A method and system for identifying abnormal vibration of a trolley mechanism of a port rail-mounted crane

Abstract

The application discloses a port rail-mounted crane trolley mechanism vibration anomaly identification method and system, and relates to the technical field of state monitoring and fault diagnosis of port hoisting equipment. The method comprises the following steps: collecting vibration signals of multiple measuring points; dividing the signals corresponding to the measuring points into multiple signal segments according to time based on the vibration signals of the measuring points or state characteristic signals converted from the vibration signals, and calculating running activity indexes of the signal segments; judging whether the measuring points are in an active state according to the running activity indexes of the signal segments; counting the number of measuring points in the active state, and determining that an effective running condition is reached when the number of measuring points in the active state reaches a preset number; under the effective running condition, performing envelope demodulation on the vibration signals of at least one measuring point and spectrum amplitude statistics in a target frequency band to obtain a robust dispersion characteristic value representing the dispersion degree of envelope spectrum amplitude value distribution; and performing vibration anomaly identification or state evaluation on the trolley mechanism according to the robust dispersion characteristic value.

Description

Technical Field

[0001] This invention relates to the field of port crane equipment condition monitoring and fault diagnosis technology, specifically to a method and system for identifying abnormal vibrations in the trolley mechanism of a port rail crane. Background Technology

[0002] Rail-mounted cranes are key large-scale equipment in container yards and port loading and unloading operations. They typically run along fixed tracks to complete container lifting, stacking, transfer, and loading / unloading connections. The trolley mechanism, as one of the important motion mechanisms of the rail-mounted crane, reciprocates along the main beam during operation. Its guide wheels, drive wheels, bearing housings, and related transmission components are subjected to long-term impacts from starting and stopping, load changes, track contact disturbances, and structural vibration transmission.

[0003] Due to the continuous nature of port operations, their dense work cycles, and high downtime costs, failure to promptly identify early anomalies in the trolley mechanism can lead to accelerated component wear, unstable trolley operation, and decreased positioning accuracy. In severe cases, this can affect the operational safety and loading / unloading efficiency of the rail-mounted crane. Therefore, identifying vibration anomalies in the trolley mechanism of port rail-mounted cranes is of significant engineering importance.

[0004] In existing technologies, common vibration monitoring methods typically extract features from the acquired vibration signals, such as root mean square value, peak value, kurtosis, vibration intensity, and spectral peak value, and then determine the state based on these features. However, in the scenario of port rail-mounted cranes, the trolley mechanism is not always in an effective operating state. When the trolley is stationary, in a low-activity state, experiences short-term start-stop, or is not in operation, the vibration signal may mainly contain background vibration, structurally transmitted vibration, or environmental disturbances. If anomaly identification is performed directly without distinguishing between effective operating conditions, invalid samples are easily included in the judgment, reducing the stability of the identification results.

[0005] Furthermore, existing methods have at least the following shortcomings: First, they lack effective operating condition screening, leading to the inclusion of static, low-activity, and start-stop transition data in anomaly identification; second, they rely heavily on single measurement points or single amplitude features, making it difficult to reflect the coordinated operation status of multiple parts of the trolley mechanism; third, some spectral or envelope spectral features are easily affected by accidental shocks and extreme spectral lines, resulting in insufficient feature stability. Therefore, a vibration anomaly identification method combining "multi-measurement condition screening + robust statistics of envelope spectrum under effective operating conditions" is needed to improve the reliability of vibration anomaly identification or condition assessment of the trolley mechanism of port rail cranes. Summary of the Invention

[0006] This invention provides a method and system for identifying vibration anomalies in the trolley mechanism of a port rail crane, aiming to solve the problems in the prior art where the identification of vibration anomalies in the trolley mechanism of a port rail crane is easily affected by invalid operating condition data, has insufficient characterization of early impact modulation anomalies, and lacks robustness of spectral features.

[0007] The technical solution of the present invention is as follows: A method for identifying abnormal vibrations in the trolley mechanism of a port rail-mounted crane includes the following steps: Step 1: Collect vibration signals from multiple measuring points on the trolley mechanism of the port rail crane. Measuring points can be set at locations such as the guide wheels, drive wheels, bearing seats, or other positions that reflect the vibration state of the trolley during operation. Assume the number of measuring points is... M , No. m The discrete vibration signals collected at each measuring point are ,in , n This is the sampling point number.

