Cleaning equipment gear box abnormal sound detection method and system based on current characteristic analysis

Abstract

This application discloses a method and system for detecting abnormal noise in the gearbox of cleaning equipment based on current characteristic analysis. The method includes: acquiring the motor current signal when the gearbox is running under no-load; extracting the periodic component corresponding to the motor rotation to determine the instantaneous motor angle; and performing angular domain resampling; dividing the angular domain current signal into segments according to the entire gear revolution; superimposing and averaging multiple rotation cycles at the same angular position to obtain a synchronous average signal; and determining the deviation area based on the baseline and deviation detection threshold; further extracting the peak amplitude and rise angle span of the deviation; calculating the steepness index in conjunction with the baseline; and comparing it with the abnormal noise judgment threshold to determine whether the gearbox has an abnormal noise defect. This application can reduce the influence of speed fluctuations and random noise on the detection results, improve the stability of identifying local tooth surface anomalies, and enhance the consistency of abnormal noise judgment.

Description

Technical Field

[0001] This application relates to the field of gearbox condition detection technology, specifically to a method and system for detecting abnormal noise in the gearbox of cleaning equipment based on current characteristic analysis. Background Technology

[0002] As cleaning equipment such as handheld vacuum cleaners, sweepers, and electric mops develop towards miniaturization, high speed, and low noise, the assembly quality and operating noise of their internal micro-reduction gearboxes have become important factors affecting product user experience and final product quality. During gear machining and assembly, defects such as burrs on the tooth surface, localized flash, pits, and tooth pitch deviations can cause periodic meshing impacts, manifesting as abnormal noises such as clicking or rattling sounds. Current production line final inspections typically rely on manual listening or using microphones and vibration sensors to collect acoustic or vibration signals for judgment. Manual listening is significantly affected by operator experience, fatigue levels, and ambient noise, resulting in limited consistency and traceability. Acoustic and vibration detection, on the other hand, places certain requirements on sensor installation location, tooling structure, and environmental vibration isolation and noise reduction conditions.

[0003] Utilizing the drive motor current to reflect changes in mechanical load is also a technical approach for gearbox condition monitoring. Existing methods typically analyze the average current, RMS value, or meshing-related components in the spectrum. However, for sharp, localized defects occurring only at individual tooth positions, the resulting current changes are short-lived, and the average current may still be within the normal range. Furthermore, during no-load testing of small gearboxes, motor speed fluctuates. When performing spectral or periodic feature analysis based on a fixed time scale, gear meshing-related features are prone to positional drift or spectral broadening, hindering the stable identification of localized tooth surface anomalies. Therefore, further improvements are needed to enhance the objective identification capability of localized tooth-level anomalies and the consistency of test results while adapting to rapid production line testing conditions. Summary of the Invention

[0004] To overcome the problems existing in related technologies, this specification provides a method and system for detecting abnormal noise in the gearbox of cleaning equipment based on current characteristic analysis.

[0005] According to a first aspect of the embodiments of this application, a method for detecting abnormal noise in the gearbox of a cleaning device based on current characteristic analysis is provided, including: The motor current signal of the cleaning equipment gearbox is collected when it is running under no-load; the periodic component corresponding to the motor rotation is extracted from the motor current signal, and the instantaneous rotation angle position of the motor is determined according to the periodic component; the motor current signal is resampled in the angular domain based on the instantaneous rotation angle position of the motor to obtain the angular domain current signal with the gear rotation angle as the independent variable. The angular current signal is segmented according to the angle range of a complete rotation of the gear. The segmented signals of multiple complete rotation cycles are superimposed and averaged at the same angular position to obtain a synchronous average signal. The overall mean of the synchronous average signal is calculated as a baseline. An angular region in the synchronous average signal that deviates from the baseline by more than the deviation detection threshold is detected as a deviation region. For each deviation region, the peak deviation amplitude and the deviation rise angle span are extracted. A steepness index is calculated based on the peak deviation amplitude, the deviation rise angle span, and the baseline. The steepness index is compared with the abnormal noise determination threshold. If the steepness index exceeds the abnormal noise determination threshold, the gearbox is determined to have an abnormal noise defect.

[0006] As an optional approach, the steepness index is the ratio of the peak deviation amplitude to the product of the deviation rise angle span and the baseline; the deviation rise angle span is the angle difference between the angle position where the current deviation in the deviation region first exceeds the deviation detection threshold and the angle position where the current deviation reaches its peak value.

[0007] As an optional approach, extracting the periodic component corresponding to the motor rotation from the motor current signal includes: performing high-pass filtering on the motor current signal to extract the commutation ripple component; performing envelope detection and peak detection on the commutation ripple component to obtain a commutation pulse time sequence; the interval between adjacent commutation pulses corresponds to a fixed angle increment of motor rotation, and the fixed angle increment is determined by the number of commutator segments of the motor.

