Digital leakage listening detection system based on stepless frequency and multi-band bandwidth adjustment

Abstract

The invention relates to the field of pipeline leakage detection, and discloses a digital leakage listening detection system based on stepless frequency and multi-band bandwidth adjustment, the system comprises a signal acquisition module, a user adjustment module, a main control module, a display module, an audio output module and a power supply module, and the system constructs a real-time simulation auditory sense and asynchronous digital vision dual-channel architecture; the core of the method is that a main control module calculates a leakage signal steady-state index by fusing an energy stability index representing signal energy time stability and a statistical randomness index representing signal statistical randomness in an asynchronous digital visual path; the index provides an objective and quantitative visual navigation basis for operators, and continuous leakage signals and transient and periodic interference can be effectively distinguished. According to the invention, auditory experience and objective data analysis of operators are considered, and the accuracy, reliability and efficiency of leakage detection are remarkably improved.

Description

Technical Field

[0001] This invention relates to the field of pipeline leak detection, specifically a digital leak detection system based on stepless frequency and multi-segment bandwidth adjustment. Background Technology

[0002] Leak detectors are key detection devices used in the operation and maintenance of water supply, gas supply, and heating pipeline networks to locate leaks in underground pressurized pipelines. Their basic working principle involves using a high-sensitivity sensor to pick up the sound waves emitted from the leak point, either from pipeline vibrations or those propagating through the ground. After amplification and filtering, the signals are then used by operators to listen to the sound through headphones or observe the numerical displays on the instrument panel to determine the exact location of the leak.

[0003] However, existing leakage detection devices still have significant shortcomings in terms of signal processing flexibility and objectivity in practical applications. On the one hand, the frequency adjustment function of many devices relies on switching between fixed levels or preset frequency bands. When the characteristic frequency of the leakage signal falls exactly between two levels, this discontinuous adjustment method makes it difficult to accurately align the signal, resulting in signal loss and decreased sensitivity. At the same time, existing devices often lack flexibility in setting the filter bandwidth to adapt to different working stages. While wideband mode is suitable for large-scale surveys, it is prone to introducing too much environmental noise. While narrowband mode is beneficial for suppressing interference, it may reduce the energy of the leakage signal because it cannot fully cover the effective frequency band of the leakage signal, and it also greatly limits the search efficiency.

[0004] Traditional leak detectors have fundamental limitations in their signal evaluation methods. Their judgments heavily rely on the operator's subjective auditory experience or a single signal strength value (such as the RMS value) displayed on the screen. This single evaluation dimension cannot effectively distinguish between the continuous random noise generated by a real leak and common environmental transient impact noise (such as passing vehicles or heavy objects falling) or periodic harmonic noise (such as the steady hum of motors, water pumps, and other machinery). These interference sources can also produce high signal strength readings, easily leading to misjudgments and severely impacting work efficiency. Therefore, current technologies generally lack a quantitative indicator that can objectively assess the inherent physical properties of the signal, thereby providing reliable decision-making assistance to operators, making it difficult to guarantee the accuracy and reliability of the detection results.

[0005] Therefore, this invention proposes a digital listening leak detection system based on stepless frequency and multi-segment bandwidth adjustment to address the shortcomings of existing technologies. Summary of the Invention

[0006] To address the shortcomings of existing technologies, this invention provides a digital listening leak detection system based on stepless frequency and multi-segment bandwidth adjustment. This solves the problem that listening leak detection technology relies too heavily on the operator's subjective auditory experience or makes judgments based solely on a single signal strength value, resulting in low detection accuracy and poor anti-interference capabilities in complex noise environments.

[0007] To achieve the above objectives, the present invention provides the following technical solution: a digital listening leak detection system based on stepless frequency and multi-segment bandwidth adjustment, comprising:

[0008] The signal acquisition module is used to convert physical vibrations into analog electrical signals;

[0009] The user adjustment module is used to receive manual instructions from operators to set filter parameters.

[0010] The main control module is used for execution control and data processing;

[0011] The display module is used to visually present information.

[0012] The audio output module includes a monitoring control circuit for outputting audio signals to the operator.

[0013] The power module provides power to the digital sound leakage detection system;

[0014] The user adjustment module includes a stepless frequency adjustment knob and a multi-segment bandwidth adjustment knob.

[0015] The digital hearing loss detection system is equipped with a real-time analog auditory path and an asynchronous digital visual analysis path. The real-time analog auditory path includes an analog bandpass filter controlled by the user adjustment module hardware. The output of the analog bandpass filter is amplified by power and adjusted by the monitoring control circuit before being output by the audio output module. The input signal of the asynchronous digital visual analysis path is obtained by splitting the analog signal of the real-time analog auditory path. The split signal is converted into a digital audio sequence by an analog-to-digital converter and processed by the main control module.

[0016] The main control module calculates energy stability index and statistical randomness index based on digital audio sequence, and integrates the energy stability index and statistical randomness index into a leakage signal steady-state index;

[0017] The main control module outputs the steady-state index of the leakage signal to the display module for visualization.