[0008] Step 2: Based on the state characterization signals obtained from the vibration signals at each measuring point, divide the signals corresponding to each measuring point into multiple signal segments according to time. Using the velocity signal as the state characterization signal, first perform detrending processing on the vibration signal to obtain the detrended vibration signal. The de-stressed vibration signal is then integrated to obtain the initial velocity signal. Then, the initial velocity signal is subjected to frequency band filtering to obtain a velocity signal used for operating condition screening. The above process can be represented as: in, This indicates detrending processing. n Indicates the discrete-time sampling point number; This is a discrete-time index for the integration and summation process. , Indicates the sampling time interval. Indicates the frequency band range as Bandpass filtering is applied. The frequency range of the bandpass filtering is determined based on the main energy frequency band of the trolley mechanism's vibration, for example, it can be from 10 Hz to 1000 Hz. The state characterization signal is divided into... The signal segment, the first m The first measuring point k The velocity signal segment is denoted as ,in .

[0009] Step 3: Calculate the operational activity index for each signal segment. The operational activity index uses the root mean square velocity value. m The first measuring point k The operational activity index of a signal segment It can be represented as: in, Indicates the first k The number of sampling points contained in a signal segment.

[0010] Step 4: Determine whether each measurement point is in an active state based on the operational activity index of each signal segment. Compare the operational activity index of multiple signal segments corresponding to the same measurement point with the preset segment threshold. Comparison to obtain fragment activity markers : Statistics m Number of active signal segments at each measurement point : When the number of signal segments Reach the preset number of segments When this occurs, the measuring point is determined to be in an active state. Measuring point active marker. It can be represented as: Step 5: Count the number of active measuring points and determine whether the trolley mechanism is in effective operating condition. Number of active measuring points C It can be represented as: Number of active measurement points C Compared with the preset threshold number of measurement points Comparison. When When the trolley mechanism is in a valid operating condition, it is determined that the trolley mechanism is in a valid operating condition; when If the trolley mechanism is in an invalid operating condition, vibration anomaly identification will not be performed on the signal under this condition. Valid operating condition flag. G It can be represented as: Step 6: Under the stated effective operating conditions, perform envelope demodulation on the vibration signals of all measuring points. Preprocess the vibration signals of the measuring points to be analyzed to obtain the demodulated signals. The preprocessing may include at least one of detrending, filtering, mean removal, normalization, and sampling rate unification. An analytical signal is constructed from the signal to be demodulated, and the envelope signal is obtained based on the modulus of the analytical signal. The analytical signal construction can be achieved through Hilbert transform. and envelope signal They can be represented as: in, Represents the Hilbert transform. j It represents the imaginary unit.

[0011] Step 7: Perform spectral amplitude statistics on the envelope signal within the target frequency band to obtain robust discrete characteristic values ​​that characterize the dispersion of the envelope spectral amplitude distribution. For the envelope signal... Perform a frequency domain transformation to obtain the envelope spectrum amplitude. : Extract the target frequency band from the envelope spectrum. amplitude sequence within : in, Indicates the first q The frequency corresponding to each frequency point. The target frequency band can be determined based on at least one of the following: the operating frequency of the trolley mechanism, the characteristic frequency of the transmission components, the structural vibration frequency, or the abnormal impact modulation frequency, for example, it can be from 5 Hz to 1500 Hz.

[0012] Calculate the amplitude sequence the median of : Calculate the absolute deviation of each amplitude in the amplitude sequence relative to the median to obtain the absolute deviation sequence. : Calculate the median of the absolute deviation sequence d : When the robust discrete eigenvalue is the standardized median absolute deviation eigenvalue, it is calculated according to the following formula: in, F Represents robust discrete eigenvalues; This represents the standardization coefficient.

[0013] Step 8: Identify vibration anomalies or evaluate the condition of the trolley mechanism based on the robust discrete eigenvalues. (The robust discrete eigenvalues ​​are then used for...) F Compared with the preset abnormal threshold Comparison is performed to obtain anomaly identification results. R : in, R =1 indicates that there is abnormal vibration in the trolley mechanism. R =0 indicates that the trolley mechanism is in a normal vibration state or has not met the abnormal judgment conditions. Preset abnormal threshold. It can be determined based on historical normal operation data, field experience, statistical analysis results, or equipment maintenance requirements.