[0008] As an optional approach, the step of resampling the motor current signal based on the instantaneous rotation angle position of the motor includes: performing linear interpolation on the rotation angle between two adjacent pulse moments in the commutation pulse time sequence, based on the assumption of constant rotation speed, to obtain the motor rotation angle value corresponding to each sampling moment; and performing interpolation resampling on the motor current signal at equal angular intervals to obtain the angular current signal.

[0009] As an optional approach, determining the deviation detection threshold specifically involves: calculating the standard deviation of the point-by-point difference between the segmented signal of each rotation cycle and the synchronous average signal at the same angular position; attenuating the standard deviation according to the number of rotation cycles involved in the superposition and averaging; and multiplying it by a preset detection factor; the detection factor ranges from 3 to 5 and is determined based on the production line false alarm rate requirements during the first batch of verifications.

[0010] As an optional approach, the abnormal noise judgment threshold is determined by statistical analysis of calibration samples of the same model of gearbox, including: taking qualified samples that have been manually confirmed to be free of abnormal noise, calculating the kurtosis index for each qualified sample according to the method described above, and calculating the mean and standard deviation of the kurtosis index for all qualified samples; adding the mean to the product of the standard deviation and the calibration coefficient as the abnormal noise judgment threshold, wherein the calibration coefficient ranges from 3 to 5 and is determined based on the distribution interval between the maximum value of the kurtosis index of qualified samples and the minimum value of the kurtosis index of unqualified samples.

[0011] As an optional approach, the number of angular domain sampling points in each gear rotation cycle of the angular domain current signal shall not be less than four times the number of teeth of the gear under test.

[0012] As an optional approach, when the angular current signal is segmented according to the angular range of one full revolution of the gear, the number of commutation pulses corresponding to one full revolution of the gear is the product of the gearbox transmission ratio and the number of motor commutator segments.

[0013] As an optional solution, when it is determined that the gearbox has an abnormal noise defect, the angle position corresponding to the deviation area is divided by the angle range corresponding to a single tooth and rounded to obtain the defective tooth number.

[0014] According to a second aspect of the embodiments of this application, a gearbox noise detection system for cleaning equipment based on current characteristic analysis is also provided, comprising: The current acquisition and angle domain resampling module is used to acquire the motor current signal when the gearbox of the cleaning equipment is running under no-load; extract the periodic component corresponding to the motor rotation from the motor current signal, determine the instantaneous rotation angle position of the motor based on the periodic component; and perform angle domain resampling on the motor current signal based on the instantaneous rotation angle position of the motor to obtain the angle domain current signal with the gear rotation angle as the independent variable. The synchronous averaging and deviation detection module is used to segment the angular domain current signal according to the angle range of a complete gear revolution, and to superimpose and average the segmented signals of multiple complete rotation cycles at the same angular position to obtain a synchronous average signal; to calculate the overall mean of the synchronous average signal as a baseline, and to detect the angular region in the synchronous average signal that deviates from the baseline by more than the deviation detection threshold as the deviation region; The abnormal noise detection module is used to extract the peak amplitude of the deviation and the range of the deviation rise angle for each deviation area, calculate the steepness index based on the peak amplitude of the deviation, the range of the deviation rise angle and the baseline; compare the steepness index with the abnormal noise detection threshold, and determine that the gearbox has an abnormal noise defect when the steepness index exceeds the abnormal noise detection threshold.

[0015] This application utilizes the existing motor current signal from the unloaded operation of the gearbox of the cleaning equipment to extract the periodic component corresponding to the motor rotation, and establishes the instantaneous rotation angle position accordingly. The time-domain current affected by speed fluctuations is converted into angular-domain current, enabling the meshing abnormalities of fixed tooth positions to be stably characterized under a unified angular reference. On this basis, the synchronous superposition of multiple complete rotation cycles suppresses random disturbances unrelated to the gear angle, highlighting the repetitive current deviation caused by local tooth surface defects. Furthermore, the steepness index is constructed by combining the peak amplitude of the deviation, the range of the deviation rise angle, and the baseline, so that local sharp impacts and slow current fluctuations can be distinguished according to a unified quantitative standard. Thus, tooth-level defects that may cause abnormal noise can be identified while the average current is still within the normal range. The defect tooth number can be further determined based on the angular position corresponding to the deviation, improving the consistency, stability, and traceability of abnormal noise detection on the production line. Attached Figure Description

[0016] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with this specification and, together with the description, serve to explain the principles of this specification.

[0017] Figure 1 This is a flowchart of a method for detecting abnormal noise in the gearbox of a cleaning equipment based on current characteristic analysis, according to one embodiment of this application.

[0018] Figure 2 This is a flowchart of corner current signal acquisition in one embodiment of this application.

[0019] Figure 3 This is a schematic diagram of the motor current angular domain resampling principle in one embodiment of this application.

[0020] Figure 4 This is a schematic diagram of the deviation area determination process in one embodiment of this application.

[0021] Figure 5 This is a schematic diagram of synchronous averaging and abnormal sound determination in one embodiment of this application.

[0022] Figure 6 This is a schematic block diagram of a gearbox noise detection system for cleaning equipment based on current characteristic analysis, as described in this application.