[0018] Preferably, the main control module is configured to operate with an intermittent working strategy; the intermittent working strategy includes:

[0019] The system periodically wakes up from a dormant state, acquires a digital audio sequence of a preset length to form a data frame, and returns to a dormant state after completing the calculation of the steady-state index of the leakage signal.

[0020] Preferably, the main control module forms an effective value time series based on the effective values ​​of multiple data frames, and determines the energy stability index based on the statistical dispersion of the effective value time series.

[0021] Preferably, the statistical dispersion of the effective value time series is determined by calculating the coefficient of variation of the effective value time series.

[0022] Preferably, the main control module calculates the zero-crossing rate of each data frame and determines a statistical randomness index based on the deviation between the zero-crossing rate and a preset theoretical value.

[0023] Preferably, the main control module obtains the steady-state index of the leakage signal by multiplying and fusing the energy stability index and the statistical randomness index according to a preset weight, wherein the preset weight is a configurable positive number.

[0024] Preferably, the stepless frequency adjustment knob is used to provide a continuously changing analog voltage signal, and the multi-segment bandwidth adjustment knob is used to output a discrete control voltage by switching the resistance value in the circuit.

[0025] Preferably, the center frequency of the analog bandpass filter is controlled by an analog voltage signal output from a stepless frequency adjustment knob, and the bandwidth of the analog bandpass filter is controlled by a control voltage output from a multi-segment bandwidth adjustment knob.

[0026] Preferably, the main control module responds to the data recording command by binding the current steady-state index of the leakage signal with the location identifier of a measurement point and storing it as a record entry.

[0027] Preferably, the main control module displays the steady-state index of the leakage signal and its corresponding location identifier from multiple stored record entries side by side on the display module, so that the operator can determine the location of the leakage point by comparing the side-by-side display of the steady-state index of the leakage signal.

[0028] This invention provides a digital listening leak detection system based on stepless frequency and multi-segment bandwidth adjustment. It has the following beneficial effects:

[0029] 1. This invention improves the accuracy and anti-interference capability of leak detection by constructing a steady-state index for the leak signal. Unlike existing technologies that rely solely on the effective value of the signal for judgment, this invention integrates the energy stability index, which characterizes the temporal stability of the signal energy, and the statistical randomness index, which characterizes the statistical randomness of the signal, enabling multi-dimensional analysis of the physical characteristics of the signal. This allows the system to effectively distinguish between the continuous random noise generated by a real leak and common environmental interference noise such as transient impacts and periodic mechanical vibrations, thereby significantly reducing the false judgment rate and improving the reliability of detection results in complex industrial environments.

[0030] 2. The parallel architecture of real-time analog hearing and asynchronous digital vision adopted in this invention optimizes the operating experience and improves detection efficiency. The real-time analog hearing path ensures that operators can obtain zero-latency, high-fidelity sound feedback, making full use of their professional hearing experience. Meanwhile, the asynchronous digital vision analysis path independently completes complex signal analysis and provides objective visual navigation basis without interfering with real-time hearing. This "hearing and vision separation" design allows operators to adopt an efficient workflow of "hearing coarse scanning and visual fine adjustment", improving the overall work efficiency from large-scale inspection to precise locking.

[0031] 3. The human-machine collaborative detection method provided by this invention, especially the multi-point data recording and parallel comparison function, objectifies the final location decision-making process and improves the positioning accuracy. Operators can manually trigger data recording at multiple locations around the suspected leak point, and the system will store and centrally display the steady-state index of the leakage signal at each point. This makes the final judgment no longer dependent on the operator's instantaneous memory or subjective feelings, but based on the direct quantitative comparison of data from multiple measurement points, thereby establishing the final confirmation of the leak point on objective data and improving the accuracy of the final location. Attached Figure Description

[0032] Figure 1 This is a schematic diagram of the dual-path signal processing architecture of the present invention;

[0033] Figure 2 This is a schematic diagram of the method flow of the present invention;

[0034] Figure 3 This is a flowchart of the human-machine collaborative closed-loop detection method of the present invention;

[0035] Figure 4 This is a schematic diagram of the system appearance of the present invention.

[0036] The components include: 10. Main body; 20. Multi-segment bandwidth adjustment knob; 30. Stepless frequency adjustment knob; 40. Dedicated button; 100. Signal acquisition module; 200. User adjustment module; 300. Main control module; 400. Display module; 500. Audio output module; and 600. Power supply module. Detailed Implementation

[0037] The technical solutions in 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.

[0038] See attached document Figure 1 , Figure 1 This is a schematic diagram of a dual-path signal processing architecture according to an embodiment of the present invention. The present invention provides a digital listening leak detection system based on stepless frequency and multi-segment bandwidth adjustment. The system may include: a signal acquisition module 100, a user adjustment module 200, a main control module 300, a display module 400, an audio output module 500, and a power supply module 600.

