Automatic counting method and device for metal pipe conveying line

Abstract

The invention relates to an automatic counting method and device for a metal pipe conveying line, and relates to the technical field of metal pipe counting, and the automatic counting method comprises the steps that a signal generator generates an analog waveform signal according to a preset frequency control word, and the analog waveform signal passes through a signal amplifier to obtain an excitation signal; the magnetic field detector generates a magnetic field according to the excitation signal, covers a quality inspection area of the metal pipe conveying line and constructs a counting detection area; when the metal tubes pass through a counting detection area, the magnetic field detector generates a signal amplitude and a signal oscillogram according to a gap between adjacent metal tubes or a gap between metal tube bodies; the controller judges the signal amplitude and the metal tube gap distance corresponding to the signal oscillogram according to a preset signal threshold range to obtain an adhesion overlapping signal; adopting a counting separation model to segment adhesion overlapping signals, determining the number of the metal tubes, generating electric signals of the corresponding number, activating a light emitter, and transmitting the electric signals to a counter to complete counting of the number of the metal tubes; and the counting precision of the metal tubes is improved.

Description

Technical Field

[0001] This invention relates to the field of metal tube counting technology, and in particular to an automated counting method and apparatus for metal tube conveying lines. Background Technology

[0002] The manufacturing industry is shifting towards a multi-variety, small-batch, lean production model, requiring full inspection and online monitoring to ensure quality. Automated counting technology has become a key tool for meeting the needs of rapid delivery and precise control.

[0003] Traditional manual counting is inefficient and error-prone, making it unsuitable for high-speed production lines. Automated counting uses machine vision and 3D inspection technologies (such as industrial cameras and image processing) to quickly and accurately identify and count the number of metal tubes. Currently, metal tube conveyor lines mainly rely on photoelectric sensors, which are used when metal tubes are closely packed on the conveyor line, with only tiny gaps between adjacent tubes.

[0004] Application number CN202211677460.9 discloses a rotary counting device and method based on electromagnetic signals, including a tray and a rotary detection device. The upper surface of the tray is partially metallized. The rotary detection device includes a coil, a pulse excitation circuit, and a signal processing circuit. The coil includes a transmitting coil and at least two receiving coils. The signal processing circuit includes a voltage amplification and detection circuit and a signal processor. The pulse excitation circuit includes a resistor R4, a capacitor C4, an inverter P1, and an inverter P2. The voltage output terminal of the signal processor is connected in series with the capacitor C4, the inverter P1, and the inverter P2 in sequence. One end of the resistor R4 is grounded, and the other end is connected between the capacitor C4 and the inverter P1, so that the capacitor C4 and the resistor R4 form a differentiating circuit. The above invention uses a signal processing algorithm to replace the comparator and improves the counting accuracy and anti-interference ability of the counting device by internal counting and then outputting through a reliable interface.

[0005] Application number CN200510029248.1 discloses a device and method for detecting the pipe gap of steel pipes using ultrasonic flaw detection. The aim is to accurately detect the pipe gap of steel pipes connected end-to-end and to trigger an alarm for large defects such as holes. The device includes a first detection coil and a second detection coil, a first signal processing unit and a second signal processing unit, a logic judgment unit, a signal output unit, and an alarm unit. The method energizes the first and second detection coils, enabling electromagnetic coupling with the steel pipe to generate magnetic flux density change signals, which are then sent to the first and second signal processing units respectively. These processing units filter and discard the signals, sending the processed signals to the logic judgment unit. The logic judgment unit then identifies the pipe gap signal and the large defect signal, sending them to the signal output unit and the alarm unit respectively to output signals. This effectively controls the tracking of material flow and avoids serious steel pipe product quality accidents.

[0006] The existing technical solutions mentioned above have the following drawbacks: 1. Photoelectric sensors cannot accurately count by detecting the light signal reflected from the end face of the metal tube, and are prone to missed detection during the metal tube transportation process; 2. Cracks or gaps may be caused during the production of the metal tube, which increases the probability of false judgment by the photoelectric sensor. Summary of the Invention

[0007] To address the shortcomings of existing technologies, the present invention aims to provide an automated counting method and apparatus for metal pipe conveying lines. The automated counting device detects the gaps between metal pipes by sensing changes in the magnetic field on the metal pipe conveying line, and counts the metal pipes based on these gaps. Detection can be completed even when metal pipes are closely packed together. Counting is performed based on the amplitude limit range of the magnetic field change signal, which can avoid misjudging cracks or gaps on the metal pipes and affecting the counting accuracy.

[0008] The above-mentioned objective of this invention is achieved through the following technical solutions: An automated counting method for metal pipe conveying lines is applied to an automated counting device including a signal generator, a signal amplifier, a magnetic field detector, a controller, a counter, and a light emitter; comprising: The signal generator generates an analog waveform signal according to the preset frequency control word, and after passing through the signal amplifier, an excitation signal is obtained. The magnetic field detector generates a magnetic field according to the excitation signal, and determines the quality inspection area of ​​the metal pipe conveyor line covered by the magnetic field as the counting detection area; When passing through a metal tube within the counting detection area, the magnetic field detector generates a signal amplitude and a signal waveform diagram based on the gap between adjacent metal tubes or the gap in the metal tube body. Based on the preset signal threshold range controller, the signal amplitude and the metal tube gap distance corresponding to the signal waveform are determined to obtain the adhesion and overlap signal; The overlapping signal is segmented using a counting separation model to determine the number of metal tubes, generate a corresponding number of electrical signals, and transmit them to the counter to count the number of metal tubes.

[0009] By employing the above technical solution, a sinusoidal signal of a specific frequency is generated by a signal generator (D / A converter). After power and voltage amplification, it provides a stable excitation signal to the magnetic field detector (probe). Under the action of the excitation signal, the probe generates an alternating magnetic field that covers the metal tube conveying area. When the metal tube passes through the counting detection area, the probe detects gaps on the metal tube or gaps between adjacent metal tubes by sensing changes in the magnetic field. Based on the different amplitudes of the changes in the gap size, different LEDs at different positions light up. When a gap between adjacent metal tubes is detected, the leftmost LED is triggered to light up, and an electrical signal is simultaneously output to the PLC control system for counting. This reduces the false count rate caused by cracks or gaps on the metal tubes and improves the counting accuracy of the metal tubes.

