A method and system for measuring rotation speed at ultra-long distance based on LoRa signal

Abstract

The application discloses a kind of ultra-long distance rotating speed sensing measurement method and system based on LoRa signal.The method comprises the following steps: sending LoRa up-modulation signal as sensing signal by signal transceiver module, and capturing the signal reflected back by rotating target as sensing original data, while recording the signal at the time of sending as sensing reference signal;Signal alignment is carried out on sensing reference signal and sensing original data by signal demodulation module, and signal demodulation is carried out on sensing original data based on predefined LoRa down-modulation signal;Sensing enhancement module carries out sensing enhancement operation on demodulated sensing original data;The ratio of peak energy in the spectrum is calculated by feature processing module, and different frequencies are automatically distinguished according to adaptive threshold.The application can solve the problem of limited distance of traditional wireless sensing, and solve the problem of low sensing upper limit caused by limited chirp frequency of LoRa signal itself.

Description

TECHNICAL FIELD

[0001] The present application belongs to the field of wireless sensing research, and particularly relates to a method and system for realizing super-long-distance high-frequency rotating speed precise sensing measurement using LoRa signals, which provides a new solution for equipment monitoring and detection in the industrial Internet of Things scenario. BACKGROUND

[0002] Rotation is a basic form of motion that exists widely in daily life and industrial scenarios, and measuring its speed is crucial for monitoring the health of rotating machinery. To achieve this purpose, people have developed various rotating speed measurement systems, which are generally divided into two categories: contact and non-contact.

[0003] Contact systems require contact with the rotating shaft, but contact can cause serious safety problems when dealing with large or high-speed rotating targets. Existing non-contact systems mainly rely on electromagnetic signals, optical signals or wireless signals to measure rotating speed. Systems based on electromagnetic signals usually use electrostatic or Hall effect sensors to detect changes in electromagnetic fields caused by rotation, but their measurement distance is very limited. Optical systems use optical sensors such as cameras and laser receivers to measure rotating speed, but they are strictly dependent on light conditions and cannot work normally in environments with insufficient light. Some systems use wireless signals such as sound waves and radio frequency signals to measure rotating speed, but they still have limitations in long-distance measurement. When dealing with large mechanical equipment, the limited measurement distance of existing technologies can pose safety risks. Therefore, in order to reduce these safety risks, it is urgent to develop a rotating speed system that can accurately measure at long distances.

[0004] In recent years, LoRa signals have shown great potential in super-long-distance transmission in the industrial Internet of Things, and some research has also used LoRa signals to realize long-distance sensing. For example, LoRa signals can extend the range of breath sensing to 70 meters, demonstrating their potential in long-distance sensing. However, due to the low chirp rate of LoRa signals themselves, existing LoRa-based sensing systems are only suitable for low-frequency motion sensing, not high-frequency motion such as rotating motion. SUMMARY

[0005] The present application aims to provide a super-long-distance rotating speed sensing measurement method and system based on LoRa signals to solve the problem of limited distance in traditional wireless sensing and to solve the low sensing problem caused by the limited chirp frequency of LoRa signals themselves, providing a new solution for equipment monitoring and detection in the industrial Internet of Things scenario.

[0006] To achieve the above invention purposes, the technical solutions of the present application are as follows:

[0007] The first aspect is a long-distance rotating speed sensing measurement method based on LoRa signal, comprising the following steps:

[0008] The signal transceiver module sends a LoRa up-modulation signal as a sensing signal, and captures a signal reflected back by the rotating target encountered by the sensing signal as sensing original data, while recording the signal at the time of sending as a sensing reference signal;

[0009] The signal demodulation module aligns the sensing reference signal and the sensing original data, and demodulates the sensing original data based on a predefined LoRa down-modulation signal;

[0010] The sensing enhancement module performs a sensing enhancement operation on the demodulated sensing original data, which includes segmenting the received continuous signal, performing frequency domain analysis on each signal segment, and extracting the frequency position and phase information corresponding to the main peak value;

[0011] The feature processing module calculates the ratio of peak energy in the spectrum according to the sensing enhanced signal spectrum, and automatically distinguishes different frequencies according to the adaptive threshold.