[0039] The signal acquisition module 100 is designed to convert the physical vibration of the pipe or medium under test into an analog electrical signal. In one specific embodiment, the signal acquisition module 100 is a piezoelectric ceramic sensor, which is encapsulated in a probe housing suitable for different working conditions, such as a pin-type probe for indoor testing or a windproof and enclosed probe for outdoor testing. The signal frequency band processed by the core algorithm of this invention is 80Hz to 8000Hz. Meanwhile, some dedicated signal acquisition modules 100, such as pin-type probes with specific mechanical conduction structures, can extend their physical response range to above 8000Hz and are also sensitive to high-frequency vibration components generated by certain high-pressure gas leaks.

[0040] In another preferred embodiment, for a signal acquisition module 100 employing a specific mechanical conduction structure (e.g., a high-rigidity alloy pin), the physical response of the probe can be extended to approximately 10 kHz to enhance sensitivity to the high-frequency components generated by high-pressure gas injection; this high-frequency response can be incorporated into the system as an optional operating mode, and in high-frequency mode, the ADC sampling rate and filter bandwidth configuration can be appropriately increased to ensure signal integrity.

[0041] The user adjustment module 200 is designed to receive manual commands from the operator to set the system's filtering parameters. The user adjustment module 200 further includes a stepless frequency adjustment knob 30 and a multi-segment bandwidth adjustment knob 20. The stepless frequency adjustment knob 30, for example, is a high-precision potentiometer that outputs a continuously changing analog voltage signal proportional to the knob's rotation angle. The multi-segment bandwidth adjustment knob 20, for example, is a multi-position mechanical band switch that, depending on the operator's selected position, changes the circuit parameters by switching different resistance values, thereby outputting a discrete control voltage for setting the filter bandwidth. These positions correspond to preset bandwidth modes, such as a broadband "survey" mode, a mid-band "transition" mode, and a narrowband "fixed-point" mode.

[0042] In another preferred embodiment, for ease of implementation and engineering, the multi-bandwidth configurations in this invention can adopt the following preferred values: FWB (Final Wideband): passband ±500Hz; First narrowband: passband ±200Hz; Second narrowband: passband ±50Hz; Fixed-point (FNB): passband ±25Hz. The above values ​​are preferred examples and can be adjusted according to the detection scenario, probe type and engineering requirements.

[0043] The main control module 300 is the core control and data processing unit of the system. In a specific embodiment, the main control module 300 can be a low-power microcontroller (MCU), such as an ARM-Cortex-M series microcontroller. The main control module 300 is responsible for parsing the parameter instructions from the user adjustment module 200 and executing subsequent digital signal processing algorithms.

[0044] Display module 400, such as liquid crystal display screen 5 (LCD), is electrically connected to main control module 300 and is used to visually present key information of the system.

[0045] In another preferred embodiment, the main interface of the display module 400 preferably simultaneously displays the peak value and effective value (Eff) corresponding to several predetermined frequency points (e.g., 200Hz, 400Hz, 1000Hz, 3000Hz, 8000Hz, etc.) so that the operator can make judgments based on traditional spectrum indicators. In addition, the sub-interface or dedicated area also displays the leakage signal steady-state index (or LSSI) calculated by the present invention. The operator can perform cross-validation and decision-making based on headphone hearing, peak / Eff visualization, and leakage signal steady-state index, thereby taking into account both experience-based and algorithmic interpretation.

[0046] The audio output module 500 includes a monitoring control circuit for adjusting the listening volume, an audio amplification circuit, and a headphone jack 5, which is used to output the processed audio signal to the operator.

[0047] The power module 600, which includes a rechargeable battery pack and power management circuitry, is responsible for providing stable power to all electronic components within the system.

[0048] In another preferred embodiment, the power module 600 preferably includes a lithium battery pack with a nominal voltage of 12.6V and a corresponding power management circuit; preferably, the charger output parameters are 12.6V / 500mA, and the full charging time under normal temperature conditions is approximately 2 hours; when using the low-power intermittent sampling strategy of the present invention and in normal usage mode, the device's continuous working time can reach approximately 52 hours (actual battery life may vary slightly depending on working parameters and usage scenarios). Furthermore, the device can be configured to automatically enter a charging protection mode during charging to ensure safety.

[0049] See attached document Figure 4 The present invention describes the specific physical form of a digital leakage detection system in one embodiment. The system includes a main body 10 for housing internal electronic components; on the front panel of the main body 10, a display module 400 is provided for visually presenting the steady-state index of the leakage signal and other system information.

[0050] In the area below the display module 400, the physical control components of the user adjustment module are centrally located. These components mainly include a stepless frequency adjustment knob 30 and a multi-segment bandwidth adjustment knob 20; by rotating the stepless frequency adjustment knob 30 and the multi-segment bandwidth adjustment knob 20, the operator can achieve continuous adjustment of the center frequency of the analog bandpass filter and discrete switching of the bandwidth, respectively.

[0051] The main body 10 is also equipped with at least one dedicated button 40, which is used to trigger data recording operations. For example, in the process of performing multi-point data recording and comparison, the operator can instruct the main control module to store the steady-state index of the leakage signal at the current measurement point by operating the dedicated button 40.