[0023] Figure 7 This is a structural diagram of an electronic device according to one embodiment of this application. Detailed Implementation

[0024] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0025] This embodiment provides a method for detecting abnormal noise in the gearbox of cleaning equipment based on current characteristic analysis. It is applicable to the final inspection of miniature reduction gearboxes in small cleaning equipment (such as handheld vacuum cleaners, electric sweepers, electric mops, etc.) on the production line. The gearbox is directly driven by a small DC brushed motor, with a rated operating voltage of, for example, 3 to 12 volts, and an unloaded operating current typically not exceeding 0.3 amps. The testing fixture consists of three parts: a gearbox positioning fixture, a regulated DC power supply, and a current acquisition unit. After the gearbox is inserted into the fixture, the power supply powers the motor, causing the motor to rotate the gearbox under no-load conditions. The current acquisition unit is connected in series in the motor power supply circuit. It converts the circuit current into a voltage signal through a precision sampling resistor or a Hall current sensor, and continuously acquires the signal via an analog-to-digital converter at a certain sampling rate, which can be configured, for example, from 20 kHz to 50 kHz. Throughout the testing process, the gearbox has no external load, and the motor operates at near-no-load rated speed. Optionally, for gearbox models using brushless DC motors, the current commutation abrupt change point can be used to replace the commutation ripple as the rotation reference. The principle is similar to that of brushed motors, except that the angle correspondence needs to be converted according to the number of motor pole pairs.

[0026] The implementation process of the method described in this application will be explained in detail below with reference to specific embodiments.

[0027] Figure 1 This is a flowchart of a method for detecting abnormal noise in the gearbox of a cleaning equipment based on current characteristic analysis, as described in one embodiment of this application. Figure 1 As shown, the method includes steps 101-103: Step 101: Extract the periodic component corresponding to the motor rotation from the motor current signal, determine the instantaneous rotation angle position of the motor, and perform angular domain resampling.

[0028] According to an embodiment of this disclosure, after the motor is powered on, it enters steady-state operation after a brief acceleration. The steady-state determination method is that when the rate of change of the average current within a continuous sampling window is lower than a preset convergence threshold (e.g., 0.5%), steady-state operation is considered achieved. After entering steady-state, formal current data acquisition begins. The acquisition duration is determined based on the gearbox transmission ratio and the required number of gear rotations, for example, 2 to 4 seconds. The acquired raw current timing signal is denoted as... ,in For time.

[0029] This step obtains the motor rotation angle information from the current signal itself, and converts the current from the time domain to the angular domain. The purpose is that the testing fixture is not a precision speed control device; the no-load speed fluctuates during the test (typically ±3% to 5% of the rated speed). If the signal is directly analyzed in the time domain, the gear meshing frequency will broaden in the spectrum as the speed drifts, weakening the ability to identify tooth defects. After switching to the angular domain, regardless of the actual speed change, the gear meshing event always corresponds to a fixed angular position.

[0030] Specifically, Figure 2 This is a flowchart of the corner current signal acquisition in one embodiment of this application. In step 201, the periodic component corresponding to the motor rotation is extracted from the motor current signal.

[0031] The armature winding of a DC brushed motor is switched on and off via a commutator. Each time the winding passes a commutator segment, the brush contact state changes, generating a brief pulse-like ripple in the current waveform. Assume the number of commutator segments in the motor is... (For motors in small cleaning equipment,) For example, if it is 5 or 7), then the motor generates [amount] per revolution. These commutation pulses constitute the periodic components of the current signal that correspond precisely to the motor's rotation.

[0032] Specifically, for the original current signal A high-pass filter is applied to remove DC components and low-frequency fluctuations, retaining commutation ripple and higher frequency components. The high-pass cutoff frequency is set to 40% to 60% of the estimated commutation frequency. The estimated commutation frequency is calculated from the motor's nominal no-load speed and the number of commutator segments. For example, with a nominal speed of 5000 rpm and 5 commutator segments, the commutation frequency is approximately 417 Hz, and the high-pass cutoff frequency can be set to 180 to 250 Hz. After filtering, the absolute value of the signal is taken, and then envelope smoothed using a low-pass filter (cutoff frequency set to 2 to 3 times the commutation frequency) to obtain the amplitude envelope of the commutation pulse.

[0033] For example, peak detection is performed on the envelope, with an amplitude threshold. Set as the root mean square value of the envelope signal 1.5 to 2.5 times, that is , To eliminate interference from noise spikes, a sliding window local maximum search strategy can be used during detection, within a window length... (Set to 30% to 50% of the minimum time interval between adjacent commutation pulses, which is calculated based on the motor's maximum speed and the number of commutator segments) Find local maxima points whose amplitude exceeds a threshold, and arrange the found local maxima points in chronological order; if the time interval between two adjacent candidate pulses is less than the minimum commutation interval... (Taking 70% of the theoretical minimum interval between adjacent commutation pulses to prevent false detection), the pulses with larger amplitudes are retained, ultimately yielding the commutation pulse time sequence. The interval between two adjacent commutation pulses corresponds to the angle through which the motor rotates by one commutator segment, i.e. radian.