[0012] Further, the signal demodulation module aligns the sensing reference signal and the sensing original data, and demodulates the sensing original data based on a predefined LoRa down-modulation signal, including:

[0013] Correlate the received sensing original data and the sensing reference signal to obtain the peak value when the two signals are aligned, determine the length of the LoRa symbol according to the position of the peak value, and cut the received signal into independent LoRa symbols;

[0014] Mix the aligned sensing original data and the predefined LoRa down-modulation signal to extract the baseband component of the signal, and perform low-pass filtering on the mixed signal to obtain the demodulated signal.

[0015] Further, the predefined LoRa down-modulation signal is a LoRa down-modulation signal generated according to preset parameters, and the preset parameters include one or more of bandwidth, chirp rate, and center frequency.

[0016] Further, the sensing enhancement operation includes segmenting the received continuous signal, performing frequency domain analysis on each signal segment, and extracting the frequency position and phase information corresponding to the main peak value, including:

[0017] The received continuous signal is segmented according to a specified time window, and the signal segment in each time window is used as an independent analysis unit.

[0018] Performing a Fast Fourier Transform on each signal segment converts the time-domain signal into a frequency-domain signal, obtaining the amplitude and phase information of the signal in the frequency domain, and generating a corresponding frequency spectrum graph;

[0019] Analyzing the generated frequency spectrum identifies the main frequency components and potential noise components in the frequency spectrum, and extracts the frequency position and phase information corresponding to the main peak values.

[0020] Further, after segmenting the received continuous signal according to the specified time window, it also includes: using the overlapping segmentation method, pre-processing each signal segment.

[0021] Further, according to the signal spectrum enhanced by perception, the ratio of peak energy in the spectrum is calculated, and different frequencies are automatically distinguished according to the adaptive threshold, including:

[0022] According to the reference signal, select the corresponding processing model, the selection of the processing model depends on the expected range of the target rotation speed;

[0023] If the target rotation speed falls within range 1, determine whether there is harmonic interference in the spectrum, if there is harmonic interference, select the lowest frequency peak as the final measurement result, if no harmonic interference is detected, select the frequency corresponding to the peak with the maximum energy in the spectrum as the representative frequency of the target rotation speed;

[0024] If the target rotation speed falls within range 2, and there are two obvious peaks in the spectrum, calculate the energy ratio between the highest peak and the surrounding noise, calculate the adaptive threshold according to the energy ratio, and calculate the ratio between the maximum peak energy and the second largest peak energy, if the ratio is greater than the calculated adaptive threshold, select the second largest peak as the final measurement result, otherwise, select the maximum peak as the measurement result; If there is only one obvious peak in the spectrum, directly select the peak as the measurement result of the rotation speed.

[0025] Further, the calculation formula of the adaptive threshold according to the energy ratio is m x 10 m , m is the energy ratio.

[0026] Further, the target rotation speed range 1 is the rotation speed less than or equal to 40 revolutions per minute; the target rotation speed range 2 is the rotation speed greater than 40 revolutions per minute.

[0027] The second aspect is a super-long-distance rotation speed perception measurement system based on LoRa signal, including:

[0028] The signal transceiver module is used for sending a LoRa up-modulation signal as a sensing signal and capturing a signal reflected back by the rotating target encountered by the sensing signal as sensing original data, while recording the signal at the time of sending as a sensing reference signal;

[0029] The signal demodulation module is used for signal alignment of the sensing reference signal and the sensing original data, and signal demodulation of the sensing original data based on a predefined LoRa down-modulation signal;

[0030] The sensing enhancement module is used for sensing enhancement operation on the demodulated sensing original data, and the sensing enhancement operation includes segment processing on the received continuous signal, frequency domain analysis on each signal segment, and extraction of frequency position and phase information corresponding to the main peak value;

[0031] The feature processing module is used for calculating the ratio of peak energy in the spectrum according to the sensing enhanced signal spectrum, and automatically distinguishing different frequencies according to the adaptive threshold.