[0052] See attached document Figure 1This is a schematic diagram of a dual-path signal processing architecture in one embodiment of the present invention. The system constructs two parallel, functionally separated signal processing paths. The signal captured by the signal acquisition module 100 (e.g., a piezoelectric element) is first amplified by a preamplifier and then sent to an analog bandpass filter. The center frequency and bandwidth of this analog bandpass filter are controlled by the stepless frequency adjustment knob 30 and the multi-segment bandwidth adjustment knob 20 in the user adjustment module 200, respectively. The filtered signal is amplified by a power amplifier and then the volume is adjusted by a monitoring control circuit. Its main signal flows to the audio output module 500 (e.g., headphones), forming a real-time analog auditory path. Simultaneously, a signal branch is led out from the power amplifier stage through a signal splitter (i.e., a frequency divider) or buffer. This analog signal branch is sent to an analog-to-digital converter (ADC) for digitization, and then the main control module 300 performs data analysis. The analysis results are finally presented on the display module 400; this is an asynchronous digital visual analysis path. Furthermore, a dedicated button 40 (setting button) is logically connected to the main control module 300 for issuing commands such as data recording.

[0053] The key feature is the construction of two parallel, functionally separated signal processing pathways: a real-time analog auditory pathway and an asynchronous digital visual analysis pathway.

[0054] The real-time simulated auditory pathway constitutes a complete analog signal link. The analog electrical signal output by the signal acquisition module 100 is first conditioned by a preamplifier and then sent to an analog bandpass filter. The center frequency of the analog bandpass filter is controlled by a continuous analog voltage output by the stepless frequency adjustment knob 30, and its bandwidth is directly controlled at the hardware level by a discrete control voltage output by the multi-segment bandwidth adjustment knob 20. The filtered signal is then amplified by a power amplifier and the volume is adjusted by the monitoring control circuit before finally being transmitted to the audio output module 500. This pathway ensures that the sound signal heard by the operator has extremely low latency and high fidelity.

[0055] An asynchronous digital visual analysis path is responsible for performing depth analysis of the signal; a signal branch is shunted from the aforementioned analog auditory path (e.g., after the power amplifier) ​​via a signal splitter (i.e., a frequency divider) or buffer; the shunted analog signal is fed into an analog-to-digital converter (ADC) and converted into a digital audio sequence; the digital audio sequence is then fed into the main control module 300 for performing the calculation of the leakage signal steady-state index proposed in this invention; the operating mode of this path is independent of the real-time audio stream, providing a basis for subsequent low-power, non-real-time calculations.

[0056] See attached document Figure 2 , Figure 2This is a schematic flowchart of a method according to an embodiment of the present invention. The present invention provides a digital method for detecting audio leakage based on stepless frequency and multi-segment bandwidth adjustment. This method is executed by a main control module 300 and includes the following steps:

[0057] S100, the main control module 300 is woken up from the low-power sleep mode at a preset time period and acquires a digital audio sequence of preset length to form a data frame;

[0058] S200 calculates the effective value (RMS) of the data frame and stores it in a sequence for storing the most recent historical effective values. Then, based on the statistical dispersion of the effective value sequence, the energy stability index (ESI) is calculated.

[0059] S300, in parallel, calculates the zero-crossing rate (ZCR) of the same data frame, and calculates the statistical randomness index (SRI) based on the deviation of the zero-crossing rate from the preset theoretical value.

[0060] S400 combines the calculated Energy Stability Index (ESI) and Statistical Randomness Index (SRI) into a single Leakage Signal Steady-State Index (LSSI) using a pre-defined weighted fusion algorithm.

[0061] S500 outputs the Leakage Signal Steady-State Index (LSSI) to the display module 400 for visualization, and after completing the calculation, it causes the main control module 300 to return to a low-power sleep mode until the next wake-up cycle.

[0062] The core algorithms and parameters involved will be explained in detail below.

[0063] See attached document Figure 1 With appendix Figure 2 Inside the main control module 300, a method for calculating the steady-state index of leakage signals was implemented through software programming.

[0064] S201, the main control module 300 executes its intermittent operating mode; by configuring its internal timer, the main control module 300 is woken up from low-power sleep mode at preset time periods (e.g., 100 milliseconds to 200 milliseconds); after each wake-up, the main control module 300 acquires a digital audio sequence of preset length through an analog-to-digital converter (ADC) to form a data frame; this data frame is represented as x. k [n], where k is the index of the current wake-up period, and n is the index of the intra-frame sampling point, ranging from 0 to L. frame -1,L frame The length of the data frame.

[0065] S202, in acquiring data frame x kAfter [n], the main control module 300 calculates the Energy Stability Index (ESI), which includes the following sub-steps:

[0066] S202-1, Calculate the current data frame x k The effective value (RMS) of [n] is denoted as R. rms [k]; its calculation formula is:

[0067]

[0068] S202-2, the main control module 300 maintains a length of N in its internal memory (RAM). win A circular buffer; this buffer is used to store the most recent N... win The effective value calculated from each wake-up cycle; the current R is calculated. rms After [k], it is stored in the circular buffer. If the buffer is full, the oldest historical value is overwritten. At this point, the data in the buffer constitutes a valid value time series.

