Elevator noise and vibration test system

Abstract

The invention provides an elevator noise and vibration testing system which comprises a hardware detection module, a data transmission module and an intelligent analysis module, automation and intelligence of elevator noise and vibration detection are achieved, the detection efficiency is greatly improved, the differentiated requirements are met, and the system is suitable for popularization and application. Signal acquisition, time synchronization, analog-to-digital conversion and storage are automatically completed through the front-end data processing module, and original data processing does not need to be manually intervened; the intelligent analysis module automatically optimizes data through algorithms such as background noise calibration and wavelet threshold denoising, realizes automatic judgment of states such as start-stop, constant speed, acceleration and deceleration of the elevator by means of a multi-feature fusion recognition algorithm, and finally automatically generates a visual chart and a standardized detection report; compared with a traditional mode of relying on a sound level meter for manual measurement and manual data summarization, the efficiency is remarkably improved, the differential detection requirements of low frequency, low noise level and the like can be met, and the single limitation of a traditional detection means is broken through.

Description

Technical Field

[0001] This invention relates to an elevator noise and vibration testing system, belonging to the field of special equipment technology. Background Technology

[0002] Existing methods for detecting elevator operating noise are relatively simple, mainly relying on inspectors to use sound level meters to take multiple measurements and then summarize and calculate data from multiple points and groups, resulting in low inspection efficiency. Furthermore, the methods only target the three locations specified in the standard and lack targeted inspection solutions for the differentiated noise measurement needs raised by owners and for measuring low-frequency, low-noise levels.

[0003] Currently, the testing methods for elevator pulley bearings are very limited. They mainly rely on visual inspection and the noise emitted after bearing damage to determine whether the bearing meets the requirements. Furthermore, quantitative analysis is not possible, which has significant limitations. Therefore, an elevator noise and vibration testing system is proposed. Summary of the Invention

[0004] In view of this, the present invention provides an elevator noise and vibration testing system to solve or alleviate the technical problems existing in the prior art, and at least provides a beneficial alternative.

[0005] The technical solution of the present invention is implemented as follows: an elevator noise and vibration testing system includes a hardware detection module, a data transmission module and an intelligent analysis module, wherein the hardware detection module is communicatively connected to the intelligent analysis module through the data transmission module; The hardware detection module includes a noise sensor, a vibration sensor, a front-end data processing module, and a detection bracket. The noise sensor is used to collect noise signals from elevator operation and the environment, and the vibration sensor is used to collect vibration signals from elevator pulleys during operation. The front-end data processing module is electrically connected to the noise sensor and the vibration sensor respectively, and performs preliminary processing on the collected raw noise and vibration signals. The detection bracket is used to fix the noise sensor and the vibration sensor, enabling rapid positioning and distance calibration of the detection location. The intelligent analysis module is intelligent analysis software, including a data processing unit, a status analysis unit, a defect diagnosis unit, and a result output unit. The data processing unit is used to denoise and filter the transmitted data and simultaneously calibrate the background noise. The status analysis unit is used to intelligently identify the elevator's operating status by combining noise waveform and vibration waveform changes. The defect diagnosis unit is used to determine the type and location of internal defects in the elevator pulleys based on vibration signal characteristics. The result output unit is used to present the entire inspection process data in chart form and generate an inspection report.

[0006] More preferably, the front-end data processing module includes a data acquisition unit, a time synchronization unit, an analog-to-digital conversion unit, and a storage unit. The acquisition unit is connected to the noise sensor and the vibration sensor respectively, and is used to receive the original noise simulation signal and vibration simulation signal output by the sensors; The time synchronization unit integrates a real-time clock module, which is used to generate a timestamp accurate to the second when the acquisition unit starts, and associate and bind the timestamp with the original signal received by the acquisition unit. The analog-to-digital conversion unit is used to convert the analog signals output by the acoustic sensor and the vibration sensor into digital signals. The storage unit is used to store the digital signal output by the analog-to-digital conversion unit and the digital signal bound to the timestamp.

[0007] More preferably, the data processing unit of the intelligent analysis module adopts a background noise calibration algorithm. Specifically, in the static state where the elevator is not running, it collects 5-10 minutes of environmental noise data and calculates the average value as the background noise reference value. During the dynamic detection process of the elevator running, it compares the collected noise data with the background noise reference value in real time, and removes interference data with a difference of less than 3dB to achieve accurate calibration of the background noise.