[0032] Beneficial effects: The present application proposes a super-long distance sensing measurement method and system based on LoRa signal, which realizes the accurate measurement of super-long distance rotating speed through innovative technical design. First of all, the sensing enhancement module of the present application significantly improves the sampling rate of the signal, so that it can maintain high accuracy in long distance sensing. Secondly, the feature processing module further optimizes the signal processing process, accurately identifies the target rotating frequency in the spectrum, and effectively eliminates the influence of environmental interference and noise. By integrating the above modules, the present application not only improves the overall robustness of the system, but also ensures accurate super-long distance rotating speed measurement in complex environments. Experimental evaluation shows that the method of the present application can realize reliable sensing within a distance of 50 meters, and maintain an average measurement error of less than 0.69% at various rotating speeds. This breakthrough provides a new solution for device monitoring and detection in industrial Internet of Things scenarios, significantly improving the application potential and actual efficiency of wireless sensing technology. BRIEF DESCRIPTION OF DRAWINGS

[0033] Figure 1 It is the overall architecture schematic diagram of the present application;

[0034] Figure 2 It is the signal demodulation module flowchart in the present application;

[0035] Figure 3 It is the sensing enhancement module schematic diagram in the present application;

[0036] Figure 4 It is the adaptive threshold schematic diagram of the feature processing module in the present application;

[0037] Figure 5A schematic diagram of measurement error at different rotational speeds in the present application. DETAILED DESCRIPTION

[0038] The present application proposes a super-long distance rotational speed sensing measurement system based on LoRa signals. This system overcomes the limitations of traditional technology through innovative signal processing methods. The present application designs four novel signal processing modules. Referring to Figure 1 , first, the present application uses a signal transceiver module and a signal demodulation module to preprocess the received signals. Next, the signal-to-noise ratio (SNR) of the signals is further improved through the sensing enhancement module, and the sampling rate of the sensing is significantly improved. Finally, through the feature processing module, the preprocessed signals are accurately analyzed and calculated to obtain the rotational speed of the target. Through the close cooperation of these four modules, the present application not only enables accurate rotational speed sensing measurement at a long distance in the presence of environmental interference and obstacles, but also significantly improves the overall performance and stability of the system.

[0039] (I) Signal transceiver module

[0040] In order to achieve non-contact rotational speed measurement, the present application transmits a LoRa up-modulation signal as a sensing signal through the signal transceiver module. This signal, after being transmitted, will encounter a rotating target and be reflected back. The signal transceiver module then captures these reflected signals as sensing raw data, while recording the signal at the time of transmission as a sensing reference signal. These two signal data are then output by the signal transceiver module, preparing for further signal processing and rotational speed calculation. The signal transceiver module includes a data acquisition unit that records and outputs raw data and reference signals to facilitate subsequent signal alignment and demodulation processing, thereby ensuring the continuity of signal processing and the integrity of data. This process ensures that the system can accurately sense and measure the rotational speed of the target without direct contact with the target, thereby avoiding the risks and limitations of traditional contact measurement methods.

[0041] According to the embodiments of the present application, the signal transceiver module includes a USRP (Universal Software Radio Peripheral) device that generates and transmits LoRa up-modulation signals through a flexible software-defined radio platform, and receives signals reflected by the target object. The USRP device accurately captures signals through its high-sensitivity receiving function, and performs preliminary digital processing on the signals to ensure the accuracy of subsequent signal demodulation and feature extraction.

[0042] (II) Signal demodulation module

[0043] In order to reduce the signal attenuation caused by long distance, the received signal needs to be demodulated by LoRa to improve the signal-to-noise ratio (SNR). The signal demodulation module is proposed to achieve this purpose. Signal demodulation is a key step in the entire sensing system, which directly affects the quality of the signal and the accuracy of subsequent processing. The main function of this module is to improve the signal-to-noise ratio of the signal by accurately aligning, demodulating and filtering the received LoRa signal.