[0010] The present invention is further configured such that: the specific steps of generating an analog waveform signal by the signal generator according to the preset frequency control word, and obtaining an excitation signal after passing through the signal amplifier, include: According to the preset frequency control word, the signal generator, in conjunction with the clock pulse, accumulates the step value one by one to obtain the phase value; The clock pulse is a time reference signal used to control the timing and stepping rhythm of waveform generation; the stepping value is the increment within each clock cycle. According to the preset waveform amplitude value, the signal generator converts the phase value into the corresponding signal amplitude value; Based on the signal amplitude value, the signal generator outputs corresponding waveform data and converts it into a corresponding analog waveform signal; The analog waveform signal is subjected to voltage gain operation and signal conditioning operation by a voltage amplification factor signal amplifier to obtain a boosted analog signal, which is then compared with a preset voltage threshold; the signal conditioning operation includes impedance matching and filtering. If the boosted analog signal is less than the voltage threshold, the voltage amplification factor of the signal amplifier is corrected using the first correction factor. If the boosted analog signal is greater than or equal to the voltage threshold, the signal amplifier performs a current gain operation on the boosted analog signal according to the current amplification factor to obtain a current boosted analog signal, calculates the gain power value with the boosted analog signal, and then compares it with the preset power threshold for judgment. If the gain power value is less than the power threshold, the current amplification factor of the signal amplifier is corrected using a second correction factor. If the gain power value is greater than or equal to the power threshold, the current current boosting analog signal is filtered by a filter circuit to remove noise, and the signal amplifier outputs an excitation signal; the excitation signal includes square wave, sawtooth wave, triangular wave, pulse wave, simple harmonic wave and step wave.

[0011] By adopting the above technical solution, the signal generator converts the phase value into an analog waveform signal according to the preset frequency and amplitude parameters. The signal amplifier first performs voltage gain and impedance matching, filtering and other conditioning on the signal. If the output voltage does not reach the threshold, the voltage amplification factor is adjusted by the correction factor. When the voltage reaches the standard, the current gain is further adjusted and the power value is calculated. If the power is insufficient, the current amplification factor is adjusted. Finally, a stable excitation signal after filtering is output. By monitoring the voltage and power threshold in real time and dynamically adjusting the amplification parameters (voltage / current gain) in combination with the correction factor, the stability and adaptability of the output signal are ensured.

[0012] The present invention is further configured such that: the specific steps of generating a magnetic field by the magnetic field detector according to the excitation signal and determining the quality inspection area of ​​the metal pipe conveyor line covered by the magnetic field as a counting detection area include: The excitation signal output by the signal amplifier is received by the driving circuit of the magnetic field detector and transmitted to the transmitting coil of the magnetic field detector. The transmitting coil generates an alternating current according to the excitation signal, and generates a magnetic field in the space surrounding the transmitting coil according to Ampere's circuital law. The magnetic field includes alternating magnetic field, pulsating magnetic field, and pulsed magnetic field; The surrounding space is a magnetic field distribution, including a near-field region, a transition region, and a far-field region; The magnetic field propagates outward in the form of electromagnetic waves, covering the quality inspection area of ​​the metal pipe conveyor line and forming a counting and detection area.

[0013] By adopting the above technical solution, the excitation signal is transmitted to the transmitting coil through the driving circuit, which excites multiple types of magnetic fields including alternating / pulsating / pulsating magnetic fields. Based on Ampere's circuital law, an electromagnetic field distribution covering the metal pipe inspection area is formed, realizing non-contact accurate detection. Through the magnetic field gradient distribution from the near field to the far field, the metal pipe inspection area is accurately covered, improving the defect detection rate.

[0014] The present invention is further configured such that, when the magnetic field detector passes through a metal tube within the counting detection area, the specific steps for generating a signal amplitude and a signal waveform based on the gap between adjacent metal tubes or the gap in the metal tube body include: When the metal tube passes through the counting and detection area, eddy currents are generated on the surface of the metal tube according to the principle of electromagnetic induction, and the rate of change of magnetic flux is calculated. The magnetic flux change rate is compared with the preset adjacent gap magnetic flux threshold and the preset self-gap magnetic flux threshold to determine the gap between adjacent metal tubes or the gap in the metal tube body. If the rate of change of magnetic flux is greater than or equal to the magnetic flux threshold of the adjacent gap, and less than the magnetic flux threshold of its own gap, then it is determined to be an adjacent metal tube gap; If the rate of change of magnetic flux is greater than or equal to the magnetic flux threshold of the gap itself, it is determined to be a gap in the metal tube body; If the rate of change of magnetic flux is less than the magnetic flux threshold of the adjacent gap, it is determined to be an invalid signal or background noise; The receiving coil senses the rate of change of magnetic flux and generates a corresponding signal amplitude and signal waveform.

[0015] By adopting the above technical solution, the magnetic flux change rate caused by the metal tube is detected by the receiving coil, and the gaps between adjacent metal tubes and the gaps in the metal tube body are distinguished based on the preset threshold. Accurate classification is achieved by combining the signal amplitude and waveform. By dynamically comparing the magnetic flux thresholds of the adjacent gaps and the gaps in the metal tube body, the gaps between the metal tubes and the defects in the body are effectively distinguished, and the detection accuracy is improved. The adjustable threshold settings and multi-parameter analysis (magnetic flux, signal amplitude, waveform morphology) enhance the system's adaptability to complex scenarios.

[0016] The present invention is further configured such that: the specific steps of determining the adhesion and overlap signal by analyzing the signal amplitude and the metal tube gap distance corresponding to the signal waveform diagram based on a preset signal threshold range controller include: The signal threshold range includes minimum amplitude, maximum amplitude, and duration thresholds; The signal amplitude and the signal waveform are filtered and denoised to obtain the effective amplitude and the effective waveform. Waveform features are extracted from the valid waveform to obtain the peak duration, and then compared with a preset signal threshold range. If the effective amplitude is greater than or equal to the minimum amplitude and less than or equal to the maximum amplitude, and the peak duration is greater than the duration threshold, then the current signal amplitude is determined to be the amplitude of the stuck metal tube signal. If the effective amplitude is greater than or equal to the minimum amplitude and less than or equal to the maximum amplitude, but the peak duration is equal to the duration threshold, then the current signal amplitude is determined to be the amplitude of a single metal tube signal. If the effective amplitude is greater than or equal to the minimum amplitude and less than or equal to the maximum amplitude, but the peak duration is less than the duration threshold, then the current signal amplitude is determined to be an invalid signal amplitude. If the effective amplitude is less than the minimum amplitude or greater than the maximum amplitude, then the current signal amplitude is determined to be an invalid signal amplitude. The individual metal tube signal amplitude is transmitted to the controller to generate a corresponding electrical signal, and simultaneously transmitted to the counter to count the number of metal tubes. Clustering the signal amplitude and corresponding effective waveform of the adhered metal tubes yields the adhered and overlapping signals.

