Indoor target positioning method and system based on acoustic device

Abstract

The invention discloses an indoor target positioning method and system based on an acoustic device, and relates to the technical field of indoor measurement positioning, and the method comprises the steps: transmitting an ultrasonic linear frequency modulation continuous wave signal and receiving a target echo through employing an existing loudspeaker and microphone array; a single reference signal is adopted to complete dechirp processing in a digital domain, and then the FFT operation length is shortened through a periodic continuation reconstruction method; and finally, respectively estimating the target distance and angle based on the distance spectrum and the array signal. According to the invention, full-digital processing is realized, an additional analog front end is not needed, hardware cost and operation complexity are obviously reduced, real-time operation on a low-power-consumption microcontroller can be realized, and the method is suitable for embedded space sensing in the fields of smart home, robots and the like.

Description

Technical Field

[0001] This invention belongs to the field of indoor measurement and positioning technology, and particularly relates to an indoor target positioning method and system based on acoustic devices. Background Technology

[0002] In the field of indoor target localization and measurement technology, current solutions primarily rely on specialized sensors such as ultrasonic radar, millimeter-wave radar, or lidar. Ultrasonic radar has gained some application due to its lower cost, but its inherent limitation is that it can only measure target distance and cannot detect angular information, thus restricting its functionality in practical applications requiring two-dimensional or three-dimensional spatial perception. Millimeter-wave radar can simultaneously measure distance and angle, but its hardware system is complex and expensive. In typical indoor home environments, its radio wave signals are easily interfered with by multipath effects caused by reflections from walls, furniture, and other objects, severely affecting measurement accuracy and reliability. While lidar offers excellent measurement results, its manufacturing cost is extremely high, making large-scale deployment in cost-sensitive consumer products difficult. More importantly, all of the above existing solutions require the installation and use of additional specialized sensing devices, which undoubtedly increases the overall hardware cost, power consumption, and physical space occupation of the system.

[0003] Many widely adopted smart home appliances and mobile robots already have built-in audio hardware such as speakers and microphones. However, existing indoor positioning technologies cannot effectively reuse these existing general-purpose acoustic devices, originally designed for audio processing, to achieve accurate spatial perception. Attempting to directly transfer mature high-frequency FMCW signal processing technology from the traditional radar field to an acoustic platform faces significant challenges, including slow sound wave propagation leading to long echo processing windows, inconsistent signal processing flows, and the difficulty of low-power embedded microcontrollers processing long data frames in real time. Therefore, the field has long lacked a solution that can fully utilize existing audio hardware to achieve high-precision indoor target positioning, including angular information, at extremely low additional cost. Summary of the Invention

[0004] To address the aforementioned technical problems, this invention proposes an indoor target localization method and system based on acoustic devices. By employing a digital deskewing processing method based on acoustic characteristics and a periodic extended FFT reconstruction algorithm, high-precision localization is achieved without additional sensors while maintaining algorithm simplicity and real-time performance.

[0005] To achieve the above objectives, this invention provides an indoor target localization method based on acoustic devices, comprising:

[0006] The signal is emitted via a loudspeaker in the ultrasonic frequency band as a linear frequency modulated continuous wave.

[0007] The analog signal containing the target's reflected echo is received through a microphone array, and the analog signal is converted into a digital echo signal;

[0008] In the digital domain, the digital echo signal is mixed with a pre-generated single digital reference signal with a duration equal to the observation window to complete the deskewing process, and the mixing result is low-pass filtered to obtain the difference frequency signal.

[0009] The difference frequency signal is subjected to periodic extension reconstruction processing to form a reconstructed signal with a shortened length, and the reconstructed signal is subjected to a Fast Fourier Transform (FFT) to obtain the distance spectrum.

[0010] Target distance estimation is performed based on the aforementioned distance spectrum;

[0011] The target angle is estimated based on the spectral data of each channel in the microphone array at the target distance.

[0012] Optionally, descrambling in the numeric field includes:

[0013] The digital reference signal is pre-generated and stored inside the microcontroller, and the frequency modulation characteristics of the digital reference signal are consistent with the transmitted linear frequency modulated continuous wave signal.