[0044] Firstly, the signal needs to be aligned and the LoRa symbol needs to be accurately extracted for subsequent demodulation. If the signal is not aligned, the signal will not only retain the influence of chirp spread spectrum modulation, but also introduce new frequency interference, thereby affecting the extraction of signal features. After signal alignment, the influence of chirp spread spectrum modulation (CSS) will be completely eliminated, and the system will be able to extract pure signal features. Further, the received signal and the reference signal are cross-correlated. When the signals are aligned, the cross-correlation operation will produce a peak, and the abscissa of the peak represents the starting position of the first LoRa symbol. The signal demodulation module will track the position of each peak and divide the received signal into independent LoRa symbols according to the symbol length. For each LoRa symbol in the received signal, the present invention performs a de-chirp operation. The signal energy in the entire frequency domain bandwidth can be concentrated on a specific frequency, thereby solving the problem of low signal-to-noise ratio and providing a stable signal for processing. The processing logic of the signal demodulation module is shown in Figure 2 The de-chirped signal will be further processed in the subsequent module to extract signal features and calculate the rotational speed.

[0045] According to the embodiment of the present invention, the signal demodulation module is integrated into a computer terminal equipped with a high-performance digital signal processor (DSP) and dedicated software for performing signal alignment, demodulation and energy concentration processing. By aligning the reference signal with the original data, the phase offset is eliminated, and the demodulation of the signal is realized by using the predefined LoRa down-modulation signal, which improves the signal-to-noise ratio of the signal and significantly improves the stability and reliability of the long-distance sensing signal.

[0046] Specifically, the specific processing logic of the signal demodulation module of the present invention is as follows:

[0047] Step 1: Signal alignment. Firstly, the received signal and the reference signal are cross-correlated. Through cross-correlation operation, when the two signals are aligned, a clear peak will be generated in the result. The position of this peak identifies the starting point of the first LoRa symbol. The present invention uses the position of this peak to track the alignment state of the signal, and according to the length of each LoRa symbol, the received signal is divided into independent LoRa symbols.

[0048] Step 2: LoRa down-modulation signal generation. According to the pre-set parameters (such as bandwidth, chirp rate, center frequency, etc.), a LoRa down-modulation signal that matches the received signal is generated. This generated down-modulation signal will serve as the reference signal for demodulation, preparing to demodulate the aligned original data.

[0049] Step 3: Signal demodulation. The aligned original data is mixed with the generated LoRa down-modulation signal (i.e., multiplied by each other) to extract the baseband component of the signal. Then, the mixed signal is low-pass filtered to remove high-frequency components and retain low-frequency parts, thereby obtaining the demodulated signal.

[0050] (Three) Perception enhancement module

[0051] Since the sampling rate of the extracted signal features does not meet the requirements of the Nyquist sampling theorem, the traditional LoRa perception method faces significant limitations in dealing with high-frequency motion. The Nyquist sampling theorem states that in order to accurately reconstruct a signal, the sampling rate must be at least twice the highest frequency component in the signal. If the sampling rate is lower than this standard, aliasing effects will occur, making it impossible to correctly capture and analyze the information of high-frequency motion, affecting subsequent measurement calculations.

[0052] A relatively direct solution to this problem is to increase the sampling rate by reducing the spreading factor of the LoRa signal or increasing its bandwidth. The spreading factor is a key parameter in LoRa signal modulation, and a lower spreading factor can increase the data transmission rate, thereby indirectly increasing the sampling rate. Increasing the bandwidth directly expands the frequency domain range that the signal can cover, allowing high-frequency components to be captured more clearly.

[0053] However, implementing this solution is not easy. First of all, such adjustments require specially designed LoRa chips to support flexible parameter configuration. In addition, although this method can increase the sampling rate, when extracting phase information directly from the original de-chirped signal, a large amount of noise is often introduced into the signal features. This is because the original signal not only contains the information of the target motion, but also contains environmental noise, interference, and other non-relevant signal components. When these noises are mixed with the useful signal, they can severely affect the quality and reliability of the signal, making subsequent signal processing and feature extraction more difficult, and even possibly leading to incorrect perception results. Therefore, although increasing the sampling rate seems to be an effective solution, it faces technical and cost challenges in practical applications, and may also introduce new problems such as increased signal noise. These factors together limit the application range of traditional LoRa perception methods in high-frequency motion detection, prompting the present applicant and inventor to explore new solutions to overcome these technical obstacles.