[0017] By adopting the above technical solution, and by setting the signal threshold range and waveform feature analysis, combined with the controller's filtering and classification logic, it can accurately distinguish between single metal tubes, adhered metal tubes, and invalid signals. It can also achieve dynamic counting and clustering of adhered signals by activating the light emitter and counter. Based on the dual threshold judgment of amplitude limit range and peak duration, it can effectively distinguish between single / adhesive metal tube signals and reduce misjudgments. Through multi-level threshold settings and signal validity judgment, it can significantly reduce the interference of background noise and abnormal signals.

[0018] The present invention is further configured such that: the specific steps of segmenting the overlapping signals using a counting separation model, determining the number of metal tubes, generating a corresponding number of electrical signals, and simultaneously transmitting them to the counter to count the number of metal tubes include: The counting separation model includes a threshold segmentation layer, a peak localization layer, a time segmentation layer, a morphology verification layer, and a dynamic correction layer; the threshold segmentation layer uses a moving average filtering algorithm to analyze the overlapping signals. Denoising was performed to obtain a noise-free, sticky signal. and with preset signal threshold Comparison and judgment:

[0019] in, For the sample time, For the size of the filter window, This is the sample location sequence number. For the overall sample index; like Then the current noiseless adhesion signal is marked as a valid detection signal; if If so, the current noiseless adhesion signal is marked as an invalid background signal; The peak location layer uses a maximum detection algorithm to detect the effective detection signal, obtain several signal peaks, and record the peak time; like and If it is, then it is marked as the signal peak value; if If so, it is determined to be a non-signal peak. in, For time window, Peak threshold As a multiple of the base number; The time division layer specifies the time interval between adjacent signal peaks. and overlapping area Perform calculations and compare with a preset time interval threshold. and preset integration area threshold Comparison and judgment; ,

[0020] in, The peak value number of the signal; like and If the signal peak value is positive, then the number of metal tubes corresponding to the current signal peak value is determined to be two; otherwise, it is determined to be a single metal tube. The morphological verification layer utilizes structural elements. For noiseless adhesion signals Perform morphological operations to obtain morphological regions. ;

[0021] in, For corrosion operation, Including rectangular and Gaussian functions, For expansion operation; The number of metal tubes was obtained by statistically analyzing the independent peak regions in the morphological region and comparing it with the number of metal tubes obtained from the time segmentation layer. If the two are consistent, the controller generates a corresponding number of electrical signals to activate the light emitter and transmit them to the counter. If the two are inconsistent, the preset signal threshold is determined by the counting separation model based on the real-time signal amplitude characteristics. Make corrections;

[0022] in, The average value of the current signal amplitude. The standard deviation of the current signal amplitude. This is a correction factor; For adjacent signal peaks The amplitude difference is calculated and compared with a preset amplitude difference threshold. The comparison was then made to verify the number of metal tubes. like If the signal peak value is positive, the number of metal tubes corresponding to the current signal peak value is determined to be two; otherwise, it is determined to be a single metal tube.

[0023] By adopting the above technical solution, the signal of the adhered metal tube is separated and counted through a multi-layer collaborative mechanism of the counting separation model (threshold segmentation, peak location, time segmentation, morphological verification, and dynamic correction): First, the signal is filtered and denoised to determine the valid signal. Then, multi-dimensional analysis is performed based on time interval, integral area, morphological operation, and amplitude difference, and dynamic threshold correction is combined to improve accuracy. Finally, the controller activates the light emitter and updates the count. By integrating statistical analysis (time / amplitude threshold) and morphological operation, the adhered signal is accurately distinguished from the independent metal tube, significantly reducing the false judgment rate. Morphological processing (corrosion / expansion) enhances signal separation, and a dual verification mechanism (time segmentation and morphological verification) ensures the reliability of the results.

[0024] Secondly, the present invention also provides an automated counting device for metal pipe conveying lines, employing the following technical solution: An automated counting device for a metal pipe conveying line, used to implement the aforementioned automated counting method, includes: A signal generator is used to generate analog waveform signals according to a preset frequency control word. A signal amplifier is used to amplify the voltage and power of the analog waveform signal to obtain an excitation signal; A magnetic field detector is used to generate a magnetic field based on the excitation signal and cover the quality inspection area of ​​the metal pipe conveyor line to construct a counting detection area. At the same time, it generates signal amplitude and signal waveform when the metal pipe passes through the counting detection area. The controller is used to determine the gap distance between the metal tubes corresponding to the signal amplitude and the signal waveform diagram based on a preset signal threshold range, obtain the adhesion and overlap signal, segment it, determine the number of metal tubes, and generate a corresponding number of electrical signals. A counter is used to record the number of electrical signals and to count the number of metal tubes. The light emitter includes several LED beads; The light emitter is used to activate the corresponding LED beads based on the electrical signal and a preset signal threshold range; Peripheral interface for connecting to third-party devices; the third-party devices include PLCs; A circuit board for integrating the signal generator, the signal amplifier, the magnetic field detector, the controller, the counter, the light emitter, and the peripheral interface; The signal generator is connected to the signal amplifier via wires on the circuit board. The signal amplifier is connected to the magnetic field detector via wires. The magnetic field detector is connected to the controller via wires. The controller is connected to the counter, the light emitter, and the peripheral interface via wires.

[0025] By adopting the above technical solution, a signal generator, signal amplifier, magnetic field detector, controller, counter, light emitter, and peripheral interface are integrated on the circuit board. The signal generator generates an analog waveform signal, the amplifier amplifies the voltage / power and outputs an excitation signal, the magnetic field detector excites a magnetic field to cover the quality inspection area of ​​the metal tube conveying line and generates a signal amplitude / waveform diagram, the controller determines the number of metal tubes based on a preset threshold to divide the adhesion signal and generates an electrical signal, the counter counts the number, the light emitter activates the LED beads to indicate the status, and the peripheral interface connects to PLC and other devices, realizing non-contact accurate counting of metal tubes in the metal tube conveying line.