[0014] The present invention also provides a vibration anomaly identification system for the trolley mechanism of a port rail crane, used to implement the above method, specifically including: The vibration acquisition module is used to collect vibration signals from multiple measuring points on the trolley mechanism of a port rail crane. The working condition screening module is used to divide the signal corresponding to each measuring point into multiple signal segments according to time based on the vibration signal of each measuring point or the state characterization signal converted from the vibration signal, calculate the operating activity index of each signal segment, determine whether each measuring point is in an active state based on the operating activity index, and determine whether the trolley mechanism is in an effective operating condition based on the number of measuring points in an active state. The feature extraction module is used to perform envelope demodulation and spectral amplitude statistics within the target frequency band of the vibration signal of the measurement point to be analyzed under effective operating conditions, so as to obtain robust discrete feature values ​​that characterize the degree of dispersion of the envelope spectrum amplitude distribution. The anomaly identification module is used to identify vibration anomalies or evaluate the condition of the trolley mechanism based on robust discrete eigenvalues.

[0015] Compared with the prior art, the beneficial effects of the present invention are: Before identifying vibration anomalies, this invention first determines whether the trolley mechanism is in an effective operating condition based on the operational activity index of signal segments corresponding to multiple measuring points. This can reduce the entry of static, low-activity, or non-working state data into the anomaly identification process and improve the effectiveness of the samples participating in the identification. This invention determines the operating condition by counting the number of active measuring points, and can improve the stability of effective operating condition judgment by utilizing the spatial information of the multiple measuring points of the trolley mechanism. This invention performs envelope demodulation and spectral amplitude statistics within the target frequency band under effective operating conditions, which can characterize the impact modulation-related changes in the vibration signals of the guide wheels, drive wheels and other parts of the trolley mechanism. The robust discrete eigenvalues ​​obtained by this invention are used to characterize the dispersion of the envelope spectrum amplitude distribution within the target frequency band. This can reduce the influence of individual extreme spectral lines or occasional disturbances on the stability of the characteristics, and is beneficial to improving the robustness of on-site vibration anomaly identification. This invention is applicable to online monitoring, offline analysis, and condition evaluation of the trolley mechanism of port rail cranes, and can provide stable characteristic basis for equipment maintenance and anomaly early warning. Attached Figure Description

[0016] Figure 1 This is a schematic diagram of the method for identifying abnormal vibrations in the trolley mechanism of a port rail crane according to the present invention; Figure 2 This is a schematic diagram of the vibration anomaly identification method for the trolley mechanism of a port rail crane provided in an embodiment of the present invention; Figure 3 This is a schematic diagram of the robust discrete eigenvalue calculation process provided in an embodiment of the present invention; Figure 4 This is a comparison chart of robust discrete eigenvalues ​​of the trolley mechanism under healthy conditions and under severe fault conditions confirmed on-site, provided in an embodiment of the present invention. Detailed Implementation

[0017] The specific embodiments of the present invention will be further described below with reference to the accompanying drawings and technical solutions.

[0018] This embodiment provides a method and system for identifying abnormal vibrations in the trolley mechanism of a port rail-mounted crane. It can be applied to port yard rail-mounted cranes and also to lifting and transport equipment with similar trolley traveling mechanisms. The specific process is as follows: Figure 1 and Figure 2 As shown, the specific steps are as follows: First, multiple vibration measuring points are arranged on the trolley mechanism. These measuring points can be located at the guide wheels, drive wheels, bearing seats, or other positions that can reflect the vibration state of the trolley during operation. The system collects the vibration signals from each measuring point and records the data. m The vibration signals collected at each measuring point are ,in , M This represents the number of measurement points.

[0019] Then, based on the vibration signals at each measuring point or the state characterization signals converted from the vibration signals, the signals corresponding to each measuring point are divided into multiple signal segments according to time. The state characterization signals may be velocity signals, filtered signals, frequency band signals, or other signals that can characterize the operating state of the vehicle, converted from the vibration signals.

[0020] In one specific implementation, the state characterization signal is a velocity signal. After detrending the vibration signal, the detrended vibration signal is integrated to obtain the velocity signal; then, the velocity signal is band-filtered to retain the main energy frequency band related to the vibration of the trolley mechanism. The frequency range of the band-filtering process can be determined according to the main energy frequency band of the vibration of the trolley mechanism, for example, it can be from 10 Hz to 1000 Hz.

[0021] The state characterization signal corresponding to each measuring point is divided into multiple signal segments, and the operational activity index of each signal segment is calculated. The operational activity index may include at least one of the following: root mean square value, energy value, peak-to-peak value, effective value, and frequency band energy value. In one specific embodiment, the operational activity index is the root mean square value of velocity, which is used to characterize the vibration intensity of the corresponding signal segment.