[0014] The digital echo signal received within the entire observation window is multiplied with the digital reference signal in the digital domain.

[0015] A digital low-pass filter is used to filter out the high-frequency components generated by the multiplication operation, while retaining the difference frequency signal.

[0016] Optionally, performing periodic extension reconstruction processing on the difference frequency signal includes:

[0017] Difference frequency signal sequence of length N is divided into multiple consecutive segments, where each segment has a length of L, and L is less than N;

[0018] The signals of all segments are superimposed in the time domain to generate a reconstructed signal with a total length of L.

[0019] Optionally, target distance estimation based on the distance spectrum includes:

[0020] Calculate the amplitude spectrum of the distance spectrum;

[0021] Detect the peak value in the amplitude spectrum and determine the frequency index corresponding to the peak value;

[0022] The distance value of the target is calculated based on the mapping relationship between the frequency index and the target distance.

[0023] Optionally, the target angle estimation based on the spectral data of each channel in the microphone array at the target distance includes:

[0024] Extract the complex value at the frequency index corresponding to the target distance from the FFT result of each microphone channel;

[0025] A spatial signal vector is formed by the complex values ​​of all channels;

[0026] The direction-of-arrival (DOA) estimation algorithm is applied to the spatial signal vector to calculate the target's angle information.

[0027] On the other hand, to achieve the above objectives, the present invention also provides an indoor target positioning system based on acoustic devices, comprising:

[0028] The signal transmitting module is used to transmit linear frequency modulated continuous wave signals in the ultrasonic frequency band through a speaker;

[0029] The signal receiving and acquisition module is used to receive an analog signal containing the reflected echo of the target through a microphone array, and convert the analog signal into a digital echo signal;

[0030] The signal processing module is used to mix the digital echo signal with a pre-generated single digital reference signal with a duration equal to the observation window in the digital domain to complete the deskewing process, and to perform low-pass filtering on the mixing result to obtain a difference frequency signal; to perform periodic extension reconstruction processing on the difference frequency signal to form a reconstructed signal with a shortened length; to perform a Fast Fourier Transform (FFT) on the reconstructed signal to obtain a range spectrum; to perform target distance estimation based on the range spectrum; and to perform target angle estimation based on the spectral data of each channel in the microphone array corresponding to the target distance.

[0031] Optionally, the signal processing module includes a descrambling unit; the descrambling unit is used to perform the descrambling process, the descrambling process including:

[0032] The digital reference signal is pre-generated and stored;

[0033] The digital echo signal and the digital reference signal are multiplied in the digital domain;

[0034] A digital low-pass filter is used to filter out the high-frequency components generated by the multiplication operation, while retaining the difference frequency signal.

[0035] Optionally, the signal processing module includes a reconstruction unit; the reconstruction unit is used to perform the periodic extension reconstruction process, specifically including:

[0036] Difference frequency signal sequence of length N is divided into multiple consecutive segments, where each segment has a length of L, and L is less than N;

[0037] The signals of all segments are superimposed in the time domain to generate a reconstructed signal with a total length of L.

[0038] Optionally, the signal processing module includes a distance estimation unit; the distance estimation unit is used to perform the target distance estimation, specifically including:

[0039] Calculate the amplitude spectrum of the distance spectrum;

[0040] Detect the peak value in the amplitude spectrum and determine the frequency index corresponding to the peak value;

[0041] The distance value of the target is calculated based on the mapping relationship between the frequency index and the target distance.

[0042] Optionally, the signal processing module includes an angle estimation unit; the angle estimation unit is used to perform the target angle estimation, specifically including:

[0043] Extract the complex value at the frequency index corresponding to the target distance from the FFT result of each microphone channel;

[0044] A spatial signal vector is formed by the complex values ​​of all channels;

[0045] The direction-of-arrival (DOA) estimation algorithm is applied to the spatial signal vector to calculate the target's angle information.

