Time-varying bandpass filtering method under linear frequency‑modulated excitation
By constructing a linear frequency modulated signal with an inverse slope and multiplying it with the target signal, converting it into an approximately steady-state signal and performing bandpass filtering, the problem of severe interference of linear frequency modulated signals in complex environments is solved, achieving efficient signal recovery and feature extraction, and improving the reliability of detection.
Patent Information
- Application Number
- PCT/CN2026/081364
- Authority / Receiving Office
- WO · WO
- Patent Type
- Applications
- Current Assignee / Owner
- Priority Date
- 2025-05-07
- Filing Date
- 2026-03-04
- Publication Date
- 2026-11-12
AI Technical Summary
The echo signal received by linear frequency modulated signal in complex environment often contains noise and high-order response interference, which leads to detection error and task failure. Traditional time-varying bandpass filters have large computational load and long time extension under large bandwidth conditions, which reduces the reliability of detection results.
By constructing a linear frequency modulated signal with an inverse slope and multiplying it with the target signal, an approximate steady-state signal is generated. This signal is then filtered using a bandpass filter and multiplied with the complex conjugate signal of the inverse slope signal to recover the linear frequency modulated excitation and its response. This avoids the bandwidth scanning process of traditional time-varying bandpass filters and reduces the computational load.
It significantly reduces the computational load of the filtering process, effectively suppresses interference components, preserves the characteristics of linear frequency modulation excitation and its response, and provides high-quality analyzable signals, laying the foundation for subsequent algorithm processing.
Smart Images

Figure CN2026081364_12112026_PF_FP_ABST
Abstract
Description
A time-varying bandpass filtering method under linear frequency modulation excitation Technical Field
[0001] This invention belongs to the field of signal processing technology, specifically relating to a time-varying bandpass filtering method under linear frequency modulation excitation. Background Technology
[0002] Linear frequency modulated (LFM) signals, whose frequency varies linearly with time, are widely used as excitation signals in radar ranging, oil well dynamic fluid level detection, pipeline blockage location, electromagnetic detection, and fiber optic sensing due to their advantages such as ultra-widebandwidth, large time-delay-bandwidth product, and ease of generation. In practical applications, an LFM signal is generated and emitted by an excitation source, propagates in the target environment, and interacts with the target or medium. The echo signal or specific response carrying target information is then collected by a receiving device. Signal processing algorithms extract feature information and key parameters, thereby calculating parameters such as target distance and medium characteristics to achieve engineering tasks such as distance measurement, target location, and imaging recognition. However, due to the influence of complex background noise in the environment, non-ideal characteristics of the transmission channel, nonlinear response of the transceiver equipment, and resonance effect of the excitation source, the received echo signal often contains noise and high-order response interference, easily drowning out the useful signal and severely affecting the extraction and identification analysis of target information, leading to detection errors or even mission failure. Therefore, before processing the collected signal using algorithms, filtering preprocessing is required to improve the signal's analyzability and anti-interference capability.
[0003] For linear frequency modulation (LFM) excitation, time-varying bandpass filters are commonly used. However, due to the time-varying frequency characteristics of LFM signals, the center frequency of the time-varying bandpass filter needs to follow the instantaneous frequency changes of the LFM signal in real time to scan the entire excitation band. Each time the frequency changes, the filtering process is equivalent to reconstructing a new time-varying bandpass filter. Therefore, when the bandwidth is set large, the filtering process generates a significant computational load, leading to a substantial increase in filtering delay, reduced feature extraction accuracy and precision, and consequently affecting the reliability of the detection results. Summary of the Invention
[0004] To improve the reliability of linear frequency modulated (LFM) excitation filtering during oil well dynamic fluid level measurement, this invention proposes a time-varying bandpass filtering method under LFM excitation. During oil well dynamic fluid level measurement, an LFM signal is sent to excite the air column inside the tubing. The excited acoustic signal is collected and processed. The fluid level depth is measured based on the mathematical relationship between the extracted resonance characteristics and the tubing length. The processing of the sensor signal obtained through continuous resonance of the LFM acoustic wave includes the following steps:
[0005] S1. Acquire the sensing signal and obtain the target signal through discrete sampling;
[0006] S2. Construct a linear frequency modulated signal with an inverse slope and multiply it with the target signal to obtain an approximate steady-state signal;
[0007] S3. Use a bandpass filter to filter the approximate steady-state signal to obtain the bandpass filtered steady-state signal;
[0008] S4. Multiply the steady-state signal after bandpass filtering with the complex conjugate signal of the inverse slope linear frequency modulation signal to recover the linear frequency modulation excitation and its response.