[0008] More preferably, the data processing unit of the intelligent analysis module employs a wavelet threshold denoising algorithm, specifically including the following processing steps: Step 1: Select the db4 wavelet basis function to perform 3-5 level discrete wavelet decomposition on the received noise and vibration digital signals to obtain low-frequency approximate components and D1-D5 high-frequency detail components, where D1 corresponds to the highest frequency electromagnetic interference and D4-D5 correspond to the mid-to-high frequency mechanical shock interference. Step 2: Apply the threshold formula λ=σ The threshold λ is calculated, where N is the signal length and σ is the noise standard deviation, which is obtained by taking the median of the absolute values ​​of the wavelet coefficients of the high-frequency detail components of D1 and dividing it by 0.6745. Step 3: Apply soft thresholding to the high-frequency detail components D1-D5. Let any wavelet coefficient in the current high-frequency detail component be w. When the wavelet coefficient |w|>λ, After processing, the coefficient w' = sign (w) × (|w| - λ); when |w| ≤ λ, w' = 0; and stratified adaptive threshold adjustment is performed using λ for D1, 0.8λ for D2-D3, and 0.5λ for D4-D5; Step 4: Input the processed D1-D5 high-frequency detail components and the low-frequency approximate components obtained in Step 1 into the discrete wavelet inverse transform module to reconstruct the noise and vibration signals after preliminary denoising. Step 5: Set a preset amplitude threshold for the reconstructed signal in Step 4, monitor the signal data points in real time, and when the amplitude of a certain data point exceeds the amplitude threshold and the duration is ≤0.1 seconds, it is determined to be an abnormal pulse. Use linear interpolation to replace the pulse point with the normal data points adjacent to the abnormal pulse.

[0009] In a further preferred embodiment, the data processing unit of the intelligent analysis module adopts a segmented processing mode with 10-second units during the wavelet threshold denoising process to ensure real-time performance. It also achieves collaborative denoising of the two signals by analyzing the temporal correlation between the noise signal and the vibration signal, which is used to accurately remove abnormal pulse data generated by electromagnetic interference and mechanical impact.

[0010] More preferably, the state analysis unit of the intelligent analysis module adopts a multi-feature fusion recognition algorithm, specifically including feature parameter extraction and operating state determination; Step 1: Feature parameter extraction. For the denoised noise signal, a 0.5-second sliding window and a 50% overlap step size are used, calculated using the formula STE(m) = Calculate the short-time energy STE(m) for each window, where x(n) is the sequence of discrete data points of the noise signal, n is the index number of the discrete data points of the noise signal, used to traverse all discrete data points of noise within a single sliding window, m is the window number, and m×125+n represents the index of the nth data point in the mth window in the overall noise signal sequence; Secondly, a sliding window with the same speed as the short-time energy is used to extract the peak value X of the vibration signal within the window after denoising. peak Then follow formula X RMS = Calculate the root mean square value X RMS, y(p) is the sequence of discrete data points of the vibration signal, where p is the index number of each discrete data point. This index is used to iterate through all discrete data points within a single sliding window. Finally, the result is calculated using the formula CF = The peak factor CF is obtained; Step 2: Determine the operating status and calculate the short-time energy change rate. ; When ΔSTE>20% and CF>3, it is determined to be the elevator start-stop state; when ΔSTE≤5% and CF≤1.5, it is determined to be the elevator constant speed running state; when the characteristic parameter is between the above two, it is determined to be the elevator acceleration / deceleration transition state.

[0011] Further preferably, the state analysis unit also associates the operating parameters output by the elevator control system with timestamps. The operating parameters output by the elevator control system include the real-time position of the car and the running speed. The identification results are then verified a second time. If the elevator is identified as being in a uniform running state but the elevator speed is <0.2m / s, it is automatically corrected to a stationary state. If the elevator is identified as being in a start-stop state but the car position does not change, it is automatically marked as an abnormal state.

[0012] More preferably, the defect diagnosis unit of the intelligent analysis module adopts a vibration feature parameter extraction algorithm. Based on the denoised vibration signal with a sampling rate of 1000Hz, the operating cycle T=60 / r and the number of data points L=T×1000 in a single cycle are calculated by the elevator pulley speed r. The vibration data point y(i) in a single cycle is extracted, where i is the index of the vibration data point in a single cycle, and the value range is from 0 to L-1. Step 1: Calculate the peak acceleration Apeak using the formula Apeak=max{∣y(0)∣,∣y(1)∣,...,∣y(L-1)∣}, and take the average value of 3-5 consecutive cycles; Step 2: According to the formula Calculate root mean square acceleration Take the average value of 3-5 consecutive periods; Step 3: Perform a 1024-point Fast Fourier Transform on the vibration data points y(i) within a single period to obtain the power spectral density S(i) in the 10-1000Hz frequency band. ), Let k be the frequency value of the k-th frequency point, where k is the index of the frequency point within this frequency band, and then follow the formula. Q= Calculate the spectral kurtosis Q, where For frequency resolution, Δf = 1000Hz / 1024, and M is the total number of frequency points within the frequency band.

