker at a speed v, the distance between the microphone and the speaker is R. At this time, the intermediate frequency signal can be represented as:

[0086]

[0087] where c represents the propagation speed of the acoustic signal in the air. Neglecting the high-order term and considering The frequency of the intermediate frequency signal can be represented as:

[0088]

[0089] The time of arrival between the microphone and the loudspeaker can be obtained by estimating the frequency of the intermediate frequency signal, but from the above formula, the frequency of the intermediate frequency signal is affected by the speed of the microphone movement, which in turn affects the final time of arrival estimate.

[0090] The frequency of the intermediate frequency signal corresponding to the descending chirp signal should be:

[0091]

[0092] In the proposed signal design scheme, the custom signal we designed satisfies the condition Therefore, the following formula is obtained:

[0093]

[0094] S5.2, jointly analyze the intermediate frequency signal spectrum of the rising chirp signal and the descending chirp signal to obtain the mean value of the direct signal frequency;

[0095] The specific steps of S5.2 are as follows:

[0096] S5.2.1, respectively, peak detection is performed on the intermediate frequency signal spectrum of the rising chirp signal and the descending chirp signal.

[0097] S5.2.2, when jointly analyzing the spectrum of the intermediate frequency signals corresponding to the rising chirp signal and the descending chirp signal, the proposed custom constraint condition is used to find the peak frequency.

[0098] In the intermediate frequency signal spectrum of the rising chirp signal / descending chirp signal, the frequency corresponding to the peak with the maximum power that satisfies the following formula is extracted as the direct signal peak frequency

[0099]

[0100] wherein and respectively represent the starting frequency of the rising chirp signal and the descending chirp signal, v max represents the maximum speed of the smartphone microphone movement.

[0101] In this embodiment, the maximum speed of the moving target is 1.5 m / s, the speed of sound is 346.15 m / s, is 22 kHz, is 16 kHz, and the calculated frequency difference is 164.67 Hz. In order to enhance the robustness of the algorithm, the threshold is amplified by 1.5 times, i.e. 247.01 Hz. In this embodiment, B is 247.01 Hz.​

[0102] Joint analysis: In order to improve the robustness of the algorithm and make accurate time of arrival estimation for the microphone with faster speed, the peak value of the intermediate frequency spectrum of the rising chirp signal and the falling chirp signal meeting the condition is extracted respectively by using the set constraint condition, that is, the above formula.

[0103] S5.2.3, the direct signal peak frequency of the rising chirp signal and the falling chirp signal is weighted and averaged to obtain the direct signal frequency mean.

[0104] Speed compensation: In order to eliminate the influence of Doppler effect on time of arrival estimation, the peak frequencies of the rising chirp signal and the falling chirp signal are added to eliminate the influence of speed on time of arrival estimation in the frequency domain. When the microphone moves faster, the peak frequency difference of the rising chirp signal and the falling chirp signal is too large, as shown in the figure, direct addition will cause the generation of false peak value. Figure 4

[0105] S5.3, according to the custom signal, the direct signal frequency mean and the time offset and calculation and processing to obtain the time of arrival estimation.

[0106] Wherein the time of arrival estimation is obtained by the following formula:

[0107]

[0108] Wherein, τ d represents the time of arrival estimation, represents the time offset corresponding to the maximum value in the result of the cross-correlation between the rising chirp signal and the custom signal, represents the time offset corresponding to the maximum value in the result of the cross-correlation between the falling chirp signal and the custom signal, A represents the time shift adjustment parameter, B up represents the bandwidth of the rising chirp signal, and T represents the duration of the signal.

[0109] The above-described embodiments are only used to describe the preferred embodiments of the present application, and do not limit the scope of the present application. Without departing from the design spirit of the present application, various modifications and improvements of the technical solutions of the present application made by those skilled in the art shall fall within the protection scope of the present application.​