[0009] Furthermore, the target signal obtained by discrete sampling include:
[0010] ;
[0011] in, The linear frequency modulation excitation and its response are expressed as:
[0012] ;
[0013] The higher-order linear frequency modulated excitation and its response are expressed as:
[0014] ;
[0015] in, Noise originating from the measurement system and the environment; e is the amplitude; e is the natural constant; j is the imaginary unit; The sampling frequency; For sampling sequences; Indicates the initial frequency; Indicates the slope; This is the initial phase; This represents the response to linear frequency modulation excitation; It represents the order of the higher-order response; This represents the response to a higher-order linear frequency modulated excitation.
[0016] Furthermore, the constructed inverse slope linear frequency modulated signal The formula is:
[0017]
[0018] in, This indicates the frequency adjustment.
[0019] Furthermore, the approximate steady-state signal The calculation formula is:
[0020]
[0021] Furthermore, the steady-state signal after bandpass filtering The formula is:
[0022] ;
[0023] in, This indicates in-band noise.
[0024] Furthermore, complex conjugate signal The calculation formula is:
[0025] .
[0026] Furthermore, the recovered linear frequency modulation excitation and its response The formula is:
[0027] .
[0028] Preferably, the sampling frequency Set to 1000Hz.
[0029] Preferably, the slope is set to 3 Hz / s.
[0030] Preferably, the bandpass filter used when filtering the near steady-state signal has a center frequency of 20Hz and a bandwidth of 5Hz.
[0031] The beneficial effects of this invention are:
[0032] This invention constructs a linear frequency modulated (LFM) signal with opposite modulation slopes and inverse instantaneous frequency changes, transforming the target LFM signal into an approximately steady-state signal. Interference components are filtered out using a conventional bandpass filter, avoiding the overall excitation band scanning process required by traditional time-varying bandpass filters. This significantly reduces the computational load and filtering time, effectively suppressing interference while preserving the characteristics of the LFM excitation and its response relatively completely, providing a high-quality analyzable signal for subsequent algorithm processing and feature extraction. Attached Figure Description
[0033] Figure 1 is a flowchart of the method of the present invention;
[0034] Figure 2 is a time-domain diagram of the target linear frequency modulated signal in the embodiment;
[0035] Figure 3 is a frequency domain diagram of the target linear frequency modulated signal in the embodiment;
[0036] Figure 4 is a time-frequency diagram of the target linear frequency modulated signal in the embodiment;
[0037] Figure 5 is the time-frequency diagram of the approximate steady-state signal obtained by multiplying the target linear frequency modulated signal and the inverse slope linear frequency modulated signal in the embodiment.
[0038] Figure 6 shows the time-frequency diagram of the recovered linear frequency modulated signal obtained by multiplying the approximate steady-state signal after bandpass filtering with the complex conjugate signal of the linear frequency modulated signal with inverse modulation slope in the embodiment.
[0039] Figure 7 shows the recovered linear frequency modulation excitation and its response in the time domain diagram of the embodiment;
[0040] Figure 8 shows the frequency domain diagram of the recovered linear frequency modulation excitation and its response in the embodiment. Detailed Implementation
[0041] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.