[0026] Threshold segmentation and dynamic correction reduce false positives and improve accurate counting; integrated components work together to enhance anti-interference capabilities and improve robustness; non-contact detection (electromagnetic induction principle ensures safety and reliability).

[0027] In summary, the beneficial technical effects of the present invention are as follows: A multi-layered collaborative mechanism (threshold segmentation, peak location, time segmentation, morphological verification, and dynamic correction) of a non-contact detection device combined with a counting separation model is used to separate and count the signals of adhered metal tubes. When a gap between adjacent metal tubes is detected, the leftmost LED is triggered to light up, and an electrical signal is simultaneously output to the PLC control system for counting. This reduces the false judgment rate caused by cracks or gaps on the metal tubes and improves the counting accuracy of the metal tubes.

[0028] Based on the dual threshold judgment of amplitude limit range and peak duration, it effectively distinguishes between signals from individual / adhesive metal tubes and reduces misjudgment.

[0029] By setting multiple threshold levels and determining signal validity, interference from background noise and abnormal signals is significantly reduced. Attached Figure Description

[0030] Figure 1 This is a flowchart illustrating an automated counting method according to one embodiment of the present invention.

[0031] Figure 2 This is a flowchart illustrating an automated counting method according to one embodiment of the present invention.

[0032] Figure 3 This is a flowchart illustrating an automated counting method according to one embodiment of the present invention.

[0033] Figure 4 This is a schematic diagram of the structure of an automated counting device according to one embodiment of the present invention. Detailed Implementation

[0034] The present invention will be further described in detail below with reference to the accompanying drawings. Example

[0035] Reference Figure 1 This invention discloses an automated counting method for metal pipe conveying lines, applied to an automated counting device including a signal generator, signal amplifier, magnetic field detector, controller, counter, and light emitter; the automated counting method includes: S1: The signal generator generates an analog waveform signal according to the preset frequency control word, and after passing through the signal amplifier, an excitation signal is obtained; S2: The magnetic field detector generates a magnetic field according to the excitation signal, and determines the quality inspection area of ​​the metal pipe conveyor line covered by the magnetic field as the counting detection area; S3: When passing through the metal tube in the counting detection area, the magnetic field detector generates a signal amplitude and a signal waveform diagram based on the gap between adjacent metal tubes or the gap in the metal tube body; S4: Based on the preset signal threshold range controller, the signal amplitude and the metal tube gap distance corresponding to the signal waveform are determined to obtain the adhesion and overlap signal; S5: The overlapping signal is segmented using a counting separation model to determine the number of metal tubes, and a corresponding number of electrical signals are generated and transmitted to the counter to count the number of metal tubes.

[0036] The implementation principle of this embodiment is as follows: The signal generator generates a highly stable analog waveform signal at the μs level based on a preset frequency control word (such as DDS technology). The signal amplifier dynamically adjusts the voltage / power (using PID closed-loop control, response time <1ms) to output the excitation signal; the transmitting coil of the magnetic field detector is excited to generate an alternating magnetic field (frequency adjustable from 1-100kHz), covering the quality inspection area of ​​the conveyor line to form a detection zone. When the metal tube passes through, its gap / slit causes a sudden change in magnetic flux (the rate of change can reach up to 500T / s). The receiving coil captures and outputs the signal amplitude and waveform in real time (sampling rate ≥100kS / s); the controller uses a parallel processing architecture: first, it performs moving average filtering (window width adaptive to production speed) and dual threshold segmentation (gap / slit threshold) on the signal to identify adhesion signals; then... The counting and separation model is launched—completing four real-time analysis steps within 5ms: ① Threshold segmentation layer filters out background noise (Gaussian filter kernel size dynamically optimized); ② Peak location layer detects local maxima by combining a time window (Δt=0.1T, where T is the tube's passage period); ③ Time segmentation layer calculates the interval between adjacent peaks and the integral area (FFT accelerated calculation); ④ Morphological verification layer uses adaptive structural elements (rectangular / circular kernel switching) to perform opening operations to separate adhered areas; finally, the results are verified by a dynamic correction layer (threshold adjusted based on the real-time signal mean μ and standard deviation σ), and an electrical signal is output to drive the LED to light up in stages. At the same time, the counter accumulates valid pulses (error rate <0.1%), and the data is synchronized to the PLC system through the EtherCAT protocol of the peripheral interface, realizing millisecond-level closed-loop control of the entire production line process. Example

[0037] Reference Figure 2 The specific steps of step S1 include: According to the preset frequency control word, the signal generator, in conjunction with the clock pulse, accumulates the step value one by one to obtain the phase value; The clock pulse is a time reference signal used to control the timing and stepping rhythm of waveform generation; the stepping value is the increment within each clock cycle. According to the preset waveform amplitude value, the signal generator converts the phase value into the corresponding signal amplitude value; Based on the signal amplitude value, the signal generator outputs corresponding waveform data and converts it into a corresponding analog waveform signal; The analog waveform signal is subjected to voltage gain operation and signal conditioning operation by a voltage amplification factor signal amplifier to obtain a boosted analog signal, which is then compared with a preset voltage threshold; the signal conditioning operation includes impedance matching and filtering. If the boosted analog signal is less than the voltage threshold, the voltage amplification factor of the signal amplifier is corrected using the first correction factor. If the boosted analog signal is greater than or equal to the voltage threshold, the signal amplifier performs a current gain operation on the boosted analog signal according to the current amplification factor to obtain an increased current analog signal, calculates the gain power value with the boosted analog signal, and then compares and judges it with a preset power threshold; If the gain power value is less than the power threshold, the current amplification factor of the signal amplifier is corrected using a second correction coefficient; If the gain power value is greater than or equal to the power threshold, the current increased analog signal is passed through a filter circuit to remove noise, and the signal amplifier outputs an excitation signal; the excitation signal includes a square wave, a sawtooth wave, a triangular wave, a pulse wave, a simple harmonic wave, and a step wave.