[0054] To solve this problem, the inventors of the present application propose a perception enhancement module that processes the received signal in segments and performs a fast Fourier transform on each signal segment to concentrate the signal's energy on specific frequencies in the frequency domain. After that, the phase information is extracted from each signal segment as the signal feature of that segment. The principle diagram of the perception enhancement module is shown in Figure 3 With these improvements, the LoRa system can reliably detect and measure high-frequency rotational motion in complex environments, providing higher precision and robustness for practical applications.

[0055] According to the embodiments of the present application, the perception enhancement module is connected to a computer terminal, and realizes real-time transmission and processing of large-capacity data through a high-speed data interface such as USB 3.0 or PCIe. The module uses a specific algorithm to process LoRa signals, significantly improves the sampling rate of signal perception, breaks through the inherent chirp frequency limit of LoRa signals, and thus enhances the perception ability of the system to high-frequency rotational motion, making it suitable for a wider range of industrial application scenarios.

[0056] The main function of the perception enhancement module in the present application is to achieve accurate signal feature extraction through a series of signal processing steps, while effectively eliminating environmental noise and other interference factors. The design and implementation of this module not only improves the sensitivity of the system to rotational motion, but also ensures that the system can maintain stable and accurate performance under changing environmental conditions.

[0057] The specific implementation logic of the perception enhancement module is as follows:

[0058] Step 1: Signal segmentation. The received continuous signal is segmented according to a certain time window. Each signal segment in the time window will be an independent analysis unit. In order to avoid spectral leakage, the method of overlapping segmentation is used to preprocess each signal segment (such as windowing), to reduce the influence of spectral edge effect and ensure the accuracy of subsequent frequency domain analysis.

[0059] Step 2: Frequency domain analysis. Perform a fast Fourier transform (FFT) on each preprocessed signal segment to convert the time domain signal to the frequency domain signal. Through FFT, the amplitude and phase information of the signal in the frequency domain is obtained, and the corresponding frequency spectrum is generated. The generated frequency spectrum is analyzed preliminarily to identify the main frequency components and potential noise components in the spectrum, preparing for the subsequent peak extraction.

[0060] Step 3: Frequency peak phase extraction. Extract the frequency position and phase information corresponding to the main peak. At the same time, use the extracted phase information to further analyze the phase change characteristics of the signal, and provide accurate phase data for subsequent rotational speed calculation. The phase information of these peaks is used as a key feature and is passed to the feature processing module for high-precision rotational speed calculation.

[0061] The method is not limited to LoRa signal types and is applicable to any LoRa chip and further improves the perceived signal-to-noise ratio.

[0062] (iv) Feature processing module

[0063] According to the findings of the current inventor during the research process, the chirp frequency of the LoRa signal itself will cause significant interference effects in the frequency spectrum. This interference is not only strong, but also difficult to ignore. We call this frequency the "perceived boundary frequency". Specifically, the perceived boundary frequency is determined by the inherent characteristics of the LoRa signal and is a fixed frequency value directly determined by the parameters in the signal modulation process. Although this frequency is theoretically fixed and can be explicitly calculated and identified, in actual operation, it is not simple to handle this problem. Directly filtering out the signal component corresponding to the perceived boundary frequency seems to be an ideal solution, but this approach will bring a series of new problems. First, since the chirp modulation of the LoRa signal is a process of gradually changing the frequency, filtering out the perceived boundary frequency may cause misfiltering of other important frequency components, thereby affecting the integrity of the overall signal. Second, the existence of the perceived boundary frequency is intertwined with other signal characteristics, and simple filtering processing can easily damage the phase information and spectral structure of the signal, thereby directly affecting the accuracy and stability of the final measurement results.