[0013] More preferably, the defect diagnosis unit further includes defect type determination, and the extracted peak acceleration Apeak and root mean square acceleration. Based on the three characteristic parameters of spectral kurtosis Q and the frequency domain characteristics in the 200-500Hz frequency band, the condition of the pulley bearing is determined in three categories. Under normal conditions, the pulley bearing's Apeak value is ≤0.5g. ≤0.15g, 3≤Q≤5, and no resonance peak with peak energy more than 3dB higher than the background in the 200-500Hz frequency band; Pulley bearing wear defects: Apeak > 0.8g >0.2g, 6<Q≤8, and no resonance peak with peak energy more than 3dB higher than the background in the 200-500Hz frequency band; Pulley bearing crack defects: Q > 8, and there is a resonance peak in the 200-500Hz frequency band that is more than 3dB higher than the background.

[0014] More preferably, the result output unit of the intelligent analysis module adopts a visual and templated output method, specifically, it automatically generates noise / vibration time-domain waveform diagrams, vibration power spectrum diagrams, and operational status statistical bar charts, and marks the operational status time periods and defect locations in the charts. At the same time, it calls standard templates to automatically generate inspection reports, filling in elevator basic information, signal processing parameters, status analysis, and defect diagnosis results.

[0015] The embodiments of the present invention have the following advantages due to the adoption of the above technical solutions: I. This invention automates and intelligentizes elevator noise and vibration detection, significantly improving detection efficiency and meeting diverse needs. The front-end data processing module automatically completes signal acquisition, time synchronization, analog-to-digital conversion, and storage, eliminating the need for manual intervention in raw data processing. The intelligent analysis module automatically optimizes data through background noise calibration and wavelet threshold denoising algorithms, and automatically determines elevator start-stop, constant speed, acceleration, and deceleration states using a multi-feature fusion recognition algorithm. Finally, it automatically generates visual charts and standardized detection reports. Compared to the traditional method of relying on manual measurement with sound level meters and manual data aggregation, this invention not only significantly improves efficiency but also covers diverse detection needs such as low-frequency and low-noise levels, breaking through the limitations of traditional detection methods.

[0016] Second, this invention achieves quantitative analysis and precise diagnosis of elevator pulley bearing defects, overcoming the limitations of traditional testing methods. Traditional pulley bearing testing relies solely on visual inspection or noise assessment after damage, failing to achieve quantitative and precise defect identification. This invention, however, establishes clear quantitative judgment criteria through a vibration characteristic parameter extraction algorithm combined with frequency domain characteristics in the 200-500Hz band. This allows for precise differentiation between three bearing states: normal, worn, and cracked, enabling quantitative diagnosis of defect types and states and providing accurate data support for elevator maintenance.

[0017] The above overview is for illustrative purposes only and is not intended to be limiting in any way. In addition to the illustrative aspects, embodiments, and features described above, further aspects, embodiments, and features of the invention will become readily apparent from the accompanying drawings and the following detailed description. Attached Figure Description

[0018] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0019] Figure 1 This is a flowchart illustrating the architecture of the present invention. Detailed Implementation

[0020] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. However, it should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of the invention. Furthermore, descriptions of well-known structures and technologies are omitted in the following description to avoid unnecessarily obscuring the concept of the invention.

[0021] In the description of this invention, it should be noted that when an element is referred to as being "fixed to" or "set on" another element, it can be directly on or indirectly on the other element. When an element is referred to as being "connected to" another element, it can be directly connected to or indirectly connected to the other element.

[0022] In the description of this invention, it should be noted that the terms "center," "length," "width," "thickness," "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "top," "bottom," "inner," and "outer," etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings, or the orientation or positional relationship commonly used when the product of this invention is in use. They are used only for the convenience of describing the invention and for simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on the invention. Furthermore, the terms "first," "second," and "third," etc., are used only to distinguish descriptions and should not be construed as indicating or implying relative importance. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of that feature. In the description of this invention, "a plurality of" means two or more, unless otherwise explicitly specified. "Several" means one or more, unless otherwise explicitly specified.

[0023] In the description of this invention, it should also be noted that, unless otherwise explicitly specified and limited, the terms "set," "install," "connect," and "link" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; and they can refer to the internal connection of two components. Those skilled in the art can understand the specific meaning of the above terms in this invention based on the specific circumstances.

[0024] The embodiments of the present invention will now be described in detail with reference to the accompanying drawings.