[0046] Technical advantages of this invention: This invention discloses an indoor target localization method and system based on acoustic devices. It utilizes a single digital reference signal to perform deskewing processing in the digital domain, completely eliminating the need for analog mixers and low-pass filters, significantly reducing hardware complexity and system cost, while avoiding errors introduced by analog circuits. By using a periodic extension reconstruction method to fold long signal sequences into shorter sequences for processing, the computational load of the Fast Fourier Transform is greatly reduced, enabling the entire algorithm to run in real-time on low-computing-power microcontrollers. Based on complex spectral data from a microphone array, direction-of-arrival estimation is achieved, enabling the system to possess complete distance and angle joint estimation capabilities while reusing ordinary audio hardware, overcoming the limitation of traditional ultrasonic radar that can only measure distance but not angle. Ultimately, a fully digital signal processing chain is realized, providing a feasible technical path for low-cost deployment of high-precision indoor positioning functions on existing smart devices. Attached Figure Description

[0047] The accompanying drawings, which form part of this application, are used to provide a further understanding of this application. The illustrative embodiments and descriptions of this application are used to explain this application and do not constitute an undue limitation of this application. In the drawings:

[0048] Figure 1This is a flowchart illustrating an indoor target localization method based on acoustic devices according to an embodiment of the present invention.

[0049] Figure 2 This is a schematic diagram of an indoor target positioning system based on acoustic devices according to an embodiment of the present invention;

[0050] Figure 3 This is a schematic diagram of a single reference signal design according to an embodiment of the present invention. Detailed Implementation

[0051] It should be noted that, unless otherwise specified, the embodiments and features described in this application can be combined with each other. This application will now be described in detail with reference to the accompanying drawings and embodiments.

[0052] It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions, and although a logical order is shown in the flowchart, in some cases the steps shown or described may be executed in a different order than that shown here.

[0053] like Figure 1 As shown, this embodiment provides an indoor target positioning method based on acoustic devices, including: transmitting a linear frequency modulated continuous wave signal in the ultrasonic frequency band through a loudspeaker;

[0054] The analog signal containing the target's reflected echo is received through a microphone array, and the analog signal is converted into a digital echo signal;

[0055] In the digital domain, the digital echo signal is mixed with a pre-generated single digital reference signal with a duration equal to the observation window to complete the deskewing process, and the mixing result is low-pass filtered to obtain the difference frequency signal.

[0056] The difference frequency signal is subjected to periodic extension reconstruction processing to form a reconstructed signal with a shortened length, and the reconstructed signal is subjected to a Fast Fourier Transform (FFT) to obtain the distance spectrum.

[0057] Target distance estimation is performed based on the aforementioned distance spectrum;

[0058] The target angle is estimated based on the spectral data of each channel in the microphone array at the target distance.

[0059] Further, signal design and acquisition:

[0060] The transmitted signal uses a linear frequency modulated continuous wave (FMCW) ultrasonic signal, the mathematical expression of which is:

[0061] ;

[0062] in The signal representing the transmission. It is the starting frequency of the transmitted signal. The time variable represents the signal. The amplitude of the transmitted signal. It is the sweep bandwidth. It is the duration of a single frequency-modulated pulse signal (chirp).

[0063] After propagating through the air, the sound waves are reflected by multiple targets and received by a microphone array. The discrete signal obtained after ADC sampling is:

[0064] ;

[0065] in Represents the received discrete signal. It is the index number of the target reflection point. It is the number of target reflection points. Characterizing the first Echo attenuation of each target It is the sampling interval. It is the duration of transmission. It refers to the entire observation period. This indicates received noise.

[0066] Furthermore, deskewing in the number domain includes:

[0067] The digital reference signal is pre-generated and stored inside the microcontroller, and the frequency modulation characteristics of the digital reference signal are consistent with the transmitted linear frequency modulated continuous wave signal.

[0068] The digital echo signal received within the entire observation window is multiplied with the digital reference signal in the digital domain.

[0069] A digital low-pass filter is used to filter out the high-frequency components generated by the multiplication operation, while retaining the difference frequency signal.

[0070] Specifically, the implementation process of this embodiment includes:

[0071] The de-skew method requires no analog circuitry, performing all processing solely in the digital domain, ensuring low-cost implementation while reducing analog mismatch errors and improving algorithm flexibility. Its key features are as follows:

[0072] like Figure 3 The single-reference signal design shown is as follows:

[0073] The length of the entire observation window Within this process, only one reference signal is used for de-chirping. This reference signal is defined as:

[0074] ;

[0075] in This represents the reference signal used for deskewing; its duration is equal to the length of the entire observation window. It is the same as the starting frequency and frequency modulation slope of the transmitted signal.