[0064] To minimize the impact of the boundary frequency on the measurement results, the present invention divides the entire speed range into two parts: range 1 (rotational speed less than or equal to 40 revolutions / minute) and range 2 (rotational speed greater than 40 revolutions / minute). Specifically, the present invention selects a LoRa signal with a boundary frequency close to 60Hz to measure the rotational speed in range 1, and selects a LoRa signal with a boundary frequency close to 30Hz to measure the rotational speed in range 2. The purpose of this is to ensure that the boundary frequency does not appear in the corresponding measurement range.

[0065] The application proposes a feature processing module to achieve high-precision measurement calculation. The core principle of this module is to calculate the ratio of peak energy based on spectral analysis and determine the physical meaning of the peak in the spectrum according to the adaptive threshold. Through experimental analysis, it is found that there are mainly three key frequencies in the spectrum: target rotation frequency, perception boundary frequency and harmonic frequency of these frequencies. In order to accurately distinguish these three frequencies, the feature processing module will analyze the spectrum of the transmitted signal and determine the appropriate processing mode according to the characteristics of these frequency components, so as to accurately calculate the target rotation frequency.

[0066] In order to improve the discrimination accuracy, the module sets adaptive thresholds. These thresholds are dynamically adjusted according to the actual environment and signal characteristics, so that the target frequency and interference frequency in the spectrum can still be effectively distinguished under different measurement conditions. Through this method, the feature processing module can accurately identify and calculate the rotation frequency in a complex signal environment, ensuring the accuracy and reliability of the measurement. In order to prove the rationality and reliability of the adaptive threshold setting, we conducted a measurement experiment on the speed of the target (10cm long propeller) in the range of 1-15m, and collected 100 groups of data to evaluate the effect of adaptive threshold. Figure 4 The identification results after applying the adaptive threshold are shown. Among them, the horizontal axis represents the ratio of the highest peak value in the frequency domain to the noise (i.e. peak reference), the vertical axis represents the ratio of the highest peak value in the frequency domain to the second highest peak (i.e. peak ratio), the purple dotted line represents the calculated adaptive threshold, the blue dot represents the correct selection of the highest peak in the frequency domain as the calculation result, the yellow dot represents the correct selection of the second highest peak in the frequency domain as the calculation result, and the red dot represents the wrong identification result. From Figure 4 It can be seen that only 4 groups of data have calculation errors in 100 groups of data, and these four groups of data are all measurement results when the test distance is 15m. This is due to the fact that when the distance is too far, the reflected signal of the small target is too weak. This test can prove the rationality and reliability of the adaptive threshold setting.

[0067] According to the embodiment of the application, the feature processing module is integrated in a computer terminal, and a high-level spectral analysis tool is used to perform comprehensive frequency domain analysis on the captured signal. The module automatically distinguishes different frequencies by calculating the ratio of peak energy in the spectrum and according to the adaptive threshold. In this way, the system can accurately identify the target rotation speed in a complex signal environment, ensuring high precision and reliability of the measurement.

[0068] The present application executes a feature processing module for accurate rotational speed calculation. The core of this module lies in identifying the physical meaning of different frequency peaks in the frequency spectrum through precise spectral analysis and accurately determining the peak corresponding to the target rotational frequency, thereby achieving high-precision rotational speed measurement. The design of the feature processing module not only focuses on identifying the main frequency components in the frequency spectrum, but also takes into account the complex interference and noise factors that may exist in the signal. This process not only improves the accuracy of rotational speed measurement, but also ensures the reliability of the system in various industrial application scenarios.

[0069] The specific processing logic of the feature processing module is as follows:

[0070] Step 1: Spectral analysis and processing model selection. According to the reference signal, select the corresponding processing model. The selection of the processing model depends on the expected range of the target rotational speed. Different rotational speed ranges may correspond to different frequency distribution characteristics, and the selection of the processing model will directly affect the subsequent rotational speed calculation. In order to preliminarily determine the approximate target rotational speed range, a LoRa signal with a boundary frequency close to 60Hz can be used to determine whether the target rotational frequency is in range 1 (preset low frequency range). If the measurement result presents obvious periodicity and the frequency domain feature presents obvious peak information, the target rotational speed is in range 1 (preset low frequency range). Otherwise, the target rotational speed is determined to be in range 2 (preset high frequency range).