[0034] For example, the original current signal Apply a fourth-order Butterworth high-pass filter (IIR type, order) (A cascaded second-order section is used to enhance numerical stability), removing DC components and low-frequency fluctuations, while retaining commutation ripple and higher frequency components. High-pass cutoff frequency. Set as the estimated commutation frequency 40% to 60%, that is: ; Among them, the estimated commutation frequency The nominal no-load speed of the motor (Revolutions per minute) and number of commutator segments calculate: (Hertz). For example, when the nominal speed is 5000 rpm and the number of commutator segments is 5, the commutation frequency is approximately 417 Hz, and the high-pass cutoff frequency can be set to 180 to 250 Hz.

[0035] After filtering, the absolute value of the signal is taken, and then it is passed through a second-order Butterworth low-pass filter (cutoff frequency). Set to 2 to 3 times the commutation frequency, i.e. , Envelope smoothing is performed to obtain the amplitude envelope of the commutation pulse.

[0036] In step 202, the instantaneous rotation angle position of the motor is determined based on the periodic components. Specifically, a mapping from time to motor rotation angle is established using the commutation pulse time sequence. The cumulative motor rotation angle corresponding to each commutation pulse is: ; in For the first The motor rotation angle corresponding to each pulse This represents the number of commutator segments. Between two adjacent pulse moments, since the commutation interval is typically no more than 3 milliseconds, the change in rotational speed during this short period is much less than 1%. Therefore, we can assume the angular velocity is constant and perform linear interpolation on the rotation angle: ; in For being in and At any sampling time between, and These represent the times of adjacent commutation pulses. Through this calculation, each sampling point in the original dataset yields a defined motor rotation angle value.

[0037] In step 203, the motor current signal is resampled in the angular domain based on the instantaneous rotation angle position of the motor to obtain the angular domain current signal.

[0038] In one embodiment, the current signal is resampled at equal angular intervals. The number of angular domain sampling points within each gear rotation cycle is set to... The corresponding angle increment is . The value of needs to meet the requirement of resolving changes in current at the single-tooth level, that is, assuming the number of teeth of the gear under test is . Each lap requires at least Only by sampling one point can the details of the current waveform be distinguished for each tooth. For example At 20:00, Take 80 to 120.

[0039] For cases where the original sampling points do not fall exactly at equal angular positions, the current value at the target angle is obtained by linear interpolation of adjacent time-domain sampling values. After resampling, the angular domain current signal is obtained, represented by the gear rotation angle. is the independent variable. This involves the gear ratio. The conversion relationship is that one revolution of the motor corresponds to one revolution of the output gear shaft. One full revolution of the output gear corresponds to... One commutation pulse. Transmission ratio. These are the design parameters for the gearbox, input into the system as the known configuration of this model before testing.

[0040] Figure 3 This is a schematic diagram illustrating the motor current angular domain resampling principle in one embodiment of this application; as shown... Figure 3 As shown, the original time-domain current signal After high-pass filtering, the commutation ripple corresponding to the motor rotation is extracted, and then the timing of the commutation pulse is determined by envelope detection and peak detection. Because adjacent commutation pulses correspond to a fixed motor rotation angle increment. ,in Since the number of commutator segments in the motor is given, the discrete rotation angle position can be determined based on the timing of each commutation pulse. Linear interpolation is then performed between adjacent pulses under the assumption of constant speed to establish a correspondence between time and the instantaneous rotation angle of the motor. Subsequently, the current signal is interpolated and resampled at equal angular intervals to obtain the angular domain current signal with the gear rotation angle as the independent variable. This allows gear rotation events, which are subject to speed fluctuations and have inconsistent time intervals in the time domain, to be converted to a fixed angular position. Combined with the gearbox transmission ratio... The output gear rotates one full revolution. A reversing pulse provides an angle reference for subsequent segmentation based on the entire gear revolution and synchronous averaging.

[0041] Step 102: Segment the angular current signal according to the angular range of one full rotation of the gear, and superimpose and average them to calculate the baseline and detect deviation areas. The angular current signal is segmented according to the angular range of one full rotation of the gear. The segmented signals from multiple complete rotation cycles are superimposed and averaged at the same angular position to obtain a synchronous average signal. The overall mean of the synchronous average signal is calculated as the baseline, and angular regions in the synchronous average signal that deviate from the baseline beyond the deviation detection threshold are detected as deviation areas.

[0042] This step performs multi-turn synchronous superposition averaging on the angular domain current signal to obtain the gear angular distribution characteristics, then establishes a judgment baseline and detects local deviations. During gear rotation, tooth surface defects at fixed positions recur at the same angle in each revolution, and superposition averaging retains these deterministic components; random noise unrelated to gear angles is uncorrelated between revolutions, and superposition averaging attenuates it by the square root of the number of rotation cycles, thereby improving the ability to identify weak tooth-level defects.