[0069] S202-3, Based on this effective value time series, calculate its coefficient of variation C. v First, calculate the arithmetic mean μ of the sequence. R and standard deviation σ R :

[0070]

[0071] Among them, R rms,i Let i be the i-th element in the sequence of valid values; subsequently, the coefficient of variation C v The calculation is as follows: The coefficient of variation C v It is a dimensionless parameter that characterizes the relative fluctuation of signal energy within the observation time window.

[0072] S202-4, the calculated coefficient of variation C v Converted to energy stability index I esi This transformation is accomplished through an exponential function, so that the index value is positively correlated with energy stability.

[0073] I esi =exp(-λ esi ·C v );

[0074] Where, λ esi It is a preset normal value used to adjust the sensitivity of the indicator; I esi The range of is (0, 1], and its theoretical value is 1 when the signal energy is absolutely stable.

[0075] S203, calculated in parallel with S202, is based on the same data frame x by the main control module 300. k [n] Calculate the Statistical Randomness Index (SRI), which includes the following sub-steps:

[0076] S203-1, Calculate the current data frame x k The zero-crossing rate Z of [n] zcr [k]; The zero-crossing rate is defined as the number of times the symbol of a signal sample changes within a time window.

[0077]

[0078] Where sgn(·) is the sign function, defined as:

[0079]

[0080] S203-2, the calculated zero-crossing rate Z zcr [k] is converted to the statistical randomness index I. sri The conversion is based on the deviation of the zero-crossing rate from the theoretical value of 0.5 for random white noise.

[0081] I sri =exp(-λ sri ·|Z zcr [k]-0.5|);

[0082] Where, λ sri This is a preset normal value used to adjust the sensitivity of the indicator; the closer the zero-crossing rate of the signal is to 0.5, the better; I sri The closer the value is to its maximum value of 1, the more similar its statistical characteristics are to random noise.

[0083] S204, the main control module 300 will calculate the energy stability index I esi and statistical randomness index I sri The data is fused to generate a single, comprehensive leakage signal steady-state index P. lssi This fusion method uses a weighted product approach, which has the advantage that the final composite index will only be high when both sub-indicators have relatively high values.

[0084]

[0085] Among them, w esi and w sri These are preset weighting coefficients used to define the relative importance of the two sub-indicators in the final evaluation, and w esi >0, w sri >0.

[0086] S205, the main control module 300 will finally calculate the steady-state index P of the leakage signal. lssi The data is sent to the display module 400 for visualization via a communication bus such as the Serial Peripheral Interface (SPI) or the Internal Integrated Circuit (I2C). After completing the calculation and data output task, the main control module 300 returns to its low-power sleep mode and waits for the next timer interrupt to start a new working cycle.

[0087] In parallel with the aforementioned real-time simulated auditory pathway, this system constructs an asynchronous digital visual analysis pathway; the function of this pathway is to perform in-depth digital analysis of signals to extract feature indicators for objective decision-making.

[0088] The signal of the asynchronous digital visual analysis path originates from the real-time analog auditory path; in one specific embodiment, an analog signal branch is drawn from a node in the analog auditory path (e.g., after the power amplification stage and before the audio output module 500) via a signal splitter (i.e., a frequency divider); the signal splitter (i.e., the frequency divider) may employ a buffer amplifier, such as a unity-gain operational amplifier, to achieve high input impedance and low output impedance, thereby avoiding a load effect on the signal quality of the analog auditory path while acquiring the signal.

[0089] The analog signal output from this branch is transmitted to the input of an analog-to-digital converter (ADC); this ADC can convert a continuous analog voltage signal into a discrete digital audio sequence; in a preferred embodiment, the ADC is an integrated peripheral of the main control module 300 (e.g., an ARM-Cortex-M series microcontroller) to simplify circuit design and reduce system cost; the ADC operates at a preset sampling frequency F. s The input signal is sampled and quantized.

[0090] One of its key features is the asynchronous digital vision analysis path, which is asynchronous to the continuous analog audio stream and designed with low power consumption as the goal. This working mode is realized through software programming of the main control module 300, which is specifically manifested as a periodic working cycle. The main control module 300 is in a low-power sleep or shutdown state for most of the time.

[0091] By configuring the timer inside the main control module 300, it can be periodically woken up at a preset, low frequency (e.g., 5 to 10 times per second, i.e., a time period of 100 to 200 milliseconds). Each time it wakes up from sleep mode, the main control module 300 immediately starts the analog-to-digital converter, performs a burst sampling operation, and acquires a short segment of data with a preset length L. frame The digital audio sequence is used to form a data frame for analysis; the duration of the data frame can be set to, for example, 20 milliseconds to 40 milliseconds.