[0071] Step 2: Signal processing mode 1. If the target rotational speed falls within range 1 (preset low frequency range), the system will first determine whether there is harmonic interference in the frequency spectrum. Harmonic interference usually manifests as multiple peaks with frequencies that are integer multiples of the fundamental frequency. If there is harmonic interference, the system will select the fundamental frequency signal (i.e. the lowest frequency peak) as the final measurement result to avoid the influence of harmonic interference on rotational speed calculation. If no harmonic interference is detected in the frequency spectrum, the system will select the frequency corresponding to the peak with the maximum energy in the frequency spectrum as the representative frequency of the target rotational speed.

[0072] Step 3: Signal processing mode 2. If the target rotational speed falls within range 2 (preset high frequency range) and two obvious peaks appear in the frequency spectrum, the system will further analyze the relationship between these peaks. First, calculate the energy ratio between the highest peak and the surrounding noise, denoted as m. Using the energy ratio m, calculate the adaptive threshold, which is calculated as m x 10 mThis is to determine which peak in the spectrum is more likely to correspond to the actual rotational speed. Then, the ratio between the maximum peak energy and the second maximum peak energy is calculated. If this ratio is greater than the adaptive threshold calculated, the system will select the second maximum peak as the final measurement result, considering it closer to the actual rotational speed; otherwise, the maximum peak is selected as the measurement result. If there is only one obvious peak in the spectrum, the system will directly select this peak as the measurement result of the rotational speed, as there is no other competing frequency component.

[0073] Step 4: Calculate the result output. After completing the above processing, the system selects the frequency corresponding to the selected peak as the final rotational speed calculation result. This rotational speed result is fed back to the user or other system modules for device state monitoring or further processing and analysis.

[0074] This feature processing method enables the system to find the target rotation-induced changes in the frequency domain and accurately extract them. It solves the problem of coexistence of LoRa's own frequency and target rotational speed in the spectrum, enabling the system to accurately perceive the rotational speed of the measurement target.

[0075] In the system, the computer terminal is equipped with a high-performance processor (such as a multi-core CPU or GPU) and a large-capacity memory to support real-time processing and storage of large-scale data. The terminal is also equipped with a graphical user interface (GUI) that displays the measurement results of rotational speed and related spectrum analysis information in real time through intuitive visualization tools, providing clear and accurate feedback to operators to support efficient monitoring and fault diagnosis of devices in industrial Internet of Things scenarios.

[0076] To verify the performance of the method of the present application, experimental tests were conducted. In the examples, the measurement performance of the system for different rotational speeds (100RPM-5100RPM) was tested, with a test distance of 6m, and the test results are shown in Figure 5 As can be seen, the system of the present application has a measurement error of less than 1.8% (MAPE represents the average absolute percentage error) for different rotational speeds. At the same time, using different signal processing modes, the system of the present application can achieve measurement of a very wide range of rotational speeds. Through experimental testing, it can be seen that the present application can solve the problem of limited sensing distance of traditional wireless sensing, and also breaks through the low sensing upper limit problem caused by the limited chirp frequency of LoRa signals. The present application ultimately uses LoRa signals to achieve precise sensing and measurement of high-frequency rotational speeds over long distances, providing a new solution for device monitoring and detection in industrial Internet of Things scenarios.

[0077] It should be pointed out finally that the above embodiments are only used to illustrate the technical solutions of the present application but not to limit it. Although the present application has been described in detail with reference to the above embodiments, it should be understood by those skilled in the art that the specific embodiments of the present application can be modified or replaced equivalently without departing from the spirit and scope of the present application, and any modification or equivalent replacement should be covered in the protection scope of the claims of the present application.