[0043] Specifically, Figure 4 This is a schematic diagram of the deviation area determination process in one embodiment of this application. In step 401, the angular current signal is segmented according to the angle range of a complete gear revolution, and the average signal is obtained by superimposing and averaging.

[0044] Specifically, according to one full revolution of the output gear (i.e. The angular current signal is segmented by radians, and the number of commutation pulses corresponding to one full revolution is the transmission ratio. With the number of commutator segments The product of these is used as the boundary for segmented pulse counting. Let the total number of pulses within the acquisition period be... One complete rotation cycle, For example, take 20 to 30. The angular current waveform of the loop is denoted as Each lap contains 1. Equally spaced sampling points. Take the arithmetic mean point-by-point for each lap at the same angular position: ; in The signal is a synchronous average, measured in amperes. The number of rotation cycles involved in the average; For the first Circle at the angle The current value at the point, in amperes. Taking 25 as an example, the random noise component is attenuated to 1 / 5 of its original value, while the amplitude of the deterministic component related to the gear angle remains unchanged.

[0045] In step S402, the overall mean of the synchronization average signal is calculated as the baseline, and the angle region that deviates from the baseline by more than the deviation detection threshold is identified as the deviation region.

[0046] Specifically, for The overall mean is obtained by taking the arithmetic mean over a complete circle. This mean is used as the baseline, denoted as . The unit is amperes. The baseline reflects the average frictional loss level of the gearbox during no-load operation.

[0047] The deviation detection threshold was then determined. This is used to distinguish between effective deviation and residual noise in a synchronously averaged signal. The method for determining this is to calculate the segmented signal for each rotation cycle. With synchronous average signal For each point difference at the same angular position, calculate the standard deviation of all differences to obtain the residual standard deviation. (Unit: Ampere), then the deviation detection threshold is obtained according to the following relationship: ; in To detect the multiple, the value range is 3 to 5; The standard deviation of the residuals; This represents the average number of rotation cycles used. A value of 3 corresponds to approximately 99.7% noise elimination probability; a value of 5 further reduces the risk of false alarms. The specific value will be selected and fixed during the initial production line validation based on the allowable false alarm rate level of the production line. For example, if... For 0.002 amperes, The value is 25, take... If it is 4, then ampere.

[0048] The specific detection process involves scanning the synchronous average signal one by one. All angles and positions, Compared with baseline The absolute value of the difference exceeds The continuous angular range is marked as a deviation area. If there is no deviation area within the entire circumference, it means that there are no detectable tooth surface defects in the gearbox, and it is directly judged as qualified.

[0049] Step 103: Extract the peak amplitude and rise angle span of the deviation region, calculate the steepness index, and compare it with the abnormal noise determination threshold. For each deviation region, extract the peak amplitude and rise angle span of the deviation, calculate the steepness index based on the peak amplitude, the rise angle span, and the baseline; compare the steepness index with the abnormal noise determination threshold, and if the steepness index exceeds the abnormal noise determination threshold, determine that the gearbox has an abnormal noise defect.

[0050] In some embodiments, the local morphology of the waveform of the deviation area detected in the aforementioned steps is analyzed to quantify the steepness of the deviation, thereby determining whether the corresponding defect causes audible abnormal noise. The physical basis is that sharp defects on the tooth surface (such as burrs, pits, sharp edges, and local flash) generate rapid force changes at the moment of meshing, concentrated within a very short angle range. The corresponding mechanical impact excitation covers a wide frequency band, encompassing the 1 to 4 kHz range sensitive to human hearing, producing "clicking" or "clattering" abnormal noises. In contrast, the smooth current fluctuations caused by slight gear eccentricity or slow, gradual changes in tooth pitch result in slow force changes and frequencies concentrated near the fundamental rotational frequency below tens of hertz, not constituting audible abnormal noise. The steeper the deviation waveform, the more intense the corresponding impact, and the higher the likelihood of producing audible abnormal noise.

[0051] In one embodiment, the peak deviation amplitude and the deviation rise angle span are extracted for the deviation region. Specifically, two feature quantities are extracted for each deviation region. The first is the peak deviation amplitude. This refers to the deviation area. Deviation from baseline The maximum absolute value, in amperes, reflects the extreme magnitude of the current deviation caused by the defect.

[0052] The second is the range of the deviation's rising angle. This refers to the first time the current deviation within the deviation region exceeds the deviation detection threshold. The angle difference between the angular position and the angular position where the current deviation reaches its peak value, in radians. The smaller the value, the narrower the angle range from near the detection threshold to the peak value, and the sharper the waveform.

[0053] For example, in a certain deviation region, It first exceeded the limit at an angle of 0.52 radians. The peak value is reached at an angle of 0.57 radians. radians; at the peak Deviation If it is 0.008 amperes, then ampere.