[0022] For the same measurement point, the activity index of multiple signal segments is compared with a preset segment threshold, and the number of signal segments with an activity index greater than the preset segment threshold is counted. When this number reaches the preset segment number, the measurement point is determined to be in an active state; otherwise, the measurement point is determined to be in an inactive state.

[0023] Furthermore, the number of active measuring points is counted and compared with a preset threshold. When the number of active measuring points is greater than or equal to the preset threshold, the trolley mechanism is determined to be in an effective operating condition; when the number of active measuring points is less than the preset threshold, the trolley mechanism is determined to be in an invalid operating condition, and no vibration anomaly identification is performed on the signal under this condition.

[0024] Through the above process, this embodiment can avoid abnormal identification of vibration signals when the trolley mechanism is stationary, in a low-activity or non-working state, thereby reducing the impact of invalid working condition data on the identification results.

[0025] like Figure 3 As shown, when the trolley mechanism is in effective operating condition, the vibration signal of the measurement point to be analyzed is envelope demodulated and the spectral amplitude in the target frequency band is statistically analyzed to obtain robust discrete characteristic values ​​used to characterize the dispersion of the envelope spectrum amplitude distribution.

[0026] Specifically, the vibration signal of the measurement point to be analyzed is selected, and the vibration signal is preprocessed to obtain the demodulated signal. Preprocessing may include at least one of detrending, filtering, mean removal, normalization, and sampling rate unification. Through preprocessing, the influence of low-frequency trend terms, DC components, amplitude scale differences, or sampling rate differences on subsequent envelope demodulation and spectral amplitude statistics can be reduced.

[0027] Subsequently, an analytic signal is constructed from the signal to be demodulated, and the envelope signal is obtained based on the modulus of the analytic signal. In one specific embodiment, the analytic signal is constructed using the Hilbert transform. Let the signal to be demodulated be... Then analyze the signal It can be represented as: in, Represents the Hilbert transform. j Represents the imaginary unit. Envelope signal. It can be represented as: Then, the envelope signal obtained by envelope demodulation is subjected to frequency domain transformation to obtain the envelope spectrum. The frequency domain transformation can be a Fast Fourier Transform, a Real-Number Fast Fourier Transform, or other frequency domain transformation methods. The amplitude sequence within the target frequency band is extracted from the envelope spectrum. The target frequency band can be determined based on at least one of the following: the operating frequency of the trolley mechanism, the characteristic frequency of the transmission components, the structural vibration frequency, or the abnormal impact modulation frequency. In one specific embodiment, the target frequency band is 5 Hz to 1500 Hz.

[0028] Let the amplitude sequence within the target frequency band be... s Calculate the median of this amplitude sequence: Calculate the magnitude of each value in the magnitude sequence relative to the median. The absolute deviation is used to obtain the absolute deviation sequence: Calculate the median of the absolute deviation sequence: Robust discrete eigenvalues ​​are obtained based on the median of the absolute deviation sequence. In one specific implementation, the robust discrete eigenvalue is the standardized median absolute deviation eigenvalue, calculated using the following formula: in, F Represents robust discrete eigenvalues. d This represents the median of the absolute deviation sequence. This represents the standardized coefficient. The value is 0.6745.

[0029] This robust discrete eigenvalue reflects the degree of dispersion of the envelope spectrum amplitude sequence within the target frequency band relative to its median level. When abnormalities occur in the guide wheels, drive wheels, bearing housings, or related components of the trolley mechanism, the distribution of the envelope spectrum amplitude within the target frequency band may change, thereby causing changes in the robust discrete eigenvalue. Therefore, this robust discrete eigenvalue can be used for vibration anomaly identification or condition assessment of the trolley mechanism.