Claims

1. A method for ultra long distance rotational speed sensing based on LoRa signals, characterized in that, The method comprises the following steps: sending a LoRa up-modulation signal as a sensing signal through a signal transceiver module, and capturing a signal reflected back by the sensing signal encountering a rotating target as sensing original data, while recording the signal at the time of sending as a sensing reference signal; performing signal alignment on the sensing reference signal and the sensing original data through a signal demodulation module, and performing signal demodulation on the sensing original data based on a predefined LoRa down-modulation signal, specifically including: performing cross-correlation operation on the received sensing original data and the sensing reference signal to obtain a peak value when the two signals reach an aligned state, determining the length of a LoRa symbol according to the position of the peak value, and cutting the received signal into independent LoRa symbols; performing mixing processing on the aligned sensing original data and the predefined LoRa down-modulation signal to extract the baseband component of the signal, and performing low-pass filtering on the mixed signal to obtain a demodulated signal; performing sensing enhancement operation on the demodulated sensing original data through a sensing enhancement module, the sensing enhancement operation including segmenting the received continuous signal, performing frequency domain analysis on each signal segment, and extracting the frequency position and phase information corresponding to the main peak value, specifically including: segmenting the received continuous signal according to a specified time window, and taking each signal segment in each time window as an independent analysis unit; performing fast Fourier transform on each signal segment to convert the time domain signal into a frequency domain signal, obtain the amplitude and phase information of the signal in the frequency domain, and generate a corresponding frequency spectrum; analyzing the generated frequency spectrum to identify the main frequency component and potential noise component in the frequency spectrum, and extracting the frequency position and phase information corresponding to the main peak value; calculating the ratio of peak energy in the frequency spectrum according to the sensing enhanced signal spectrum through a feature processing module, and automatically distinguishing different frequencies according to an adaptive threshold, specifically including: selecting a corresponding processing model according to the reference signal, and the selection of the processing model depends on the expected range of the target rotating speed; if the target rotating speed falls within range 1, determining whether there is harmonic interference in the frequency spectrum, if there is harmonic interference, selecting the peak value of the lowest frequency as the final measurement result, if no harmonic interference is detected, selecting the frequency corresponding to the peak value with the maximum energy in the frequency spectrum as the representative frequency of the target rotating speed; the range 1 is a preset low frequency range; if the target rotating speed falls within range 2, and there are two obvious peak values in the frequency spectrum, calculating the energy ratio between the highest peak value and the surrounding noise, calculating the adaptive threshold according to the energy ratio, and calculating the ratio between the maximum peak energy and the second maximum peak energy, if the ratio is greater than the calculated adaptive threshold, selecting the second maximum peak as the final measurement result, otherwise, selecting the maximum peak as the measurement result; if there is only one obvious peak value in the frequency spectrum, directly selecting the peak value as the measurement result of the rotating speed; the range 2 is a preset high frequency range.

2. The method of claim 1, wherein, The predefined LoRa down-modulation signal is a LoRa down-modulation signal matched with the received signal according to preset parameters, and the preset parameters include one or more of bandwidth, chirp rate and center frequency.

3. The method of claim 1, wherein, After the received continuous signal is segmented according to the specified time window, the method further comprises: using an overlapping segmentation method to pre-process each signal segment.

4. The method of claim 1, wherein, The calculation formula of the adaptive threshold according to the energy ratio is , and m is the energy ratio.

5. The method of claim 1, wherein, The target rotating speed range 1 is less than or equal to 40 rpm, and the target rotating speed range 2 is greater than 40 rpm.

6. A system for ultra long range rotational speed sensing based on LoRa signals, characterized in that, The method is used for the LoRa signal-based ultra-long distance rotating speed sensing measurement method as claimed in any one of claims 1-5, comprising: The signal transceiver module is used for sending a LoRa up-modulation signal as a sensing signal, and capturing a signal reflected back by the rotating target as the sensing original data, while recording the signal at the time of sending as a sensing reference signal; The signal demodulation module is used for signal alignment of the sensing reference signal and the sensing original data, and signal demodulation of the sensing original data based on a predefined LoRa down-modulation signal; The sensing enhancement module is used for sensing enhancement operation on the demodulated sensing original data, and the sensing enhancement operation comprises segmenting the received continuous signal, performing frequency domain analysis on each signal segment, and extracting frequency position and phase information corresponding to the main peak value; The feature processing module is used for calculating the ratio of peak energy in the spectrum according to the sensing enhanced signal spectrum, and automatically distinguishing different frequencies according to the adaptive threshold.

Images