[0038] The implementation principle of this embodiment is as follows: high-precision excitation signal generation is achieved through a real-time closed-loop control signal chain. That is, the signal generator is based on DDS technology, and a 32-bit phase accumulator is driven by a 100MHz clock within the FPGA, and a digital waveform is output in real time through a phase-amplitude mapping table (pre-stored 12bit waveform data), and a basic analog signal is generated through a 16-bit DAC (conversion rate 1μs); the signal amplifier adopts a two-stage dynamic regulation architecture: Voltage gain layer: Based on an instrumentation amplifier, impedance matching and band-pass filtering (-3dB bandwidth 10Hz - 1MHz) are performed. The output signal is compared with a preset voltage threshold V_th by a high-speed comparator (response time 10ns). If it is lower than the threshold, the first correction coefficient k_v is calculated in real time using the PID algorithm (k_v = 1 + α·(V_th - V_act) / V_th, α = 0.05 - 0.2), and the gain multiple is dynamically increased; Current drive layer: When the voltage meets the standard, current amplification is performed through an H-bridge power amplifier, and the instantaneous power P_out = V_rms × I_rms is calculated synchronously and compared with the power threshold P_th: If P_out < P_th, the second correction coefficient k_i is generated using the gradient descent method based on the power error ΔP (k_i ← k_i + β·ΔP, β = 0.01), and the bias voltage is updated; Noise suppression layer: The power达标 signal is passed through an elliptic filter (stopband attenuation -60dB) to eliminate switching noise, and finally six types of standard waveforms are output (square wave rising edge < 50ns, sine wave THD < 0.1%). Embodiment

[0039] The specific steps in step S2 include: The excitation signal output by the signal amplifier is received through the drive circuit of the magnetic field detector and transmitted to the transmitting coil of the magnetic field detector; The transmitting coil generates an alternating current according to the excitation signal, and generates a magnetic field in the space surrounding the transmitting coil according to Ampere's circuital law. The magnetic field includes alternating magnetic field, pulsating magnetic field, and pulsed magnetic field; The surrounding space is a magnetic field distribution, including a near-field region, a transition region, and a far-field region; The magnetic field propagates outward in the form of electromagnetic waves, covering the quality inspection area of ​​the metal pipe conveyor line and forming a counting and detection area.

[0040] The implementation principle of this embodiment is as follows: The driving circuit receives the excitation signal through a real-time impedance matching network (dynamically adjusting LC parameters, response time <10μs), drives the transmitting coil (wound with Litz wire, Q value >50) to generate a high-intensity alternating current (peak value 20A@100kHz), and excites three types of magnetic fields in space based on Ampere's circuital law: Alternating magnetic field: radiates in the form of electromagnetic waves (frequency adjustable from 1 to 100 kHz), and forms a high gradient magnetic field (intensity 0.1-1 T / m) in the near field region (distance < λ / 2π). Pulsating magnetic field: Directional penetration is achieved by modulating the carrier wave (duty cycle 0.1-0.9 programmable); Pulsed magnetic field: eddy current effect is excited by nanosecond-level rise time (<100ns); Magnetic field distribution is divided spatially as follows: Near-field region (0-30cm): Magnetic induction intensity B∝1 / r³, forming a high-sensitivity detection area; Transition zone (30cm-1m): B∝1 / r², used for testing large-size metal tubes; Far-field region (>1m): B∝1 / r, attenuation is suppressed by waveguide structure; By optimizing the Helmholtz coil (coaxial double coil spacing / diameter ratio = 0.8), a uniform counting detection zone (magnetic field non-uniformity within ±2%) is constructed in the quality inspection area of ​​the conveyor line, which induces a sudden change in electromagnetic properties when the metal tube passes through. Example

[0041] Reference Figure 3 The specific steps in step S3 include: When the metal tube passes through the counting and detection area, eddy currents are generated on the surface of the metal tube according to the principle of electromagnetic induction, and the rate of change of magnetic flux is calculated. The magnetic flux change rate is compared with the preset adjacent gap magnetic flux threshold and the preset self-gap magnetic flux threshold to determine the gap between adjacent metal tubes or the gap in the metal tube body. If the rate of change of magnetic flux is greater than or equal to the magnetic flux threshold of the adjacent gap, and less than the magnetic flux threshold of its own gap, then it is determined to be an adjacent metal tube gap; If the rate of change of magnetic flux is greater than or equal to the magnetic flux threshold of the gap itself, it is determined to be a gap in the metal tube body; If the rate of change of magnetic flux is less than the magnetic flux threshold of the adjacent gap, it is determined to be an invalid signal or background noise; The receiving coil senses the rate of change of magnetic flux and generates a corresponding signal amplitude and signal waveform.

[0042] The implementation principle of this embodiment is as follows: The receiving coil (using a differential structure with a sensitivity of 0.1mV / μT) captures the change in the eddy current magnetic field excited when the metal tube enters the detection area in real time. The magnetic flux change rate ΔΦ / Δt (accuracy ±3%) is calculated by a differentiating circuit (time constant τ=50μs), and dynamic judgment of dual thresholds is performed simultaneously. Gap identification: If ΔΦ / Δt ∈ [Φ_gap, Φ_flaw) (Φ_gap=5mWb / s, Φ_flaw=20mWb / s), it is determined to be a gap between adjacent steel pipes, triggering the gap measurement algorithm (based on Δt to calculate the gap width); Crack detection: If ΔΦ / Δt ≥ Φ_flaw, start the transient analysis module (sampling rate 1MS / s), extract the rising slope of the signal (>100V / ms indicates crack characteristics), and combine it with the waveform symmetry index (ASI>0.7 indicates a crack); The receiving coil synchronously outputs signal amplitude (range ±10V) and waveform (converted by 16-bit ADC). The spectrum characteristics are analyzed by FFT (abnormalities are marked by main frequency offset >5%), providing four-dimensional data to the controller: magnetic flux, amplitude, waveform, and spectrum, achieving a defect classification accuracy of >99.2% (actual measured data). Example

[0043] The specific steps in step S4 include: The signal threshold range includes minimum amplitude, maximum amplitude, and duration thresholds; The signal amplitude and the signal waveform are filtered and denoised to obtain the effective amplitude and the effective waveform. Waveform features are extracted from the valid waveform to obtain the peak duration, and then compared with a preset signal threshold range. If the effective amplitude is greater than or equal to the minimum amplitude and less than or equal to the maximum amplitude, and the peak duration is greater than the duration threshold, then the current signal amplitude is determined to be the amplitude of the stuck metal tube signal. If the effective amplitude is greater than or equal to the minimum amplitude and less than or equal to the maximum amplitude, but the peak duration is equal to the duration threshold, then the current signal amplitude is determined to be the amplitude of a single metal tube signal. If the effective amplitude is greater than or equal to the minimum amplitude and less than or equal to the maximum amplitude, but the peak duration is less than the duration threshold, it is determined that the current signal amplitude is an invalid signal amplitude; If the effective amplitude is less than the minimum amplitude or greater than the maximum amplitude, it is determined that the current signal amplitude is an invalid signal amplitude; Transmit the signal amplitude of the single metal tube to the controller to generate a corresponding electrical signal, and at the same time transmit it to the counter to count the number of metal tubes; Cluster the signal amplitudes of the adhered metal tubes and the corresponding effective waveform diagrams to obtain the adhered overlapping signals.