[0092] After successfully acquiring a complete data frame, the main control module 300 immediately executes the subsequent signal processing algorithm, namely, the calculation of the proposed steady-state index of the leakage signal. After the calculation task is completed, the main control module 300 sends the calculation result to the display module 400 and then returns to its low-power sleep state until the next timer interrupt event wakes it up again. This timer-driven, intermittent "wake-up, sampling, calculation, sleep" work cycle constitutes the asynchronous and low-power operation mode of this path. Its refresh rate matches the operator's need for visual information updates, while reducing the overall energy consumption of the system.

[0093] The main control module 300 implements an intermittent working strategy through internal software programming to minimize the overall power consumption of the system while ensuring effective visual information updates; this strategy constructs a periodic working loop driven by a timer.

[0094] In one specific embodiment, the internal hardware timer of the main control module 300 is configured to generate an interrupt request at a preset time period; the time period can be set to 100 milliseconds to 200 milliseconds, which makes the visual information refresh frequency on the display module 400 5 to 10 times per second, which is sufficient for the operator's visual observation.

[0095] During system operation, the main control module 300 is in a low-power sleep state for most of the time, such as stop mode or standby mode. In this state, its core clock is turned off and the power consumption is extremely low.

[0096] When the aforementioned hardware timer reaches its preset period and generates an interrupt, the interrupt event will wake up the main control module 300 from its sleep state and put it into a fully operational active state. After entering the active state, the main control module 300 immediately starts its internally integrated analog-to-digital converter (ADC).

[0097] The analog-to-digital converter performs a short burst sampling operation, that is, within a short time window (e.g., 20 milliseconds to 40 milliseconds), sampling at a preset sampling frequency F. s Analog signals from an asynchronous digital vision analysis path are continuously acquired to obtain a length of L. frame The digital audio sequence forms a data frame for subsequent analysis.

[0098] After the data frame acquisition is completed, the main control module 300, while remaining active, immediately executes all the calculation tasks of the Leakage Signal Steady-State Index (LSSI) proposed in this invention. This calculation task includes the calculation of the Energy Stability Index (ESI), the calculation of the Statistical Randomness Index (SRI), and the fusion of the two into the final comprehensive index.

[0099] After the calculation task is completed and the final steady-state index of the leakage signal is obtained, the main control module 300 sends the index value to the display module 400 through the corresponding communication interface; after the data transmission task is completed, the main control module 300 returns to the preset low-power sleep state by executing specific instructions.

[0100] The main control module 300 will remain in this sleep state until the next timer interrupt event arrives. At that time, the complete working cycle of "wake-up, sampling, calculation, output, and sleep" will be triggered again and repeated. In this way, the active time of the main control module 300 is strictly limited to a very short time segment within each working cycle, thereby achieving low-power operation.

[0101] Within each activity cycle of the main control module 300, the Energy Stability Index (ESI) is calculated to quantify the degree of energy fluctuation of the acquired signal over a period of time; this calculation process specifically includes the following steps:

[0102] First, the main control module 300 calculates the root mean square (RMS) value, i.e., the effective value R, based on the data frames collected within the current activity cycle. rms [k], where k is the index of the current activity cycle, and this value reflects the signal energy of the current data frame;

[0103] The main control module 300 maintains a length of N in its internal memory. win A circular buffer; this buffer, as a first-in-first-out (FIFO) data structure, is used to store the most recently accessed N... win The effective value calculated from each activity cycle; the currently calculated effective value R rms [k] is written to the circular buffer, overwriting the earliest historical value stored therein; in this way, the buffer always dynamically stores an effective time series of values ​​representing the recent trend of signal energy changes.

[0104] The main control module 300 is based on all N stored in the circular buffer. win Given one effective value, calculate the coefficient of variation C of the time series of that effective value. v This calculation process includes: first calculating the arithmetic mean μ of the sequence. R With standard deviation σ R Then, the obtained standard deviation σ RDivide by the arithmetic mean μ R The coefficient of variation C was obtained. v The coefficient of variation is a dimensionless parameter whose value is not affected by the overall signal strength, but rather purely characterizes the relative fluctuation of signal energy within the observation time window.

[0105] The calculated coefficient of variation C v Used to generate the final energy stability index l esi This generation process is accomplished through a preset exponential function; this function will convert the input coefficient of variation C... v The value is non-linearly mapped to a value in the interval (0, 1); this mapping relationship ensures that the more stable the signal energy (i.e., C), the more stable the index value becomes. v When the value is smaller, the output energy stability index l esi The closer the value is to its maximum value of 1, the better. esi The value is the energy stability index calculated for this activity cycle and used for subsequent fusion.

[0106] In parallel with the calculation of energy stability indicators, the main control module 300 calculates based on the same data frame x k [n] performs the calculation of the Statistical Randomness Index (SRI), which is designed to assess how similar a signal is to random noise in terms of statistical properties.