[0054] In one embodiment, a steepness index is calculated based on the peak deviation magnitude, the deviation rise angle span, and the baseline; wherein, the steepness index It comprehensively reflects the magnitude and concentration of the deviation, and is defined as the ratio of the peak deviation magnitude to the product of the deviation rise angle span and the baseline: ; in The peak deviation is expressed in amperes. The deviation rises by an angle span in radians (radians themselves are dimensionless). The baseline is given in amperes. The physical meaning of is the normalized rate of change of current per unit angle. That is... A larger value indicates that the deviation changes more drastically within a shorter angular span, corresponding to stronger mechanical impact and higher broadband noise radiation intensity.

[0055] Following the example above, ampere, radians, if the baseline Ampère, then In contrast, another gearbox exhibits a deviation in a certain region. Both are 0.008 amperes, but It is 0.3 radians (the deviation is distributed within a range close to the width of a full tooth pitch). It remains at 0.22 amperes, then Both have the same peak amplitude, but the former is concentrated in a very narrow angle range (steep type), while the latter is distributed in a wide angle range (gentle type). The steepness index differs by about 6 times, and the former is far more likely to produce audible abnormal sounds.

[0056] In one embodiment, the steepness index is compared with an abnormal noise detection threshold to determine whether the gearbox has an abnormal noise defect. Specifically, the abnormal noise detection threshold... The determination was made through statistical analysis of calibration samples of the same model of gearbox. The calibration process involved taking qualified samples (e.g., no fewer than 30 pieces) confirmed to be free of abnormal noise by manual listening and unqualified samples (e.g., no fewer than 10 pieces) confirmed to have abnormal noise. The aforementioned steps were performed once for each sample, and the steepness index was calculated for each. The mean steepness index values ​​of all qualified samples were then calculated. and standard deviation The threshold for detecting abnormal sounds is set as follows: ; in The average steepness index of qualified products; The standard deviation of the kurtosis index for qualified products; This is the calibration coefficient, with a value ranging from 3 to 5. The specific value is determined based on the distribution interval between the maximum value of the steepness index for qualified products and the minimum value of the steepness index for unqualified products. Within this interval, the quantile corresponding to the point that clearly separates the two types of samples is selected. value.

[0057] As an example, if qualified products , The minimum defective product The value is 0.45, take hour, The boundary is clearly defined, falling between the upper limit for acceptable products and the lower limit for unacceptable products. Recalibration is required every time the gearbox model is changed. In some embodiments, during continuous production of the same model, a small number of sampled units can be used for verification at regular intervals (e.g., every 500 units). Is it still applicable?

[0058] The specific judgment logic is as follows: for the gearbox under inspection, if no deviation area is detected in the aforementioned steps, it is directly judged as qualified. If deviation areas exist, the steepness index is calculated for each deviation area. Any deviation region The value exceeds the threshold for detecting abnormal sounds. At that time, the gearbox was determined to have an abnormal noise defect and marked as unqualified.

[0059] Figure 5 This is a schematic diagram illustrating synchronization averaging and abnormal sound determination in one embodiment of this application. For example... Figure 5 As shown, the angular current signal is segmented according to a full revolution of the output gear. The average signal is obtained by summing and averaging the signals at the same angular position across multiple complete rotation cycles. Since fixed tooth surface defects recur at the same angular position in each rotation cycle, while random noise is uncorrelated across multiple revolutions, synchronous averaging can highlight local anomalies related to gear angles. The overall mean of the synchronous average signal is used as the baseline. When a local signal deviates from the baseline by more than the detection threshold, the corresponding continuous angular range is defined as the deviation region, and the peak amplitude of the deviation is extracted. And the angle span of deviation increase from the first time the detection threshold is exceeded to the peak value. Based on this, the steepness index is calculated. The more concentrated the deviation and the smaller the range of the rising angle, the greater the steepness index; when it exceeds the abnormal noise judgment threshold, the gearbox is judged to have an abnormal noise defect. Further, based on the relationship between the angular position corresponding to the deviation area and the single tooth angle range, the corresponding defective tooth number can be determined.

[0060] Specifically, after determining that it is an abnormal sound defect, the angular position corresponding to the deviation area is divided by the angular range corresponding to a single tooth (i.e., ,in The defect tooth number is calculated by rounding down the number of teeth on the inspected gear to obtain the defect tooth number, which is used for subsequent process traceability and location. For example, if the center angle of the deviation area is 1.2 radians, The single tooth angle range is radian, Rounded down to 4, the defect is located on the 4th tooth.

[0061] Optionally, there are multiple deviation areas and some Value close to (For example, in the state) to Gearboxes between [specific parameters] can be marked as defective items for manual re-inspection. In some embodiments, the inspection results, along with the defective tooth number and [other parameters], [are included]. The numerical values ​​are stored in the database and associated with the gearbox serial number to calculate the defect rate of a specific process or batch of materials.