[0025] like Figure 1 As shown, this embodiment of the invention provides an elevator noise and vibration testing system, including a hardware detection module, a data transmission module, and an intelligent analysis module. The hardware detection module is communicatively connected to the intelligent analysis module through the data transmission module. In one embodiment, the hardware detection module includes a noise sensor, a vibration sensor, a front-end data processing module, and a detection bracket. The noise sensor is used to collect noise signals from elevator operation and the environment, and the vibration sensor is used to collect vibration signals from elevator pulley operation. The front-end data processing module is electrically connected to the noise sensor and the vibration sensor respectively, and performs preliminary processing on the collected raw noise and vibration signals. The detection bracket is used to fix the noise sensor and the vibration sensor, and can realize rapid positioning of the detection position and distance calibration. The intelligent analysis module is intelligent analysis software, including a data processing unit, a status analysis unit, a defect diagnosis unit, and a result output unit. The data processing unit is used to denoise and filter the transmitted data and synchronously calibrate the background noise. The status analysis unit is used to intelligently identify the elevator's operating status by combining noise waveform and vibration waveform changes. The defect diagnosis unit is used to determine the type and location of internal defects in the elevator pulleys based on vibration signal characteristics. The result output unit is used to present the entire inspection process data in chart form and generate an inspection report. In this embodiment, the hardware detection module uses a detection bracket to fix noise and vibration sensors, enabling rapid positioning and distance calibration of the detection location. The noise sensor collects elevator operation and environmental noise signals, while the vibration sensor collects elevator pulley vibration signals. The front-end data processing module performs preliminary processing on the two raw signals. The processed signals are then sent to the intelligent analysis module via a data transmission module. In this module, the data processing unit is responsible for noise reduction and background noise calibration, the status analysis unit combines noise and vibration waveforms to identify the elevator's operating status, the defect diagnosis unit determines internal defects in the pulley, and the result output unit presents the data in charts and generates a detection report. This architecture, through standardized sensor fixing, coordinated signal acquisition, and intelligent analysis, reduces manual operation steps, improves detection efficiency, and covers both elevator operating status and core component detection, solving the problem of limited detection methods in traditional approaches.

[0026] In one embodiment, the front-end data processing module includes an acquisition unit, a time synchronization unit, an analog-to-digital conversion unit, and a storage unit. The acquisition unit is connected to the noise sensor and vibration sensor respectively, and is used to receive the original noise simulation signal and vibration simulation signal output by the sensors; The time synchronization unit integrates a real-time clock module, which is used to generate a timestamp accurate to the second when the acquisition unit starts up, and associates and binds the timestamp with the raw signal received by the acquisition unit. The analog-to-digital converter is used to convert the analog signals output by the acoustic sensor and vibration sensor into digital signals. The storage unit is used to store the digital signals output by the analog-to-digital converter and the timestamp-bound digital signals; In this embodiment, the acquisition unit of the front-end data processing module first receives the raw analog signals output by the noise sensor and vibration sensor; then, the time synchronization unit integrates a real-time clock module to generate a timestamp accurate to the second and associates it with the raw signal to ensure the time consistency of the two signals; the analog-to-digital conversion unit converts the analog signal into a digital signal to meet the digital processing requirements of the intelligent analysis module; finally, the storage unit stores the timestamp-bound digital signal locally. The timestamp binding provides a benchmark for subsequent signal collaborative analysis, the analog-to-digital conversion achieves signal format adaptation, and local storage ensures the integrity of the raw data, effectively improving the accuracy and security of subsequent data processing.

[0027] In one embodiment, the data processing unit of the intelligent analysis module employs a background noise calibration algorithm. Specifically, in the static state where the elevator is not running, it collects 5-10 minutes of environmental noise data and calculates the average value as the background noise baseline. During the dynamic detection process of the elevator operation, it compares the collected noise data with the background noise baseline in real time, eliminating interference data with a difference of less than 3dB to achieve accurate background noise calibration. The data processing unit of the intelligent analysis module also employs a wavelet threshold denoising algorithm, specifically including the following processing steps. Step 1: Select the db4 wavelet basis function to perform 3-5 level discrete wavelet decomposition on the received noise and vibration digital signals to obtain low-frequency approximate components and D1-D5 high-frequency detail components, where D1 corresponds to the highest frequency electromagnetic interference and D4-D5 correspond to the mid-to-high frequency mechanical shock interference. Step 2: Apply the threshold formula λ=σ The threshold λ is calculated, where N is the signal length and σ is the noise standard deviation, which is obtained by taking the median of the absolute values ​​of the wavelet coefficients of the high-frequency detail components of D1 and dividing it by 0.6745. Step 3: Apply soft thresholding to the high-frequency detail components D1-D5. Let any wavelet coefficient in the current high-frequency detail component be w. When the wavelet coefficient |w|>λ, After processing, the coefficient w' = sign (w) × (|w| - λ); when |w| ≤ λ, w' = 0; and stratified adaptive threshold adjustment is performed using λ for D1, 0.8λ for D2-D3, and 0.5λ for D4-D5; Step 4: Input the processed D1-D5 high-frequency detail components and the low-frequency approximate components obtained in Step 1 into the discrete wavelet inverse transform module to reconstruct the noise and vibration signals after preliminary denoising. Step 5: A preset amplitude threshold is set for the reconstructed signal from Step 4. Signal data points are monitored in real time. When the amplitude of a data point exceeds the threshold and its duration is ≤0.1 seconds, it is determined to be an abnormal pulse. Linear interpolation is used to replace the abnormal pulse with adjacent normal data points. The data processing unit of the intelligent analysis module uses a segmented processing mode with 10-second units during wavelet threshold denoising to ensure real-time performance. It also achieves collaborative denoising of the two signals by analyzing the temporal correlation between the noise signal and the vibration signal, accurately eliminating abnormal pulse data generated by electromagnetic interference and mechanical impact. A duration of ≤0.1 seconds is a critical limitation, as these abnormal pulses often originate from transient electromagnetic interference during elevator operation and brief mechanical impacts between pulleys and guide rails, typically lasting only 0.01-0.1 seconds. This effectively distinguishes them from normal signal fluctuations lasting longer than 0.5 seconds during elevator acceleration and deceleration, preventing the misjudgment of useful signals as interference. In this embodiment, the data processing unit first achieves interference filtering through a background noise calibration algorithm: 5-10 minutes of ambient noise are collected when the elevator is not running, and the average value is calculated as a baseline. During elevator operation, the collected data is compared with the baseline in real time, and interference data with a difference of less than 3dB is eliminated. Then, a wavelet threshold denoising algorithm is used to optimize the signal: a db4 wavelet basis is used to perform 3-5 levels of discrete wavelet decomposition on the digital signal to obtain high-frequency detail components and low-frequency approximate components corresponding to different interferences. After calculating the threshold λ according to the formula, layered soft thresholding is applied to the D1-D5 high-frequency components. The processed components are reconstructed through inverse transform, and then abnormal pulses are determined and replaced using amplitude thresholding. Simultaneously, 10-second segmented processing and signal time-domain correlation analysis are used to ensure real-time performance and denoising accuracy. This algorithm combination specifically filters environmental interference, electromagnetic interference, and mechanical shock interference, significantly improving signal purity and providing a high-quality data foundation for subsequent state recognition and defect diagnosis.