[0107] The calculation process first determines the zero-crossing rate Z of the current data frame. zcr [k]; The zero-crossing rate is defined as the frequency at which a signal sample crosses zero within a time window; In the specific implementation, the main control module 300 traverses all adjacent sample point pairs within the data frame, and accumulates the number of signal symbol changes by comparing the difference in the sign function of each pair of adjacent sample points; This accumulated number, after normalization, yields the zero-crossing rate Z of the current data frame. zcr [k].

[0108] Calculated zero-crossing rate Z zcr [k] is compared with a preset theoretical value; this theoretical value is determined based on an ideal random white noise model with a Gaussian distribution, and its theoretical zero-crossing rate is 0.5; the main control module 300 calculates the current zero-crossing rate Z. zcr [k] is the absolute deviation from the theoretical value of 0.5; this deviation quantifies the degree to which the statistical characteristics of the acquired signal deviate from the ideal random noise model; a smaller deviation means that the signal is more random.

[0109] This absolute deviation value is used as input to a preset exponential function to generate the final statistical randomness index I. sriThe exponential function performs a nonlinear mapping, converting the deviation value into an index value within the interval (0, 1). This mapping relationship is set such that the smaller the input deviation value (i.e., the stronger the signal randomness), the higher the output statistical randomness index I. sri The closer the value is to its theoretical maximum value of 1, the better. sri The value is a statistical randomness indicator calculated for this activity cycle and used for subsequent integration.

[0110] The main control module 300 completes the energy stability index. esi and statistical randomness index l sri After calculation, the main control module 300 performs a fusion operation, combining the two independent sub-indices into a single, comprehensive leakage signal steady-state index P. lssi .

[0111] The fusion operation employs a weighted product approach. This approach is chosen based on the following technical considerations: the target signal must simultaneously satisfy both energy stability and statistical randomness; if the value of any sub-index is low, the product operation will lead to a corresponding decrease in the final composite index; in the fusion calculation, the two sub-indexes are each assigned a preset weight coefficient w. esi and w sri This is used to adjust the relative importance of energy stability and statistical randomness in the final evaluation.

[0112] The calculated steady-state exponent P of the final leakage signal lssi The main control module 300 sends the data to the display module 400 through its communication interface; in a specific embodiment, the communication interface may be a serial peripheral interface (SPI) or an internal integrated circuit (I2C) bus.

[0113] After receiving the index value, the display module 400 presents it to the operator in one or more preset visualization forms. These visualization forms may include: displaying it directly on the screen as a numerical value; presenting it as a bar chart or progress bar whose length or height changes dynamically with the index value; or mapping it to a custom graphic symbol, such as a circular or bar indicator whose fill degree changes with the index value.

[0114] See attached document Figure 3 , attached Figure 3 The flowchart illustrates the human-machine collaborative closed-loop detection method. This invention combines the operator's experience with the system's objective quantitative analysis capabilities. The specific steps are as follows:

[0115] S301, the operator starts the equipment and performs a preliminary auditory guided scan through a real-time simulated auditory pathway. In this step, the operator listens to the real-time sound captured by the signal acquisition module 100 and processed by the analog circuit through the audio output module 500. The operator can use the user adjustment module 200 to set the bandwidth to a wider level (e.g., "survey" mode) and move the signal acquisition module 100 to perform a wide-range scan on the possible area of ​​the pipeline under test, while rotating the stepless frequency adjustment knob 30 to preliminarily determine the suspicious areas where abnormal sounds exist based on auditory experience.

[0116] S302, after identifying the suspicious area, the operator switches to fine-tuning under visual navigation; in this step, the operator's main attention shifts from auditory perception to the steady-state index P of the leakage signal presented on the display module 400. lssi The operator visualizes the readings; by making combined adjustments to the user adjustment module 200, the operator switches between several discrete, narrower bandwidth levels (e.g., "transition" or "fixed-point" modes) and makes continuous, fine adjustments to the stepless frequency adjustment knob 30 at each level; during the adjustment process, the operator continuously observes the steady-state index P of the leakage signal. lssi The changes are used to find parameter combinations that can significantly increase the value of the index.

[0117] S303, the operator performs closed-loop optimization to finally confirm the location of the leak point; this step is a continuation and final confirmation of S302; the operator's goal is to find a unique physical location and a unique set of filtering parameters (i.e., the optimal bandwidth and the optimal center frequency), under which the steady-state index P of the leak signal presented on display module 400 is determined. lssi It can reach its maximum value and the reading remains stable.

[0118] S304, To further refine the pinpointing and eliminate the randomness of environmental interference, the operator can perform multi-point data recording and comparison; in this step, the operator selects several (e.g., four) measurement points with clear spatial relationships near the maximum value point determined in S303; at each measurement point, the operator stably places the signal acquisition module 100 and waits for the steady-state index P of the leakage signal on the display module 400. lssi After the reading stabilizes, a data recording operation is triggered via a dedicated button (e.g., the SET button); upon receiving the recording command, the main control module 300 will record the current P... lssi The value is bound to the location identifier of the measurement point (e.g., A, B, C, D) and stored in internal memory.