[0076] Implementation method:

[0077] The reference signal is pre-generated in the MCU's internal memory and is not input through an external analog mixer. The received signal is multiplied by this reference signal in the digital domain to complete the de-chewing operation.

[0078] ;

[0079] in This represents the result of the de-scratching operation. These are high-frequency components, which are removed using a digital filter. Representing the first The difference frequency signal of each target is represented as:

[0080] ;

[0081] in It is the first The propagation delay of each target For its corresponding distance, This represents the speed of sound in air. This frequency is directly related to the target distance, realizing a "distance-encoded" mapping relationship.

[0082] Low-pass filtering and time-domain results:

[0083] The deskewing result is filtered by a digital filter to remove high-frequency terms, resulting in a product containing only the low-frequency components corresponding to the target. , is represented as:

[0084] ;

[0085] in This indicates a low-pass FIR filter.

[0086] Furthermore, the periodic extension reconstruction processing of the difference frequency signal includes:

[0087] Difference frequency signal sequence of length N is divided into multiple consecutive segments, where each segment has a length of L, and L is less than N;

[0088] The signals of all segments are superimposed in the time domain to generate a reconstructed signal with a total length of L.

[0089] Specifically, the implementation process of this embodiment includes:

[0090] In the FMCW system, distance estimation relies on FFT spectral analysis of the de-angled signal. To cover a ranging range of several meters, the observation time... It often takes tens of milliseconds, corresponding to a large number of sampling points (e.g. or This is difficult to process in real time on low-performance MCUs.

[0091] To address this issue, this invention proposes a computational optimization scheme that utilizes the periodic extension characteristics of FFT, significantly reducing the computational burden of FFT, improving system real-time performance, and enabling the scheme to run stably on low-end MCUs (such as the STM32 series).

[0092] Segmented overlay reconstruction:

[0093] The length is The descrambled sequence is divided into several segments, each with a length of [length missing]. And perform periodic superposition in the time domain:

[0094] ;

[0095] in Indicates the index of the segmented signal. This indicates the index of the sampling point within each signal segment.

[0096] The length is obtained as Reconstruction sequence .

[0097] Single FFT operation:

[0098] right Execute once Point Fast Fourier Transform (FFT) yields the complete distance spectrum:

[0099] .

[0100] in This represents the result after FFT.

[0101] Furthermore, target distance estimation based on the distance spectrum includes:

[0102] Calculate the amplitude spectrum of the distance spectrum;

[0103] Detect the peak value in the amplitude spectrum and determine the frequency index corresponding to the peak value;

[0104] The distance value of the target is calculated based on the mapping relationship between the frequency index and the target distance.

[0105] Specifically, the implementation process of this embodiment includes:

[0106] Distance estimation:

[0107] right Perform amplitude detection and obtain the first value using the constant false alarm rate (CFAR) algorithm or threshold detection. Target peak index of each target And based on the relationship between frequency and delay:

[0108] ;

[0109] in Indicates the first Distance to each target The sampling frequency (e.g., 64kHz).

[0110] Furthermore, based on the spectral data of each channel in the microphone array corresponding to the target distance, target angle estimation includes:

[0111] Extract the complex value at the frequency index corresponding to the target distance from the FFT result of each microphone channel;

[0112] A spatial signal vector is formed by the complex values ​​of all channels;

[0113] The direction-of-arrival (DOA) estimation algorithm is applied to the spatial signal vector to calculate the target's angle information.

[0114] Specifically, the implementation process of this embodiment includes:

[0115] Extract the complex values ​​of each channel in the microphone array at the same index. The target space signal vector is constructed, and the direction of arrival (DOA) is estimated using the MUSIC or Capon algorithm to obtain the target angle.

[0116] The MCU outputs distance and angle information via serial port or Bluetooth, which can be interfaced with a host computer or robot navigation system.