[0028] In one embodiment, the state analysis unit of the intelligent analysis module adopts a multi-feature fusion recognition algorithm, specifically including feature parameter extraction and running status determination; Step 1: Feature parameter extraction. For the denoised noise signal, a 0.5-second sliding window and a 50% overlap step size are used, calculated using the formula STE(m) = Calculate the short-time energy STE(m) for each window, where x(n) is the sequence of discrete data points of the noise signal, n is the index number of the discrete data points of the noise signal, used to traverse all discrete data points of noise within a single sliding window, m is the window number, and m×125+n represents the index of the nth data point in the mth window in the overall noise signal sequence; Secondly, a sliding window with the same speed as the short-time energy is used to extract the peak value X of the vibration signal within the window after denoising. peak Then follow formula X RMS = Calculate the root mean square value X RMS, y(p) is the sequence of discrete data points of the vibration signal, where p is the index number of each discrete data point. This index is used to iterate through all discrete data points within a single sliding window. Finally, the result is calculated using the formula CF = The peak factor CF is obtained; Step 2: Determine the operating status and calculate the short-time energy change rate. ; When ΔSTE > 20% and CF > 3, the elevator is determined to be in a start-stop state; when ΔSTE ≤ 5% and CF ≤ 1.5, the elevator is determined to be in a constant-speed running state; when the characteristic parameters are between the above two, the elevator is determined to be in an acceleration / deceleration transition state. The state analysis unit also associates the operating parameters output by the elevator control system with timestamps. The operating parameters output by the elevator control system include the real-time position of the car and the running speed. The identification results are then verified a second time. If the elevator is identified as being in a constant-speed running state but the elevator speed is < 0.2 m / s, it is automatically corrected to a stationary state; if the elevator is identified as being in a start-stop state but the car position does not change, it is automatically marked as an abnormal state. In this embodiment, the multi-feature fusion recognition algorithm of the state analysis unit first extracts feature parameters: for the denoised noise signal, a 0.5-second sliding window and a 50% overlap step size are used to calculate the short-time energy STE (m); for the vibration signal, the same window is used to extract the peak value Xpeak and calculate the root mean square value. The system first calculates the peak factor CF; then it determines the operating status by calculating the short-time energy change rate ΔSTE (m). The system determines the start-stop state based on "ΔSTE > 20% and CF > 3", the constant speed state based on "ΔSTE ≤ 5% and CF ≤ 1.5", and the acceleration / deceleration transition state based on the state between these two. Finally, it performs a secondary verification by linking the car position and operating speed parameters of the elevator control system with timestamps, correcting misjudgments and marking anomalies. Multi-feature fusion avoids the limitations of a single parameter, and the secondary verification of elevator operating parameters further reduces the misjudgment rate, achieving accurate identification of the elevator's operating status.