[0044] The implementation principle of this embodiment is as follows: The controller realizes accurate counting of metal tubes through a real-time signal processing chain: First, use adaptive Gaussian filtering (window width σ = 0.5 × tube speed) to denoise the original signal, and extract the effective amplitude (accuracy ±0.1V) and waveform diagram; Based on the gradient detection algorithm (third-order derivative zero-crossing method), calculate the peak duration T_p (resolution 1ms), and combine triple-threshold dynamic determination: Single tube recognition: When the effective amplitude ∈ [V_min, V_max] (such as 2 - 10V) and T_p = T_ref (reference value 50ms), trigger the counter to increment and activate the green LED; Adhered tube detection: If T_p > T_ref + ΔT (ΔT = 5ms), start K-means clustering (feature dimension: amplitude integral area + waveform kurtosis) to separate the overlapping signals (maximum support for 4 tubes to adhere); Invalid signal rejection: When T_p < T_ref or the amplitude exceeds the limit, enable spectral entropy analysis (FFT 1024 points), and when the entropy value > 7.5, it is determined as electromagnetic interference; The adhered signals are clustered by DBSCAN (ε = 0.1V, minPts = 3) to generate sub-waveform clusters, and the separation results are verified through a convolutional neural network (CNN, lightweight MobileNetV3) (accuracy 98.7%), output an electrical signal to drive the yellow LED to alarm, and synchronously update the counter (compensation value = number of clusters). Embodiment

[0045] The specific steps in step S5 include: The threshold segmentation layer uses a moving average filtering algorithm for the adhered overlapping signals to denoise and obtain the noise-free adhered signals , and compare and judge with a preset signal threshold :

[0046] Among them, <​​For the size of the filter window, This is the sample location sequence number. For the overall sample index; like Then the current noiseless adhesion signal is marked as a valid detection signal; if If so, the current noiseless adhesion signal is marked as an invalid background signal; The peak location layer uses a maximum detection algorithm to detect the effective detection signal, obtain several signal peaks, and record the peak time; like and If it is, then it is marked as the signal peak value; if If so, it is determined to be a non-signal peak. in, For time window, Peak threshold This is a multiple of the base number.

[0047] The implementation principle of this embodiment is as follows: Adhesive signals are separated through a two-stage real-time processing model—the threshold segmentation layer uses a dynamic window moving average filter (window size N = 2k + 1, k adaptively adjusts with signal frequency, ranging from 3 to 31) to eliminate high-frequency noise, outputting a noise-free signal and comparing it with a preset threshold (when...). (Time markers are considered valid signals; otherwise, they are discarded); the peak localization layer is based on a dual-condition extremum detection algorithm, scanning for signals that meet the conditions within a time window (default 10ms). and , ( The system identifies local maxima points (with 1.2 as a base multiple) and outputs peak coordinates and timestamps in real time, providing millisecond-level positioning data for subsequent counting (processing latency <0.5ms). Example

[0048] The specific steps in step S5 also include: The time division layer specifies the time interval between adjacent signal peaks. and overlapping area Perform calculations and compare with a preset time interval threshold. and preset integration area threshold Comparison and judgment; ,

[0049] in, The peak value number of the signal; like and If the signal peak value is positive, then the number of metal tubes corresponding to the current signal peak value is determined to be two; otherwise, it is determined to be a single metal tube. The morphological verification layer utilizes structural elements. Perform morphological operations on the noise-free adhesion signal to obtain a morphological region ;

[0050] Among them, is the erosion operation, including a rectangle and a Gaussian function, is the dilation operation; Statistically analyze the independent peak regions in the morphological region to obtain the number of metal pipes, and compare it with the number of metal pipes obtained from the time segmentation layer: If they are consistent, generate corresponding electrical signals through the controller to activate the illuminator and transmit them to the counter at the same time; If they are inconsistent, then use the counting separation model to correct the preset signal threshold according to the real-time signal amplitude characteristics ;

[0051] Among them, is the mean value of the current signal amplitude, is the standard deviation of the current signal amplitude, is the correction coefficient; Calculate the amplitude difference between adjacent signal peaks and compare it with the preset amplitude difference threshold to verify the number of metal pipes; If , it is determined that the number of metal pipes corresponding to the current signal peak is two; otherwise, it is determined to be a single metal pipe.

[0052] The implementation principle of this embodiment is as follows: The time segmentation layer calculates the adjacent peak time interval Δt and the integral area A in real time. When Δt < T_min (the minimum interval threshold, default 0.2s) and A > τ_area (the area threshold, taking 1.5 times the single-pipe area), it is determined that there is double-pipe adhesion; the morphological verification layer uses multi-core morphological processing: first, perform erosion operation with a rectangular structural element B_rect (size 3×3) to eliminate burrs, and then dilate with a Gaussian kernel B_gauss (σ = 0.8) to restore the signal contour, separate the adhesion region and count the number of independent peaks; when the results of the two layers conflict, the dynamic correction layer updates the threshold based on the real-time signal mean μ and standard deviation σ according to θ_new = μ + σ ( = 1.5~2.5), and at the same time verify that the amplitude difference |Δp| = |p_i - p_{i - 1}| > δ (δ = 0.3×p_avg) to strengthen the double-pipe determination. Finally, generate electrical signals through the controller to drive the counting (processing delay < 10ms, separation accuracy 98%). Example

[0053] Reference Figure 4 An automated counting device for a metal pipe conveyor line, applied to the aforementioned automated counting method, comprising: A signal generator is used to generate analog waveform signals according to a preset frequency control word. A signal amplifier is used to amplify the voltage and power of the analog waveform signal to obtain an excitation signal; A magnetic field detector is used to generate a magnetic field based on the excitation signal and cover the quality inspection area of ​​the metal pipe conveyor line to construct a counting detection area. At the same time, it generates signal amplitude and signal waveform when the metal pipe passes through the counting detection area. The controller is used to determine the gap distance between the metal tubes corresponding to the signal amplitude and the signal waveform diagram based on a preset signal threshold range, obtain the adhesion and overlap signal, segment it, determine the number of metal tubes, and generate a corresponding number of electrical signals. A counter is used to record the number of electrical signals and to count the number of metal tubes. The light emitter includes several LED beads; The light emitter is used to activate the corresponding LED beads based on the electrical signal and a preset signal threshold range; Peripheral interface for connecting to third-party devices; the third-party devices include PLCs; A circuit board for integrating the signal generator, the signal amplifier, the magnetic field detector, the controller, the counter, the light emitter, and the peripheral interface.