[0117] like Figure 2 As shown, this embodiment provides an indoor target positioning system based on acoustic devices, including: a signal transmitting module for transmitting a linear frequency modulated continuous wave signal in the ultrasonic band through a loudspeaker;

[0118] The signal receiving and acquisition module is used to receive an analog signal containing the reflected echo of the target through a microphone array, and convert the analog signal into a digital echo signal;

[0119] The signal processing module is used to mix the digital echo signal with a pre-generated single digital reference signal with a duration equal to the observation window in the digital domain to complete the deskewing process, and to perform low-pass filtering on the mixing result to obtain a difference frequency signal; to perform periodic extension reconstruction processing on the difference frequency signal to form a reconstructed signal with a shortened length; to perform a Fast Fourier Transform (FFT) on the reconstructed signal to obtain a range spectrum; to perform target distance estimation based on the range spectrum; and to perform target angle estimation based on the spectral data of each channel in the microphone array corresponding to the target distance.

[0120] Furthermore, the signal processing module includes a descrambling unit; the descrambling unit is used to perform the descrambling process, the descrambling process including:

[0121] The digital reference signal is pre-generated and stored;

[0122] The digital echo signal and the digital reference signal are multiplied in the digital domain;

[0123] A digital low-pass filter is used to filter out the high-frequency components generated by the multiplication operation, while retaining the difference frequency signal.

[0124] Furthermore, the signal processing module includes a reconstruction unit; the reconstruction unit is used to perform the periodic extension reconstruction process, specifically including:

[0125] Difference frequency signal sequence of length N is divided into multiple consecutive segments, where each segment has a length of L, and L is less than N;

[0126] The signals of all segments are superimposed in the time domain to generate a reconstructed signal with a total length of L.

[0127] Furthermore, the signal processing module includes a distance estimation unit; the distance estimation unit is used to perform the target distance estimation, specifically including:

[0128] Calculate the amplitude spectrum of the distance spectrum;

[0129] Detect the peak value in the amplitude spectrum and determine the frequency index corresponding to the peak value;

[0130] The distance value of the target is calculated based on the mapping relationship between the frequency index and the target distance.

[0131] Furthermore, the signal processing module includes an angle estimation unit; the angle estimation unit is used to perform the target angle estimation, specifically including:

[0132] Extract the complex value at the frequency index corresponding to the target distance from the FFT result of each microphone channel;

[0133] A spatial signal vector is formed by the complex values ​​of all channels;

[0134] The direction-of-arrival (DOA) estimation algorithm is applied to the spatial signal vector to calculate the target's angle information.

[0135] This invention discloses an indoor target localization method and system based on acoustic devices. It utilizes a single digital reference signal to perform deskewing processing in the digital domain, completely eliminating the need for analog mixers and low-pass filters, significantly reducing hardware complexity and system cost, while avoiding errors introduced by analog circuits. By using a periodic extension reconstruction method to fold long signal sequences into shorter sequences for processing, the computational load of the Fast Fourier Transform is greatly reduced, enabling the entire algorithm to run in real-time on low-computing-power microcontrollers. Based on complex spectral data from a microphone array, direction-of-arrival estimation is achieved, giving the system complete range and angle joint estimation capabilities while reusing ordinary audio hardware, overcoming the limitation of traditional ultrasonic radar that can only measure range and not angle. Ultimately, a fully digital signal processing chain is realized, providing a feasible technical path for low-cost deployment of high-precision indoor positioning functions on existing smart devices.

[0136] The above are merely preferred embodiments of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A method for indoor target positioning based on acoustic devices, characterized in that, comprises: emitting a linear frequency modulated continuous wave signal in ultrasonic band through a speaker; receiving analog signals containing target reflected echoes through a microphone array, and converting the analog signals into digital echo signals; in the digital domain, mixing the digital echo signals with a single digital reference signal with the same length as the observation window to complete the dechirp process, and low-pass filtering the mixing result to obtain a difference frequency signal; performing a period extension reconstruction process on the difference frequency signal to form a reconstructed signal with a shortened length, and performing a fast Fourier transform (FFT) on the reconstructed signal to obtain a range spectrum; performing target distance estimation based on the range spectrum; performing target angle estimation based on the spectral data of each channel of the microphone array corresponding to the target distance.