[0029] In one embodiment, the defect diagnosis unit of the intelligent analysis module adopts a vibration feature parameter extraction algorithm. Based on the denoised vibration signal with a sampling rate of 1000Hz, it first calculates the running cycle T=60 / r and the number of data points L=T×1000 in a single cycle by the elevator pulley speed r, and then extracts the vibration data point y(i) in a single cycle, where i is the index of the vibration data point in a single cycle, and the value range is 0 to L-1. Step 1: Calculate the peak acceleration Apeak using the formula Apeak=max{∣y(0)∣,∣y(1)∣,...,∣y(L-1)∣}, and take the average value of 3-5 consecutive cycles; Step 2: According to the formula Calculate root mean square acceleration Take the average value of 3-5 consecutive periods; Step 3: Perform a 1024-point Fast Fourier Transform on the vibration data points y(i) within a single period to obtain the power spectral density S(i) in the 10-1000Hz frequency band. ), Let k be the frequency value of the k-th frequency point, where k is the index of the frequency point within this frequency band, and then follow the formula. Q= Calculate the spectral kurtosis Q, where For frequency resolution, Δf = 1000Hz / 1024, and M is the total number of frequency points within the frequency band; the defect diagnosis unit also includes defect type determination, and the extracted peak acceleration Apeak and root mean square acceleration. Based on the three characteristic parameters of spectral kurtosis Q and the frequency domain characteristics in the 200-500Hz frequency band, the condition of the pulley bearing is determined in three categories. Under normal conditions, the pulley bearing's Apeak value is ≤0.5g. ≤0.15g, 3≤Q≤5, and no resonance peak with peak energy more than 3dB higher than the background in the 200-500Hz frequency band; Pulley bearing wear defects: Apeak > 0.8g >0.2g, 6<Q≤8, and no resonance peak with peak energy more than 3dB higher than the background in the 200-500Hz frequency band; Pulley bearing crack defects: Q>8, and there is a resonance peak in the 200-500Hz frequency band that is more than 3dB higher than the background. In this embodiment, the defect diagnosis unit, based on the denoised vibration signal with a sampling rate of 1000Hz, first calculates the operating period T=60 / r and the number of data points L=T×1000 using the pulley rotation speed r, and extracts vibration data for a single period; then it extracts feature parameters: calculates the peak acceleration Apeak and root mean square acceleration for 3-5 consecutive periods. The vibration data is subjected to a 1024-point fast Fourier transform to obtain the power spectral density of 10-1000Hz and the spectral kurtosis Q is calculated. Finally, the defect type is determined by combining the characteristic parameters with the frequency domain features of 200-500Hz. This method realizes the quantitative and accurate diagnosis of pulley bearing defects by combining quantitative characteristic parameters with frequency domain features, and solves the limitations of traditional appearance inspection and subjective judgment.

[0030] In one embodiment, the result output unit of the intelligent analysis module adopts a visual and templated output method, specifically, it automatically generates noise / vibration time-domain waveform diagrams, vibration power spectrum diagrams, and operation status statistical bar charts, and marks the operation status time periods and defect locations in the charts. At the same time, it calls standard templates to automatically generate inspection reports, and fills in elevator basic information, signal processing parameters, status analysis, and defect diagnosis results. Visual charts intuitively present the test data and key conclusions, facilitating quick problem identification. Templated reports reduce manual work, improve report standardization and traceability, and provide clear and accurate decision-making basis for elevator maintenance.

[0031] When this invention is in operation: First, the noise sensor and vibration sensor are fixed by the detection bracket to complete the rapid positioning and distance calibration of the detection position; then the hardware detection module is activated, the noise sensor collects the noise simulation signal of the elevator operation and the environment, and the vibration sensor collects the vibration simulation signal of the elevator pulley operation, and both are transmitted to the front-end data processing module simultaneously. The front-end data processing module's acquisition unit receives two analog signals, the time synchronization unit generates a timestamp accurate to the second and binds it to the signal, the analog-to-digital conversion unit converts the analog signal into a digital signal, and finally the storage unit stores the digital signal bound to the timestamp locally, while the data transmission module sends the digital signal to the intelligent analysis module. The data processing unit of the intelligent analysis module first filters environmental noise through a background noise calibration algorithm, then uses a wavelet threshold denoising algorithm to remove electromagnetic interference and mechanical shock noise, while ensuring real-time performance through 10-second segmented processing, and achieving collaborative denoising through signal time domain correlation. Based on the denoised signal, the state analysis unit extracts the short-time energy of the noise signal, the peak value and peak factor of the vibration signal, calculates the short-time energy change rate, and determines the elevator start-stop, constant speed or acceleration / deceleration state by combining the threshold. Then, it performs secondary verification by associating the car position and speed parameters of the elevator control system with the timestamp, corrects misjudgments and marks anomalies. The defect diagnosis unit uses a denoised vibration signal with a sampling rate of 1000Hz to calculate the operating cycle based on the pulley rotation speed and extracts single-cycle data. It extracts peak acceleration, root mean square acceleration and spectral kurtosis, and combines the frequency domain characteristics of the 200-500Hz band to determine whether the pulley bearing is in a normal, worn or cracked state.