[0030] This embodiment uses measured vibration data from the trolley mechanism of a port rail crane at a port site as an example to illustrate the practical application of the method of the present invention. The trolley mechanism of this equipment experienced a serious malfunction during on-site operation, and the malfunction status was confirmed on-site. To identify this serious malfunction, eight vibration measurement points were arranged at the guide wheel, the guide-side drive wheel, and the non-guide-side drive wheel of the trolley mechanism. In this embodiment, the vibration signal sampling frequency is 12800 Hz, and each analysis window contains 128000 sampling points, corresponding to approximately 10 seconds of vibration signal. When processing the acceleration signals collected from each measurement point, effective operating conditions are first screened according to the aforementioned embodiment; after determining that the trolley mechanism is in an effective operating condition, robust discrete eigenvalue extraction is performed. Specifically, the vibration signal to be analyzed is demodulated using the Hilbert envelope to obtain the envelope signal; the envelope signal is frequency-domain transformed, and the envelope spectrum amplitude sequence within the target frequency band from 5 Hz to 1500 Hz is extracted; the standardized median absolute deviation is calculated based on this amplitude sequence to obtain the robust discrete eigenvalues. F To establish a health status baseline, 20 historical analysis windows of the trolley mechanism in a healthy operating state were selected as baseline samples. Robust discrete eigenvalues ​​were calculated for each measurement point, and the median value of the baseline samples at each measurement point was used as the health status eigenvalue. Subsequently, the same calculation was performed on the analysis windows corresponding to the periods of severe faults confirmed on-site, obtaining the robust discrete eigenvalues ​​under severe fault conditions. The calculation results for healthy and severe fault conditions are shown in Table 1 below.

[0031] Table 1

[0032] like Figure 4As shown, under the confirmed severe fault condition on-site, the robust discrete eigenvalues ​​(IDEVs) of multiple measuring points of the trolley mechanism all showed a significant increase compared to the healthy state baseline. The increase was particularly pronounced at guide wheel 1, guide wheel 2, and guide-side drive wheel 1, indicating a significant enhancement in the dispersion of the envelope spectrum amplitude distribution within the target frequency band under severe fault conditions. During anomaly identification, the anomaly judgment threshold can be constructed from the median value of the healthy samples and the robust discreteness of the healthy samples. For example, for the guide-side drive wheel 1 measuring point, the median of its robust discrete eigenvalue in the healthy state baseline sample is approximately 0.374, and the anomaly judgment threshold determined based on the dispersion of the healthy state samples is approximately 0.907. In the corresponding severe fault analysis window, the robust discrete eigenvalue of this measuring point is approximately 163.551, significantly exceeding the corresponding anomaly judgment threshold, thus triggering a severe vibration anomaly alarm for the trolley mechanism. Anomaly alarms can be output for other measuring points in the same manner.

Claims

1. A method for identifying abnormal vibrations in the trolley mechanism of a port rail-mounted crane, characterized in that, Includes the following steps: Step 1: Collect vibration signals from multiple measuring points on the trolley mechanism of the port rail crane; Step 2: Based on the state characterization signals obtained from the vibration signals of each measuring point, divide the signals corresponding to each measuring point into multiple signal segments according to time. Step 3: Calculate the operational activity index for each signal segment; Step 4: Determine whether each measuring point is in an active state based on the activity index of each signal segment; Step 5: Count the number of active measuring points and determine whether the trolley mechanism is in effective operating condition; Step 6: Under the effective operating conditions, perform envelope demodulation on the vibration signals of all measuring points; Step 7: Perform spectral amplitude statistics on the envelope signal within the target frequency band to obtain robust discrete characteristic values ​​that characterize the degree of dispersion of the envelope spectral amplitude distribution; Step 8: Identify vibration anomalies or evaluate the condition of the trolley mechanism based on the robust discrete eigenvalues.

2. The method for identifying abnormal vibrations in the trolley mechanism of a port rail crane according to claim 1, characterized in that, Step 1 is as follows: Measuring points are set at the guide wheels, drive wheels, bearing seats, or other locations that can reflect the vibration state of the trolley during operation; the number of measuring points is set to... M , No. m The discrete vibration signals collected at each measuring point are ,in , n This is the sampling point number.

3. The method for identifying abnormal vibrations in the trolley mechanism of a port rail crane according to claim 1, characterized in that, Step 2 is as follows: Using the velocity signal as the state characterization signal, the vibration signal is first detrended to obtain the detrended vibration signal. ; The de-stressed vibration signal is then integrated to obtain the initial velocity signal. Then, the initial velocity signal is subjected to frequency band filtering to obtain a velocity signal used for operating condition screening. The above process can be represented as: in, This indicates detrending processing. n Indicates the discrete-time sampling point number; This is a discrete-time index for the integration and summation process. , Indicates the sampling time interval. Indicates the frequency band range as Bandpass filtering; the frequency range of bandpass filtering is determined based on the main energy frequency band of the trolley mechanism's vibration; the state characterization signal is divided into... The signal segment, the first m The first measuring point k The velocity signal segment is denoted as ,in .