[0062] It is understandable that the entire calculation process described above, including high-pass filtering, envelope detection, peak detection, linear interpolation, superposition averaging, deviation detection, kurtosis calculation, and threshold comparison, consists of basic numerical operations that can be completed in real time on embedded processors or industrial control computers on production lines. Taking a sampling rate of 20 kHz and a data acquisition time of 3 seconds as an example, the original data volume is 60,000 sampling points; after angular domain resampling and multi-cycle averaging, the data volume is reduced to a single cycle. 100 points (e.g., 100 points), subsequent deviation detection and The value calculation is performed only on these 100 points, resulting in a very small computational load. Taking a motor speed of 5000 rpm and a transmission ratio of 10 as an example, the output gear speed is 500 rpm, and one revolution takes 0.12 seconds. 25 revolutions would take about 3 seconds to complete the data acquisition. Including data processing time, the total time for single-item detection can be controlled within 4 seconds.

[0063] Therefore, this method extracts the rotation angle reference from the motor current signal itself, converts the current to the angular domain, and performs synchronous superposition and averaging to obtain gear tooth-by-tooth resolution current distribution information. Then, by analyzing the steepness of the deviation waveform, it quantifies whether the deviation will produce audible abnormal noise. Compared with the existing average current threshold method, this method can detect gearboxes with normal average current but local sharp tooth surface defects that cause abnormal noise. Compared with manual listening judgment, this method achieves objective quantitative detection, with consistent and traceable judgment standards, and is unaffected by operator status and environmental noise.

[0064] Figure 6 This is a schematic block diagram of a gearbox noise detection system for cleaning equipment based on current characteristic analysis, as described in this application. Figure 6 As shown, the system includes: The current acquisition and angle domain resampling module 601 is used to acquire the motor current signal when the gearbox of the cleaning equipment is running under no-load; extract the periodic component corresponding to the motor rotation from the motor current signal, determine the instantaneous rotation angle position of the motor based on the periodic component; and perform angle domain resampling on the motor current signal based on the instantaneous rotation angle position of the motor to obtain the angle domain current signal with the gear rotation angle as the independent variable. The synchronous averaging and deviation detection module 602 is used to segment the angular domain current signal according to the angle range of a complete rotation of the gear, and to superimpose and average the segmented signals of multiple complete rotation cycles at the same angular position to obtain a synchronous average signal; to calculate the overall mean of the synchronous average signal as a baseline, and to detect the angular region in the synchronous average signal that deviates from the baseline by more than the deviation detection threshold as the deviation region; The abnormal noise determination module 603 is used to extract the peak amplitude of the deviation and the range of the deviation rise angle for each deviation region, calculate the steepness index based on the peak amplitude of the deviation, the range of the deviation rise angle and the baseline, compare the steepness index with the abnormal noise determination threshold, and determine that the gearbox has an abnormal noise defect when the steepness index exceeds the abnormal noise determination threshold.

[0065] The implementation process of the functions and roles of each module in the above system is detailed in the implementation process of the corresponding steps in the above method, and will not be repeated here.

[0066] Figure 7 This is a schematic diagram of the structure of an electronic device according to one embodiment of this application. Figure 7 As shown, at the hardware level, the electronic device includes a processor 1102, an internal bus 1104, a network interface 1106, memory 1108, and non-volatile memory 1110, and may also include other hardware required for business operations. One or more embodiments of this specification can be implemented in software, for example, the processor 1102 reads the corresponding computer program from the non-volatile memory 1110 into the memory 1108 and then runs it. Of course, in addition to software implementation, one or more embodiments of this specification do not exclude other implementation methods, such as logic devices or a combination of hardware and software, etc. That is to say, the execution subject of the following processing flow is not limited to each logic module, but can also be hardware or logic devices.

[0067] Based on the same concept as the above-described method, this application also provides a machine-readable storage medium storing a plurality of computer instructions. When executed by a processor, these computer instructions can implement the method for detecting abnormal noise in a cleaning equipment gearbox based on current characteristic analysis disclosed in the above examples of this application. The machine-readable storage medium can be any electronic, magnetic, optical, or other physical storage device, and can contain or store information such as executable instructions, data, etc.

[0068] Based on the same application concept as the above method, this application embodiment also provides a computer program product, which includes a computer program that, when executed by a processor, implements the cleaning equipment gearbox abnormal noise detection method based on current characteristic analysis disclosed in the above examples of this application.

[0069] In the above embodiments, the descriptions of each embodiment have different focuses. Parts not described in detail in a certain embodiment can be referred to in the relevant descriptions of other embodiments. The above descriptions are merely preferred embodiments of this application and explanations of the technical principles used. Those skilled in the art should understand that the scope of the invention involved in this application is not limited to the technical solutions formed by specific combinations of the above technical features, but should also cover other technical solutions formed by arbitrary combinations of the above technical features or their equivalent features without departing from the inventive concept.