[0032] The output unit automatically generates noise / vibration time-domain waveforms, vibration power spectrum diagrams, and operational status statistical bar charts, marking the operational status time periods and defect locations. Simultaneously, it calls standard templates to fill in elevator basic information, signal processing parameters, status analysis, and defect diagnosis results, generating a standardized inspection report.

[0033] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any person skilled in the art can easily conceive of various variations or substitutions within the technical scope disclosed in the present invention, and these should all be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. An elevator noise and vibration testing system, characterized in that: It includes a hardware detection module, a data transmission module, and an intelligent analysis module. The hardware detection module is communicatively connected to the intelligent analysis module through the data transmission module. The hardware detection module includes a noise sensor, a vibration sensor, a front-end data processing module, and a detection bracket. The noise sensor is used to collect noise signals from elevator operation and the environment, and the vibration sensor is used to collect vibration signals from elevator pulleys during operation. The front-end data processing module is electrically connected to the noise sensor and the vibration sensor respectively, and performs preliminary processing on the collected raw noise and vibration signals. The detection bracket is used to fix the noise sensor and the vibration sensor, enabling rapid positioning and distance calibration of the detection location. The intelligent analysis module is intelligent analysis software, including a data processing unit, a status analysis unit, a defect diagnosis unit, and a result output unit. The data processing unit is used to denoise and filter the transmitted data and simultaneously calibrate the background noise. The status analysis unit is used to intelligently identify the elevator's operating status by combining noise waveform and vibration waveform changes. The defect diagnosis unit is used to determine the type and location of internal defects in the elevator pulleys based on vibration signal characteristics. The result output unit is used to present the entire inspection process data in chart form and generate an inspection report.

2. The elevator noise and vibration testing system according to claim 1, characterized in that: The front-end data processing module includes a data acquisition unit, a time synchronization unit, an analog-to-digital conversion unit, and a storage unit. The acquisition unit is connected to the noise sensor and the vibration sensor respectively, and is used to receive the original noise simulation signal and vibration simulation signal output by the sensors; The time synchronization unit integrates a real-time clock module, which is used to generate a timestamp accurate to the second when the acquisition unit starts, and associate and bind the timestamp with the original signal received by the acquisition unit. The analog-to-digital conversion unit is used to convert the analog signals output by the acoustic sensor and the vibration sensor into digital signals. The storage unit is used to store the digital signal output by the analog-to-digital conversion unit and the digital signal bound to the timestamp.

3. The elevator noise and vibration testing system according to claim 1, characterized in that: The data processing unit of the intelligent analysis module adopts a background noise calibration algorithm. Specifically, in the static state where the elevator is not running, it collects 5-10 minutes of environmental noise data and calculates the average value as the background noise benchmark. During the dynamic detection process of the elevator running, it compares the collected noise data with the background noise benchmark in real time, and removes interference data with a difference of less than 3dB to achieve accurate calibration of background noise.

4. The elevator noise and vibration testing system according to claim 3, characterized in that: The data processing unit of the intelligent analysis module employs a wavelet threshold denoising algorithm, specifically including the following processing steps. Step 1: Select the db4 wavelet basis function to perform 3-5 level discrete wavelet decomposition on the received noise and vibration digital signals to obtain low-frequency approximate components and D1-D5 high-frequency detail components, where D1 corresponds to the highest frequency electromagnetic interference and D4-D5 correspond to the mid-to-high frequency mechanical shock interference. Step 2: Apply the threshold formula λ=σ The threshold λ is calculated, where N is the signal length and σ is the noise standard deviation, which is obtained by taking the median of the absolute values ​​of the wavelet coefficients of the high-frequency detail components of D1 and dividing it by 0.6745. Step 3: Apply soft thresholding to the high-frequency detail components D1-D5. Let any wavelet coefficient in the current high-frequency detail component be w. When the wavelet coefficient |w|>λ, After processing, the coefficient w' = sign (w) × (|w| - λ); when |w| ≤ λ, w' = 0; and stratified adaptive threshold adjustment is performed by using λ for D1, 0.8λ for D2-D3, and 0.5λ for D4-D5; Step 4: Input the processed D1-D5 high-frequency detail components and the low-frequency approximate components obtained in Step 1 into the discrete wavelet inverse transform module to reconstruct the noise and vibration signals after preliminary denoising. Step 5: Set a preset amplitude threshold for the reconstructed signal in Step 4, monitor the signal data points in real time, and when the amplitude of a certain data point exceeds the amplitude threshold and the duration is ≤0.1 seconds, it is determined to be an abnormal pulse. Use linear interpolation to replace the pulse point with the normal data points adjacent to the abnormal pulse.