2. The indoor target positioning method based on acoustic devices according to claim 1, wherein the dechirp process in the digital domain comprises: the digital reference signal is pre-generated and stored in a microcontroller, and the frequency modulation characteristic of the digital reference signal is consistent with the emitted linear frequency modulated continuous wave signal; the digital echo signals received in the entire observation window are multiplied with the digital reference signal in the digital domain; a digital low-pass filter is used to filter out high-frequency components generated by the multiplication operation and retain the difference frequency signal.

3. The indoor target positioning method based on acoustic devices according to claim 1, wherein the period extension reconstruction process on the difference frequency signal comprises: the difference frequency signal sequence with a length of N is divided into multiple continuous sub-segments, each with a length of L, and L is less than N; all the sub-segment signals are superimposed in the time domain to generate a reconstructed signal with a total length of L.

4. The indoor target positioning method based on acoustic devices according to claim 1, wherein the target distance estimation based on the range spectrum comprises: calculating the amplitude spectrum of the range spectrum; detecting the peak value in the amplitude spectrum and determining the frequency index corresponding to the peak value; calculating the distance value of the target according to the mapping relationship between the frequency index and the target distance.

5. The indoor target positioning method based on acoustic devices according to claim 4, wherein the target angle estimation based on the spectral data of each channel of the microphone array corresponding to the target distance comprises: extracting the complex value at the frequency index corresponding to the target distance from the FFT result of each microphone channel; constructing a spatial signal vector from the complex values of all channels; applying a direction of arrival estimation algorithm to the spatial signal vector to calculate the angle information of the target. The system for implementing the indoor target positioning method based on acoustic devices according to any one of claims 1-5 comprises: a signal emitting module for emitting a linear frequency modulated continuous wave signal in ultrasonic band through a speaker; a signal receiving and collecting module for receiving analog signals containing target reflected echoes through a microphone array, and converting the analog signals into digital echo signals; ​ 6. An acoustic device based indoor target positioning system, characterized by ​ ​ ​ The signal processing module is configured to mix the digital echo signal with a single digital reference signal with a length equal to the observation window in a digital domain to complete a dechirp process, and to perform low-pass filtering on a mixing result to obtain a difference frequency signal; to perform a period extension reconstruction process on the difference frequency signal to form a reconstructed signal with a shortened length; to perform a fast Fourier transform (FFT) on the reconstructed signal to obtain a range profile; to perform target distance estimation based on the range profile; and to perform target angle estimation based on spectral data corresponding to the target distance of each channel in the microphone array.

7. The acoustic device-based indoor target positioning system of claim 6, wherein the signal processing module comprises a dechirp unit configured to perform the dechirp process, and the dechirp process comprises: pre-generating and storing the digital reference signal; multiplying the digital echo signal with the digital reference signal in the digital domain; filtering out high-frequency components generated by the multiplication operation using a digital low-pass filter to retain the difference frequency signal.

8. The acoustic device-based indoor target positioning system of claim 6, wherein the signal processing module comprises a reconstruction unit configured to perform the period extension reconstruction process, and the reconstruction unit comprises: dividing a sequence of the difference frequency signal with a length of N into a plurality of continuous sub-segments, each with a length of L, and L is less than N; superimposing signals of all sub-segments in a time domain to generate a reconstructed signal with a total length of L.

9. The acoustic device-based indoor target positioning system of claim 6, wherein the signal processing module comprises a distance estimation unit configured to perform the target distance estimation, and the distance estimation unit comprises: calculating an amplitude spectrum of the range profile; detecting a peak value in the amplitude spectrum and determining a frequency index corresponding to the peak value; calculating a distance value of the target according to a mapping relationship between the frequency index and the target distance.

10. The acoustic device-based indoor target positioning system of claim 6, wherein the signal processing module comprises an angle estimation unit configured to perform the target angle estimation, and the angle estimation unit comprises: extracting a complex value at the frequency index corresponding to the target distance from an FFT result of each microphone channel; constructing a spatial signal vector from the complex values of all channels; and applying a direction of arrival estimation algorithm to the spatial signal vector to calculate angle information of the target. ​

Images