Claims

1. A high-precision near-ultrasonic arrival time estimation method based on velocity compensation, characterized in that, Including the following steps: S1. A stationary speaker emits a custom signal to the surrounding environment, and the custom signal is received as a return signal by the microphone of a smartphone in motion. The customized signal in S1 is set according to the following formula: s(t)=s up (t)+s down (t) The customized signal mainly consists of a rising linear frequency modulated (RF) signal and a falling linear frequency modulated (RF) signal, wherein and B represents the starting frequencies of the rising linear frequency modulated (RFM) signal and the falling linear frequency modulated (RFM) signal, respectively. up and B down Let T and t represent the bandwidths of the rising linear frequency modulated (RFM) signal and the falling linear frequency modulated (RFM) signal, respectively, where T represents the duration of the signal, t represents time, and t∈[0,T]. S2. Perform filtering and separation processing on the returned signal to separate the rising linear frequency modulated signal and the falling linear frequency modulated signal from the returned signal; S3. Perform cross-correlation processing on the rising linear frequency modulation signal, the falling linear frequency modulation signal and the customized signal respectively, output their respective cross-correlation results, and extract the time offset corresponding to the maximum value of each of the two cross-correlation results. S4. Based on the two time offsets, the customized signal is time-shifted and frequency-spread to generate two different reference signals, which are used as reference signals for the rising linear frequency modulation signal and the falling linear frequency modulation signal, respectively. S5. Based on each linear frequency modulation signal in the rising linear frequency modulation signal and the falling linear frequency modulation signal and its corresponding reference signal, perform mixing, filtering and fast Fourier transform processing in sequence, and then combine the two time offsets for joint analysis and calculation to obtain the arrival time estimate. The specific steps of S5 are as follows: S5.1 For each linear frequency modulation signal and its corresponding reference signal in the rising linear frequency modulation signal and the falling linear frequency modulation signal, perform mixing, filtering and fast Fourier transform processing in sequence to obtain the intermediate frequency signal spectrum of each signal. S5.

2. Perform joint analysis on the intermediate frequency signal spectrum diagrams of the rising linear frequency modulated signal and the falling linear frequency modulated signal to obtain the average frequency of the direct signal; S5.3, Based on the customized signal, the average frequency of the direct signal, and the time offset. and The calculation process yields the estimated arrival time. The arrival time estimate in S5.3 is obtained using the following formula: Where, τ d Indicates the estimated arrival time. This represents the time offset corresponding to the maximum value in the cross-correlation result between the rising linear frequency modulated signal and the custom signal. This represents the time offset corresponding to the maximum value in the cross-correlation result of the falling linear frequency modulated signal and the customized signal. A represents the time shift adjustment parameter, and B represents the time shift adjustment parameter. up The signal represents the bandwidth of the rising linear frequency modulated signal, and T represents the duration of the signal.

2. The near-ultrasonic high-precision time-of-arrival estimation method based on velocity compensation according to claim 1, characterized in that, The cross-correlation described in step S3 adopts generalized cross-correlation.

3. The near-ultrasonic high-precision time-of-arrival estimation method based on velocity compensation according to claim 1, characterized in that, In S4, time-shift adjustment and frequency expansion of the customized signal are performed according to the following settings: The start time of the reference signal corresponding to the rising linear frequency modulation signal is A represents the time shift adjustment parameter, with the frequency according to... Expanding the ratio while keeping it constant; The start time of the reference signal corresponding to the falling linear frequency modulated signal is Frequency according to Expanding the ratio while keeping it constant; in, This represents the time offset corresponding to the maximum value in the cross-correlation result between the rising linear frequency modulated signal and the custom signal. B represents the time offset corresponding to the maximum value in the cross-correlation result of the falling linear frequency modulated signal and the custom signal. up and B down These represent the bandwidths of the rising linear frequency modulated (RF) signal and the falling linear frequency modulated (RF) signal, respectively, and T represents the duration of the signal.

4. The near-ultrasonic high-precision time-of-arrival estimation method based on velocity compensation according to claim 1, characterized in that, The specific steps in S5.1 are as follows: S5.1.1 Mix the rising linear frequency modulation signal and the falling linear frequency modulation signal with their corresponding reference signals to obtain their respective mixed signals, and then filter them to obtain their respective intermediate frequency signals; S5.1.

2. Perform Fast Fourier Transform on each intermediate frequency signal to obtain the intermediate frequency signal spectrum of each linear frequency modulated signal.

5. The near-ultrasonic high-precision arrival time estimation method based on velocity compensation according to claim 1, characterized in that, The specific steps in S5.2 are as follows: S5.2.1 Perform peak detection on the intermediate frequency signal spectrum diagrams of the rising linear frequency modulated signal and the falling linear frequency modulated signal, respectively; S5.2.2 In the intermediate frequency signal spectrum diagram of the rising linear frequency modulated signal / falling linear frequency modulated signal, extract the frequency corresponding to the peak with the highest power that satisfies the following formula as the peak frequency of the direct signal. in and These represent the starting frequencies of the rising linear frequency modulated (RFM) signal and the falling linear frequency modulated (RFM) signal, respectively. max This represents the maximum speed at which the smartphone microphone moves, and c represents the speed at which sound signals travel through the air. S5.3.

3. The average direct signal frequency is obtained by weighting the peak frequencies of the rising linear frequency modulated signal and the falling linear frequency modulated signal.

Images