[0054] The signal generator is connected to the signal amplifier via wires on the circuit board. The signal amplifier is connected to the magnetic field detector via wires. The magnetic field detector is connected to the controller via wires. The controller is connected to the counter, the light emitter, and the peripheral interface via wires.

[0055] The implementation principle of this embodiment is as follows: High-speed and accurate counting of metal tubes is achieved through a fully hardware-integrated signal chain. The signal generator, based on DDS technology, drives a 32-bit phase accumulator within the FPGA with a 100MHz clock to generate digital waveforms, which are then output as six basic analog signals via a 16-bit DAC (conversion delay 1μs). The signal amplifier adopts a two-stage dynamic gain architecture: the voltage stage uses an instrumentation amplifier (adjustable gain 40-60dB) to perform impedance matching and bandpass filtering (10Hz-1MHz), while the current stage is driven by an H-bridge power amplifier (efficiency ≥85%), and the voltage / power threshold is compared in real time. The gain coefficient is dynamically corrected using a PID algorithm (iteration period 5μs) with a response time of 200ns. The magnetic field detector uses a Helmholtz coil to excite three types of magnetic fields (alternating / pulsating / pulsating) to create a uniform detection zone (non-uniformity <±2%) 30cm above the conveyor line. When the metal tube passes through, the receiving coil (differential structure, sensitivity 0.1mV / μT) captures the magnetic flux change (ΔΦ / Δt accuracy ±3%). After dual threshold judgment (gap threshold 5mWb / s, slit threshold 20mWb / s), the signal amplitude and waveform are output. The controller performs five layers of real-time processing. Adaptive Gaussian filtering (window σ = 0.5 × tube speed) extracts the effective signal. Triple threshold determination (amplitude limit range 2-10V, peak duration baseline 50ms) identifies single tubes / adhered tubes. The sticky signal is filtered by moving average (dynamic window N=0.2f_s+1) and subjected to dual-condition peak detection (local maximum + amplitude > 1.2θ_th). The time-division layer calculates the interval Δt between adjacent peaks and the integral area A (accelerated by the trapezoidal method). The morphology verification layer uses multi-kernel morphology (3×3 rectangular erosion + σ=0.8 Gaussian dilation) to separate the adhesion regions. When the results conflict, the dynamic correction layer updates the threshold by θ_new=μ+1.8σ and verifies it by amplitude difference (|Δp|>0.3p_avg). Finally, when the steel pipe passes through the detection area, the probe detects the gaps on the steel pipe or the gaps between adjacent steel pipes by sensing changes in the magnetic field, and lights up the LEDs at different positions according to the different amplitudes of the changes in the gap size.

[0056] When a gap between adjacent steel pipes is detected, the leftmost LED bead is triggered to light up, and an electrical signal is simultaneously output to the PLC control system for counting.

[0057] The embodiments described herein are preferred embodiments of the present invention and are not intended to limit the scope of protection of the present invention. Therefore, all equivalent changes made in accordance with the structure, shape, and principle of the present invention should be covered within the scope of protection of the present invention.

Claims

1. An automated counting device for a metal pipe conveyor line, characterized in that, include: A signal generator is used to generate analog waveform signals according to a preset frequency control word. A signal amplifier is used to amplify the voltage and power of the analog waveform signal to obtain an excitation signal; A magnetic field detector is used to generate a magnetic field based on the excitation signal and cover the quality inspection area of ​​the metal pipe conveyor line to construct a counting detection area. At the same time, it generates signal amplitude and signal waveform when the metal pipe passes through the counting detection area. The controller is used to determine the gap distance between the metal tubes corresponding to the signal amplitude and the signal waveform diagram based on a preset signal threshold range, obtain the adhesion and overlap signal, segment it, determine the number of metal tubes, and generate a corresponding number of electrical signals. A counter is used to record the number of electrical signals and to count the number of metal tubes. The signal generator is connected to the signal amplifier via a wire, the signal amplifier is connected to the magnetic field detector via a wire, the magnetic field detector is connected to the controller via a wire, and the controller is connected to the counter via a wire.

2. The automated counting device for metal pipe conveying lines according to claim 1, characterized in that, The automated counting device also includes: A light emitter is activated based on the electrical signal and a preset signal threshold range. Peripheral interface for connecting to third-party devices; the third-party devices include PLCs; A circuit board for integrating the signal generator, the signal amplifier, the magnetic field detector, the controller, the counter, the light emitter, and the peripheral interface.

3. The automated counting device for metal pipe conveying lines according to claim 2, characterized in that: The light emitter includes several LED beads, with one LED bead corresponding to each preset signal threshold range.

4. An automated counting method for a metal pipe conveyor line, applied to the automated counting device according to any one of claims 1-3; characterized in that, include: The signal generator generates an analog waveform signal according to the preset frequency control word, and after passing through the signal amplifier, an excitation signal is obtained. The magnetic field detector generates a magnetic field according to the excitation signal, and determines the quality inspection area of ​​the metal pipe conveyor line covered by the magnetic field as the counting detection area; When passing through a metal tube within the counting detection area, the magnetic field detector generates a signal amplitude and a signal waveform diagram based on the gap between adjacent metal tubes or the gap in the metal tube body. Based on the preset signal threshold range controller, the signal amplitude and the metal tube gap distance corresponding to the signal waveform are determined to obtain the adhesion and overlap signal; The overlapping signal is segmented using a counting separation model to determine the number of metal tubes, generate a corresponding number of electrical signals, and transmit them to the counter to count the number of metal tubes.