Claims

1. A method for detecting abnormal noise in the gearbox of cleaning equipment based on current characteristic analysis, characterized in that, include: Collect the motor current signal when the gearbox of the cleaning equipment is running under no-load. The periodic component corresponding to the motor rotation is extracted from the motor current signal, and the instantaneous rotation angle position of the motor is determined based on the periodic component; the motor current signal is resampled in the angular domain based on the instantaneous rotation angle position of the motor to obtain the angular domain current signal with the gear rotation angle as the independent variable; The angular current signal is segmented according to the angle range of a complete rotation of the gear. The segmented signals of multiple complete rotation cycles are superimposed and averaged at the same angular position to obtain a synchronous average signal. The overall mean of the synchronous average signal is calculated as a baseline. An angular region in the synchronous average signal that deviates from the baseline by more than the deviation detection threshold is detected as a deviation region. For each deviation region, the peak deviation amplitude and the deviation rise angle span are extracted. A steepness index is calculated based on the peak deviation amplitude, the deviation rise angle span, and the baseline. The steepness index is compared with an abnormal noise determination threshold. If the steepness index exceeds the abnormal noise determination threshold, the gearbox is determined to have an abnormal noise defect. The steepness index is the ratio of the product of the peak deviation amplitude, the deviation rise angle span, and the baseline. The deviation rise angle span is the angle difference between the angle position where the current deviation in the deviation region first exceeds the deviation detection threshold and the angle position where the current deviation reaches its peak value.

2. The method according to claim 1, characterized in that, The step of extracting the periodic component corresponding to the motor rotation from the motor current signal includes: performing high-pass filtering on the motor current signal to extract the commutation ripple component; performing envelope detection and peak detection on the commutation ripple component to obtain a commutation pulse time sequence; the interval between adjacent commutation pulses corresponds to a fixed angle increment of motor rotation, and the fixed angle increment is determined by the number of commutator segments of the motor.

3. The method according to claim 2, characterized in that, The step of resampling the motor current signal in the angular domain based on the instantaneous rotation angle position of the motor includes: performing linear interpolation on the rotation angle between two adjacent pulse moments in the commutation pulse time sequence, based on the assumption of constant rotation speed, to obtain the motor rotation angle value corresponding to each sampling moment; and performing interpolation resampling on the motor current signal at equal angular intervals to obtain the angular domain current signal.

4. The method according to claim 1, characterized in that, The specific steps for determining the deviation detection threshold are as follows: calculate the standard deviation of the point-by-point difference between the segmented signal of each rotation cycle and the synchronous average signal at the same angular position, attenuate the standard deviation according to the number of rotation cycles involved in the superposition and average, and then multiply it by a preset detection factor; the detection factor ranges from 3 to 5 and is determined according to the production line false alarm rate requirements during the first batch of verification.

5. The method according to claim 1, characterized in that, The abnormal noise judgment threshold is determined by statistical analysis of calibration samples of the same model of gearbox, including: taking qualified samples that have been manually confirmed to be free of abnormal noise, calculating the kurtosis index for each qualified sample according to the method described above, and calculating the mean and standard deviation of the kurtosis index for all qualified samples; adding the mean to the product of the standard deviation and the calibration coefficient as the abnormal noise judgment threshold, wherein the calibration coefficient ranges from 3 to 5 and is determined based on the distribution interval between the maximum value of the kurtosis index of qualified samples and the minimum value of the kurtosis index of unqualified samples.

6. The method according to claim 1, characterized in that, The number of angular sampling points in each gear rotation cycle of the angular current signal shall not be less than four times the number of teeth of the gear under test.

7. The method according to claim 2, characterized in that, When the angular current signal is segmented according to the angular range of one full revolution of the gear, the number of commutation pulses corresponding to one full revolution of the gear is the product of the gearbox transmission ratio and the number of motor commutator segments.

8. The method according to claim 1, characterized in that, When it is determined that the gearbox has an abnormal noise defect, the angle position corresponding to the deviation area is divided by the angle range corresponding to a single tooth and rounded to obtain the defective tooth number.

9. A noise detection system for a gearbox of a cleaning equipment based on current characteristic analysis, characterized in that, include: The current acquisition and corner domain resampling module is used to acquire the motor current signal when the gearbox of the cleaning equipment is running under no-load conditions. The periodic component corresponding to the motor rotation is extracted from the motor current signal, and the instantaneous rotation angle position of the motor is determined based on the periodic component; the motor current signal is resampled in the angular domain based on the instantaneous rotation angle position of the motor to obtain the angular domain current signal with the gear rotation angle as the independent variable; The synchronous averaging and deviation detection module is used to segment the angular domain current signal according to the angle range of a complete gear revolution, and to superimpose and average the segmented signals of multiple complete rotation cycles at the same angular position to obtain a synchronous average signal; to calculate the overall mean of the synchronous average signal as a baseline, and to detect the angular region in the synchronous average signal that deviates from the baseline by more than the deviation detection threshold as the deviation region; The abnormal noise detection module is used to extract the peak amplitude and the rise angle span of the deviation for each deviation region, and calculate the steepness index based on the peak amplitude, the rise angle span, and the baseline; compare the steepness index with the abnormal noise detection threshold, and determine that the gearbox has an abnormal noise defect when the steepness index exceeds the abnormal noise detection threshold; the steepness index is the ratio of the peak amplitude of the deviation to the product of the rise angle span and the baseline; the rise angle span is the angle difference between the angle position where the current deviation in the deviation region first exceeds the deviation detection threshold and the angle position where the current deviation reaches its peak value.

Images