4. The method for identifying abnormal vibrations in the trolley mechanism of a port rail-mounted crane according to claim 1, characterized in that, Step 3 is as follows: The activity level metric uses the root mean square value of speed. m The first measuring point k The operational activity index of a signal segment Represented as: in, Indicates the first k The number of sampling points contained in a signal segment.

5. The method for identifying abnormal vibrations in the trolley mechanism of a port rail crane according to claim 1, characterized in that, Step 4 is as follows: The operational activity index of multiple signal segments corresponding to the same measurement point is compared with the preset segment threshold. Compare and obtain fragment activity markers. : Statistics m Number of active signal segments at each measurement point : When the number of signal segments Reach the preset number of segments When the time is right, the measuring point is determined to be in an active state; the active measuring point is marked. Represented as: 。 6. The method for identifying abnormal vibrations in the trolley mechanism of a port rail-mounted crane according to claim 1, characterized in that, Step 5 is as follows: Number of active measurement points C Represented as: Number of active measurement points C Compared with the preset threshold number of measurement points Compare; when When the trolley mechanism is in a valid operating condition, it is determined that the trolley mechanism is in a valid operating condition; when When the trolley mechanism is in an invalid operating condition, it is determined that the signal under this condition is not identified as having abnormal vibration; valid operating condition is marked. G Represented as: 。 7. The method for identifying abnormal vibrations in the trolley mechanism of a port rail crane according to claim 1, characterized in that, Step 6 is as follows: The vibration signal from the measurement point to be analyzed is preprocessed to obtain the demodulated signal. ; The preprocessing includes at least one of detrending, filtering, mean removal, normalization, and sampling rate unification; the signal to be demodulated is constructed using analytical signals, and the envelope signal is obtained based on the modulus of the analytical signals; the analytical signal construction is achieved through Hilbert transform, and the analytical signal... and envelope signal They are represented as follows: in, Represents the Hilbert transform. j It represents the imaginary unit.

8. The method for identifying abnormal vibrations in the trolley mechanism of a port rail-mounted crane according to claim 1, characterized in that, Step 7 is as follows: For envelope signal Perform a frequency domain transformation to obtain the envelope spectrum amplitude. : Extract the target frequency band from the envelope spectrum. amplitude sequence within : in, Indicates the first q The frequency corresponding to each frequency point; the target frequency band is determined based on at least one of the following: the operating frequency of the trolley mechanism, the characteristic frequency of the transmission components, the structural vibration frequency, or the abnormal impact modulation frequency; Calculate the amplitude sequence the median of : Calculate the absolute deviation of each amplitude in the amplitude sequence relative to the median to obtain the absolute deviation sequence. : Calculate the median of the absolute deviation sequence d : When the robust discrete eigenvalue is the standardized median absolute deviation eigenvalue, it is calculated according to the following formula: in, F Represents robust discrete eigenvalues; This represents the standardization coefficient.

9. The method for identifying abnormal vibrations in the trolley mechanism of a port rail crane according to claim 1, characterized in that, Step 8 is as follows: robust discrete eigenvalues F Compared with the preset abnormal threshold Comparison is performed to obtain anomaly identification results. R : in, R =1 indicates that there is abnormal vibration in the trolley mechanism. R =0 indicates that the trolley mechanism is in a normal vibration state or has not met the abnormal judgment condition; preset abnormal threshold. Determined based on historical normal operation data, field experience, statistical analysis results, or equipment maintenance requirements.

10. A vibration anomaly identification system for a port rail-mounted crane trolley mechanism, used to implement the method described in any one of claims 1-9, characterized in that, Specifically, it includes: The vibration acquisition module is used to collect vibration signals from multiple measuring points on the trolley mechanism of a port rail crane. The working condition screening module is used to divide the signal corresponding to each measuring point into multiple signal segments according to time based on the vibration signal of each measuring point or the state characterization signal converted from the vibration signal, calculate the operating activity index of each signal segment, determine whether each measuring point is in an active state based on the operating activity index, and determine whether the trolley mechanism is in an effective operating condition based on the number of measuring points in an active state. The feature extraction module is used to perform envelope demodulation and spectral amplitude statistics within the target frequency band of the vibration signal of the measurement point to be analyzed under effective operating conditions, so as to obtain robust discrete feature values ​​that characterize the degree of dispersion of the envelope spectrum amplitude distribution. The anomaly identification module is used to identify vibration anomalies or evaluate the condition of the trolley mechanism based on robust discrete eigenvalues.

Images