[0042] This invention proposes a time-varying bandpass filtering method under linear frequency modulation excitation. In the process of measuring the dynamic fluid level in an oil well, a linear frequency modulation signal is sent to excite the air column inside the pipe. The acoustic signal after excitation is collected and processed. The fluid level depth is measured based on the mathematical relationship between the extracted resonance characteristics and the pipe length. The processing of the sensor signal obtained through continuous resonance of the linear frequency modulation sound wave includes the following steps:
[0043] S1. Acquire the sensing signal and obtain the target signal through discrete sampling;
[0044] S2. Construct a linear frequency modulated signal with an inverse slope and multiply it with the target signal to obtain an approximate steady-state signal;
[0045] S3. Use a bandpass filter to filter the approximate steady-state signal to obtain the bandpass filtered steady-state signal;
[0046] S4. Multiply the steady-state signal after bandpass filtering with the complex conjugate signal of the inverse slope linear frequency modulation signal to recover the linear frequency modulation excitation and its response.
[0047] Oil well dynamic fluid level measurement is an important application scenario for linear frequency modulated (LFM) excitation methods. This involves sending a LFM signal to excite the air column inside the tubing, collecting the excited acoustic signal, processing it, and then measuring the fluid level depth based on the mathematical relationship between the extracted resonance characteristics and the tubing length. In this embodiment, the collected tubing acoustic signal obtained through continuous resonance of the LFM sound wave is filtered using a time-varying bandpass filtering method under LFM excitation mentioned in this invention.
[0048] Install the measuring device; place the excitation sound source at the inlet of the experimental pipe, and install the sound sensor at the front end of the excitation sound source. Set the signal sampling frequency f. s =1000Hz, total sampling time T=10s, linear frequency modulation excitation starting frequency f L =10Hz, termination frequency f U =40Hz, tuning slope k=3Hz / s.
[0049] As shown in Figure 1, the acquired sensor signal is discretely sampled to obtain the target signal. The target signal is multiplied by the inverse-slope linear frequency modulated signal to obtain an approximate steady-state signal, which is then filtered using a bandpass filter. The bandpass-filtered steady-state signal is then multiplied by the complex conjugate signal of the inverse-slope linear frequency modulated signal to obtain the recovered linear frequency modulated excitation and its response. The following embodiments will describe the above steps in detail.
[0050] Acoustic signals obtained from continuous excitation by linear frequency modulated sound waves are acquired, and the target signal is obtained through discrete sampling. The time-domain plot is shown in Figure 2. It can be seen that the useful signal reflecting the resonance characteristics is completely submerged. Figure 3 is a frequency domain signal plot of the acoustic signal in the frequency range of 0~100Hz. It can be seen that there is a lot of interference in the signal spectrum, making it difficult to extract the resonance characteristics.
[0051] The time-frequency diagram is shown in Figure 4. It can be seen that the frequency ranges of the higher-order response, noise interference, and linear frequency modulated signal overlap significantly, making them impossible to suppress directly through conventional filtering. The components are represented as follows:
[0052] ;
[0053] in, For linear frequency modulation excitation and its response; For higher-order linear frequency modulation excitation and its response; Noise originating from the measurement system and the environment.
[0054] Linear frequency modulation excitation and its response The calculation formula is:
[0055] ;
[0056] in, e is the amplitude; e is the natural constant; j is the imaginary unit; The sampling frequency; For sampling sequences; Indicates the initial frequency; Indicates the slope; This is the initial phase; This represents the response to a linear frequency modulation excitation, with a frequency similar to that of the linear frequency modulation excitation.
[0057] Linear frequency modulation higher-order excitation and its response The calculation formula is:
[0058] ;
[0059] in, It represents the order of the higher-order response; This represents the response to a higher-order linear frequency modulated excitation, with a frequency similar to that of the higher-order linear frequency modulated excitation.
[0060] Based on the frequency characteristics of the linear frequency modulated signal, construct the inverse slope linear frequency modulated signal. The calculation formula is:
[0061]
[0062] in, This indicates frequency adjustment; in this embodiment, it is set to... =10Hz.