5. The automated counting method for metal pipe conveying lines according to claim 4, characterized in that, The specific steps of generating an analog waveform signal according to a preset frequency control word, and then amplifying it to obtain an excitation signal include: According to the preset frequency control word, the signal generator, in conjunction with the clock pulse, accumulates the step value one by one to obtain the phase value; The clock pulse is a time reference signal used to control the timing and stepping rhythm of waveform generation; the stepping value is the increment within each clock cycle. According to the preset waveform amplitude value, the signal generator converts the phase value into the corresponding signal amplitude value; Based on the signal amplitude value, the signal generator outputs corresponding waveform data and converts it into a corresponding analog waveform signal; The analog waveform signal is subjected to voltage gain operation and signal conditioning operation by a voltage amplification factor signal amplifier to obtain a boosted analog signal, which is then compared with a preset voltage threshold; the signal conditioning operation includes impedance matching and filtering. If the boosted analog signal is greater than or equal to the voltage threshold, the signal amplifier performs a current gain operation on the boosted analog signal according to the current amplification factor to obtain a current boosted analog signal, calculates the gain power value with the boosted analog signal, and then compares it with the preset power threshold for judgment. If the gain power value is greater than or equal to the power threshold, the current current-boosting analog signal is filtered by a filter circuit to remove noise, and the signal amplifier outputs an excitation signal.

6. The automated counting method for metal pipe conveying lines according to claim 4, characterized in that, The specific steps of generating a magnetic field by the magnetic field detector based on the excitation signal and determining the quality inspection area of ​​the metal pipe conveyor line covered by the magnetic field as the counting detection area include: The excitation signal output by the signal amplifier is received by the driving circuit of the magnetic field detector and transmitted to the transmitting coil of the magnetic field detector. The transmitting coil generates an alternating current according to the excitation signal, and generates a magnetic field in the space surrounding the transmitting coil according to Ampere's circuital law. The magnetic field includes alternating magnetic field, pulsating magnetic field, and pulsed magnetic field; The surrounding space is a magnetic field distribution, including a near-field region, a transition region, and a far-field region; The magnetic field propagates outward in the form of electromagnetic waves, covering the quality inspection area of ​​the metal pipe conveyor line and forming a counting and detection area.

7. The automated counting method for metal pipe conveying lines according to claim 4, characterized in that, When the magnetic field detector passes through a metal tube within the counting detection area, the specific steps for generating the signal amplitude and signal waveform based on the gap between adjacent metal tubes or the gap in the metal tube body include: When the metal tube passes through the counting and detection area, eddy currents are generated on the surface of the metal tube according to the principle of electromagnetic induction, and the rate of change of magnetic flux is calculated. The magnetic flux change rate is compared with the preset adjacent gap magnetic flux threshold and the preset self-gap magnetic flux threshold to determine the gap between adjacent metal tubes or the gap in the metal tube body. If the rate of change of magnetic flux is greater than or equal to the magnetic flux threshold of the adjacent gap and less than the magnetic flux threshold of its own gap, it is determined to be a gap between adjacent metal tubes. The receiving coil senses the rate of change of magnetic flux and generates a corresponding signal amplitude and signal waveform.

8. The automated counting method for metal pipe conveying lines according to claim 4, characterized in that, The specific steps for determining the adhesion and overlap signal by analyzing the signal amplitude and the metal tube gap distance corresponding to the signal waveform based on a preset signal threshold range controller include: The signal threshold range includes an amplitude limiting interval and a duration threshold; the amplitude limiting interval includes a minimum amplitude and a maximum amplitude. The signal amplitude and the signal waveform are filtered and denoised to obtain the effective amplitude and the effective waveform. Waveform feature extraction is performed on the effective waveform to obtain the peak duration: If the effective amplitude is within the amplitude limit range and the peak duration is greater than the duration threshold, then the current signal amplitude is determined to be the amplitude of the bonded metal tube signal. Clustering the signal amplitude and corresponding effective waveform of the adhered metal tubes yields the adhered and overlapping signals.

9. The automated counting method for metal pipe conveying lines according to claim 4, characterized in that, The specific steps of segmenting the overlapping signals using a counting separation model, determining the number of metal tubes, generating a corresponding number of electrical signals, and transmitting them to the counter to count the number of metal tubes include: The counting separation model uses a moving average filtering algorithm to analyze the overlapping signals. Denoising was performed to obtain a noise-free, sticky signal. and with preset signal threshold Comparison and judgment: in, For the sample time, For the size of the filter window, For sample location sequence number, For the overall sample index; like If so, the current noiseless adhesion signal is marked as a valid detection signal; The counting separation model uses a maximum detection algorithm to detect the effective detection signal, obtain several signal peaks, and record the peak time; like and If so, it is marked as the signal peak value; in, For time window, Peak threshold This is a multiple of the base number.

10. The automated counting method for a metal pipe conveyor line according to claim 9, characterized in that, The specific steps of segmenting the overlapping signals using a counting separation model, determining the number of metal tubes, generating a corresponding number of electrical signals, and transmitting them to the counter to count the number of metal tubes further include: The counting separation model considers the time interval between adjacent signal peaks. and overlapping area Perform calculations and compare with a preset time interval threshold. and preset integration area threshold Comparison and judgment; , in, The peak value number of the signal; like and If the signal peak value is positive, then the number of metal tubes corresponding to the current signal peak value is determined to be two; otherwise, it is determined to be a single metal tube. The counting separation model utilizes structural elements. For noiseless adhesion signals Perform morphological operations to obtain morphological regions. ; in, For corrosion operation, Includes rectangular and Gaussian functions. For expansion operation; Morphological regions The number of metal tubes was obtained by statistically analyzing the independent peak regions in the data, and then compared with the number of metal tubes obtained from the time segmentation layer. If the two are consistent, the controller generates a corresponding number of electrical signals to activate the light emitter and transmit them to the counter. If the two are inconsistent, the preset signal threshold is determined by the counting separation model based on the real-time signal amplitude characteristics. Make corrections; in, The average value of the current signal amplitude. The standard deviation of the current signal amplitude. This is a correction factor; For adjacent signal peaks The amplitude difference is calculated and compared with a preset amplitude difference threshold. The comparison was then made to verify the number of metal tubes. like If the signal peak value is positive, the number of metal tubes corresponding to the current signal peak value is determined to be two; otherwise, it is determined to be a single metal tube.

Images