5. The elevator noise and vibration testing system according to claim 4, characterized in that: The data processing unit of the intelligent analysis module adopts a segmented processing mode with 10-second units to ensure real-time performance during wavelet threshold denoising. It also achieves collaborative denoising of the two signals by analyzing the temporal correlation between the noise signal and the vibration signal, which is used to accurately remove abnormal pulse data generated by electromagnetic interference and mechanical impact.

6. The elevator noise and vibration testing system according to claim 1, characterized in that: The state analysis unit of the intelligent analysis module adopts a multi-feature fusion recognition algorithm, specifically including feature parameter extraction and operation status determination; Step 1: Feature parameter extraction. For the denoised noise signal, a 0.5-second sliding window and a 50% overlap step size are used, calculated using the formula STE(m) = Calculate the short-time energy STE(m) for each window, where x(n) is the sequence of discrete data points of the noise signal, n is the index number of the discrete data points of the noise signal, used to traverse all discrete data points of noise within a single sliding window, m is the window number, and m×125+n represents the index of the nth data point in the mth window in the overall noise signal sequence; Secondly, a sliding window with the same speed as the short-time energy is used to extract the peak value X of the vibration signal within the window after denoising. peak , Then press formula X RMS = Calculate the root mean square value X RMS, y(p) is the sequence of discrete data points of the vibration signal, where p is the index number of each discrete data point. This index is used to iterate through all discrete data points within a single sliding window. Finally, the result is calculated using the formula CF = The peak factor CF is obtained; Step 2: Determine the operating status and calculate the short-time energy change rate. ； When ΔSTE>20% and CF>3, the elevator is determined to be in start-stop state; when ΔSTE≤5% and CF≤1.5, the elevator is determined to be in constant speed running state. When the characteristic parameter is between the two mentioned above, it is determined to be an elevator acceleration / deceleration transition state.

7. The elevator noise and vibration testing system according to claim 6, characterized in that: The state analysis unit also associates the operating parameters output by the elevator control system with timestamps. The operating parameters output by the elevator control system include the real-time position of the car and the running speed. The identification results are then verified. If the elevator is identified as running at a constant speed but the elevator speed is <0.2m / s, it is automatically corrected to a stationary state. If the elevator is identified as starting or stopping but the car position does not change, it is automatically marked as an abnormal state.

8. The elevator noise and vibration testing system according to claim 1, characterized in that: The defect diagnosis unit of the intelligent analysis module adopts a vibration feature parameter extraction algorithm. Based on the denoised vibration signal with a sampling rate of 1000Hz, it first calculates the running cycle T=60 / r and the number of data points L=T×1000 in a single cycle by the elevator pulley speed r, and then extracts the vibration data point y(i) in a single cycle, where i is the index of the vibration data point in a single cycle, and the value range is 0 to L-1. Step 1: Calculate the peak acceleration Apeak using the formula Apeak=max{∣y(0)∣,∣y(1)∣,...,∣y(L-1)∣}, and take the average value of 3-5 consecutive cycles; Step 2: According to the formula Calculate root mean square acceleration Take the average value of 3-5 consecutive periods; Step 3: Perform a 1024-point Fast Fourier Transform on the vibration data points y(i)j within a single period to obtain the power spectral density S(i) in the 10-1000Hz frequency band. ), Let Q be the frequency value of the k-th frequency point, where k is the index of the frequency point within this frequency band, and then calculate it according to the formula Q= Calculate the spectral kurtosis Q, where Δf is the frequency resolution, Δf = 1000Hz / 1024, and M is the total number of frequency points within the frequency band.

9. The elevator noise and vibration testing system according to claim 8, characterized in that: The defect diagnosis unit also includes defect type determination, and the extracted peak acceleration Apeak and root mean square acceleration. Based on the three characteristic parameters of spectral kurtosis Q and the frequency domain characteristics in the 200-500Hz frequency band, the condition of the pulley bearing is determined in three categories. Under normal conditions, the pulley bearing's Apeak value is ≤0.5g. ≤0.15g, 3≤Q≤5, and no resonance peak with peak energy more than 3dB higher than the background in the 200-500Hz frequency band; Pulley bearing wear defects: Apeak > 0.8g >0.2g, 6<Q≤8, and no resonance peak with peak energy more than 3dB higher than the background in the 200-500Hz frequency band; Pulley bearing crack defects: Q > 8, and there is a resonance peak in the 200-500Hz frequency band that is more than 3dB higher than the background.

10. The elevator noise and vibration testing system according to claim 1, characterized in that: The intelligent analysis module's output unit adopts a visual and template-based output method. Specifically, it automatically generates noise / vibration time-domain waveform diagrams, vibration power spectrum diagrams, and operational status statistical bar charts, and marks the operational status time periods and defect locations in the charts. At the same time, it calls standard templates to automatically generate inspection reports, filling in elevator basic information, signal processing parameters, status analysis, and defect diagnosis results.

Images