[0119] S305, after completing the data recording of all preset measurement points, the main control module 300 will store multiple P...lssi The values ​​and their corresponding location identifiers are simultaneously displayed side-by-side on a dedicated interface (e.g., a sub-interface) of the display module 400; by intuitively comparing the magnitudes of these values, the operator can objectively determine the center of the leakage signal energy; the steady-state index P of the leakage signal... lssi The measurement point with the highest value is the location that is ultimately confirmed as the leak point.

[0120] In another preferred embodiment, during the precise positioning stage, a phased point deployment strategy can be adopted: first, coarse positioning is performed at intervals of 0.5m-1.0m; then, fine positioning is performed at intervals of 0.2m-0.3m; finally, precise positioning is performed around the suspected point at intervals of 0.1m, and data is recorded; the data recording operation is triggered by the SET key: after the measuring point stabilizes, the SET key is pressed to record the current steady-state index P of the leakage signal. lssi The values ​​and location information (e.g., A, B, C, D) are stored in the internal memory for subsequent parallel comparison and determination.

[0121] 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 digital listen-fool detection system based on stepless frequency and multi-section bandwidth adjustment, characterized in that, The digital leak detection system comprises: a signal acquisition module for converting physical vibration into analog electrical signals; a user adjustment module for receiving manual instructions from an operator to set filter parameters; a main control module for performing control and data processing; a display module for visual presentation of information; an audio output module including a monitoring control circuit for outputting audio signals to the operator; a power module for providing power for the digital leak detection system; wherein the user adjustment module comprises a stepless frequency adjustment knob and a multi-section bandwidth adjustment knob; the digital leak detection system is provided with a real-time analog auditory pathway and an asynchronous digital visual analysis pathway; the real-time analog auditory pathway comprises an analog band-pass filter controlled by the user adjustment module hardware, the output of the analog band-pass filter is amplified by a power amplifier and adjusted by a monitoring control circuit, and then output by the audio output module; the input signal of the asynchronous digital visual analysis pathway is obtained by splitting the analog signal of the real-time analog auditory pathway, the split signal is converted into a digital audio sequence by an analog-to-digital converter, and processed by the main control module; the main control module calculates an energy stability index and a statistical randomness index based on the digital audio sequence, and fuses the energy stability index and the statistical randomness index into a leak signal steady-state index; the main control module outputs the leak signal steady-state index to the display module for visual presentation.

2. The digital listen-fool detection system based on stepless frequency and multi-section bandwidth adjustment according to claim 1, characterized in that, The main control module is configured to operate in an intermittent working strategy; the intermittent working strategy comprises: periodically waking up from a dormant state, collecting a digital audio sequence of a preset length to form a data frame, and returning to the dormant state after completing the calculation of the leak signal steady-state index.

3. The digital listen-fool detection system based on stepless frequency and multi-section bandwidth adjustment according to claim 1, characterized in that, The main control module forms an effective value time sequence based on the effective values of multiple data frames, and determines the energy stability index based on the statistical dispersion degree of the effective value time sequence.

4. The digital listen-fool detection system based on stepless frequency and multi-section bandwidth adjustment according to claim 3, characterized in that, The statistical dispersion degree of the effective value time sequence is determined by calculating the coefficient of variation of the effective value time sequence.

5. The digital listen-fool detection system based on stepless frequency and multi-section bandwidth adjustment according to claim 1, characterized in that, The main control module calculates the zero-crossing rate of each data frame, and determines the statistical randomness index based on the deviation of the zero-crossing rate from a preset theoretical value.

6. The digital listen-fool detection system based on stepless frequency and multi-section bandwidth adjustment according to claim 1, characterized in that, The main control module fuses the energy stability index and the statistical randomness index by multiplication according to a preset weight to obtain the leak signal steady-state index, and the preset weight is a configurable positive number.

7. The digital listen-fool detection system based on stepless frequency and multi-section bandwidth adjustment according to claim 1, characterized in that, The stepless frequency adjustment knob is used to provide a continuously changing analog voltage signal, and the multi-section bandwidth adjustment knob is used to output a discrete control voltage by switching the resistance value in the circuit.

8. The digital listen-fool detection system based on stepless frequency and multi-section bandwidth adjustment according to claim 7, characterized in that, The center frequency of the analog band-pass filter is controlled by the analog voltage signal output by the stepless frequency adjustment knob, and the bandwidth of the analog band-pass filter is controlled by the control voltage output by the multi-section bandwidth adjustment knob.

9. The digital listen-fool detection system based on stepless frequency and multi-section bandwidth adjustment according to claim 1, characterized in that, The main control module binds the current leak signal steady-state index with the position identification of a measurement point and stores it as a record entry in response to a data recording instruction.

10. The digital listen-fool detection system based on stepless frequency and multi-section bandwidth adjustment according to claim 9, characterized in that, The main control module displays the leak signal steady-state indexes and the corresponding position identifications in the stored multiple record entries side by side on the display module, so that the operator can determine the leak point position by comparing the side-by-side displayed leak signal steady-state indexes.

Images