[0063] target signal With the inverse slope linear frequency modulation signal Multiplication yields an approximate steady-state signal The calculation formula is:
[0064]
[0065] The time-frequency plot of the obtained approximate steady-state signal is shown in Figure 5. It can be seen that the overall signal amplitude has increased, the frequency has been converted to an approximate steady-state frequency, and the frequency components are clearly separable. For the approximate steady-state signal... Using the center frequency f c =20Hz, bandwidth is F b The filtered signal is obtained by filtering with a bandpass filter of 5Hz. The calculation formula is:
[0066]
[0067] in, This indicates in-band noise.
[0068] Constructing an inverse slope linear frequency modulation signal complex conjugate signal The calculation formula is:
[0069] .
[0070] The filtered result is compared with the complex conjugate signal. The recovered linear frequency modulation excitation and its response are obtained by multiplication. The calculation formula is:
[0071] .
[0072] The time-frequency diagram of the recovered linear frequency modulation signal is shown in Figure 6. It can be seen that other interference components have been effectively filtered out, while the linear frequency modulation excitation and its response have been well preserved and recovered.
[0073] To further observe the filtering effect, the time and frequency domain plots of the recovered linear frequency modulated excitation and its response are shown in Figures 7 and 8. It can be seen that the higher-order linear frequency modulated excitation and its response are effectively suppressed by noise interference, and the signal exhibits obvious periodicity, which is beneficial for extracting the signal resonance characteristics.
[0074] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.
Claims
1. A time-varying bandpass filtering method under linear frequency modulation excitation, wherein in the process of measuring the dynamic fluid level in an oil well, a linear frequency modulation signal is sent to excite the air column inside the tubing, the acoustic signal after excitation is collected and processed, and the fluid level depth is measured based on the mathematical relationship between the extracted resonance characteristics and the tubing length, characterized in that... The processing of the sensor signal obtained through continuous resonance of linear frequency modulated sound waves includes the following steps: S1. The target signal is obtained by discretely sampling the sensor signal; S2. Construct a linear frequency modulated signal with inverse slope, i.e.: Where j is the imaginary unit; The sampling sequence; Sampling frequency; Indicates the frequency adjustment; Indicates the slope; Multiplying this by the target signal yields an approximate steady-state signal, i.e.: in, The target signal is obtained through discrete sampling; It is a linear frequency modulated signal with inverse slope; is the amplitude; e is the natural constant; Indicates the initial frequency; Indicates the frequency adjustment; This is the initial phase; It represents the order of the higher-order response; Noise originating from the measurement system and the environment; This represents the response to linear frequency modulation excitation; This represents the response to a higher-order linear frequency modulated excitation; S3. A bandpass filter is used to filter the approximately steady-state signal to obtain the bandpass-filtered steady-state signal. The formula is: ; in, Indicates in-band noise; S4. Multiply the bandpass-filtered steady-state signal with the complex conjugate signal of the inverse-slope linear frequency modulated signal to recover the linear frequency modulated excitation and its response. The formula is: ; in, For inverse slope linear frequency modulation signal The complex conjugate signal is represented as 。 2. The method according to claim 1, characterized in that, Target signal obtained by discrete sampling include: ; in, The linear frequency modulation excitation and its response are expressed as: ; The higher-order linear frequency modulated excitation and its response are expressed as: ; in, Noise originating from the measurement system and the environment; e is the amplitude; e is the natural constant; j is the imaginary unit; The sampling frequency; The sampling sequence; Indicates the initial frequency; Indicates the slope; This is the initial phase; This represents the response to linear frequency modulation excitation; It represents the order of the higher-order response; This represents the response to a higher-order linear frequency modulated excitation.
3. The method according to any one of claims 1 to 2, characterized in that, Sampling frequency Set to 1000Hz.
4. The method according to any one of claims 1 to 2, characterized in that, The slope was set to 3Hz / s.
5. The method according to claim 1, characterized in that, The bandpass filter used for filtering the near steady-state signal has a center frequency of 20Hz and a bandwidth of 5Hz.
