A digital testing method and device for key components of lifting equipment

Abstract

This application discloses a digital detection method and apparatus for key components of lifting equipment, relating to the field of digital detection of lifting equipment. The method includes: acquiring daily maintenance and repair records of the lifting equipment; determining high-frequency fault points of the lifting equipment based on these records; acquiring detection data collected by sensors installed at each high-frequency fault point; extracting time-frequency domain features from the collected detection data and optimizing these features to determine core features; and determining abnormal data in the detection data and the corresponding component locations based on these core features. This application analyzes high-frequency fault points of the lifting equipment through daily maintenance and repair records, and by installing targeted sensors at these high-frequency fault points to collect component operating data in real time, it can detect component abnormalities in real time through digital detection.

Description

Technical Field

[0001] This application relates to the field of digital inspection of lifting equipment, and in particular to a method and apparatus for digital inspection of key components of lifting equipment. Background Technology

[0002] In current crane maintenance and repair work, the operating status of cranes is often judged by visual inspection or sound. This method cannot accurately locate fault points or predict trends, and is also affected by external environmental factors such as oil stains, paint, and noise, leading to significant errors in the final judgment of the crane's current status. This results in significant safety hazards and can easily lead to various safety accidents. Furthermore, traditional crane maintenance and repair work requires personnel to conduct comprehensive observations of multiple cranes, without the ability to perform digital detection of fault points. This wastes a significant amount of time, increases the risk of data confusion, and slows down the maintenance process, greatly reducing personnel efficiency. Additionally, routine maintenance cannot compare and analyze historical data, cannot understand the operating trends of high-frequency fault points in real time, cannot predict fault occurrences in advance, and does not pay attention to the equipment's operating condition, thus failing to gain a more detailed understanding of the mechanism and the true condition of the equipment. Summary of the Invention

[0003] The purpose of this application is to provide a digital detection method and device for key components of lifting equipment, which can detect abnormalities of components in real time through digital detection.

[0004] To achieve the above objectives, this application provides the following solution: Firstly, this application provides a digital inspection method for key components of lifting equipment, including: Obtain daily maintenance and repair records of lifting equipment; Identify the high-frequency failure points of the lifting equipment based on its daily maintenance and repair records; the components corresponding to the high-frequency failure points are considered critical components. Acquire detection data collected by sensors installed at each high-frequency fault point; The collected detection data are subjected to time-frequency domain feature extraction and feature optimization to determine the core features; Based on the core features, identify the abnormal data in the detection data and the location of the corresponding components.

[0005] Secondly, this application provides a computer device, including: a memory, a processor, and a computer program stored in the memory and capable of running on the processor, wherein the processor executes the computer program to implement the above-described digital detection method for key components of lifting equipment.

[0006] Thirdly, this application provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the aforementioned digital detection method for key components of lifting equipment.

[0007] According to the specific embodiments provided in this application, this application has the following technical effects: This application provides a digital detection method and apparatus for key components of lifting equipment. The method includes: acquiring daily maintenance and repair records of the lifting equipment; determining high-frequency fault points of the lifting equipment based on the daily maintenance and repair records; acquiring detection data collected by sensors installed at each high-frequency fault point; extracting time-frequency domain features from the collected detection data and optimizing the features to determine core features; and determining abnormal data in the detection data and the corresponding component locations based on the core features. This application analyzes high-frequency fault points of the lifting equipment through daily maintenance and repair records, and collects component operating data in real time by installing targeted sensors at high-frequency fault points, thereby enabling real-time detection of component abnormalities through digital detection. Attached Figure Description

[0008] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the embodiments 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.

[0009] Figure 1 This is an application environment diagram of a digital detection method for key components of lifting equipment according to an embodiment of this application; Figure 2 A flowchart illustrating a digital inspection method for key components of lifting equipment, provided as an embodiment of this application; Figure 3 This is a schematic diagram of the structure of a computer device provided in an embodiment of this application. Detailed Implementation

[0010] 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.

[0011] To make the above-mentioned objectives, features and advantages of this application more apparent and understandable, the application will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0012] The digital inspection method for key components of lifting equipment provided in this application can be applied to, for example... Figure 1 In the application environment shown, terminal 102 communicates with server 104 via a network. A data storage system can store the data that server 104 needs to process. The data storage system can be set up independently, integrated into server 104, or placed in the cloud or on another server. Terminal 102 can send the daily maintenance and repair records of the lifting equipment to server 104. After receiving the daily maintenance and repair records, server 104 acquires the records; determines the high-frequency fault points of the lifting equipment based on the records; considers the components corresponding to the high-frequency fault points as critical components; acquires the detection data collected by sensors installed at each high-frequency fault point; performs time-frequency domain feature extraction and feature optimization on the collected detection data to determine core features; and determines the abnormal data in the detection data and the location of the corresponding components based on the core features. Server 104 can then feed back the obtained abnormal data and the location of the corresponding components to terminal 102.

[0013] Among them, terminal 102 can be, but is not limited to, various desktop computers, laptops, smartphones, tablets, IoT devices and portable wearable devices, and server 104 can be implemented by independent servers or server clusters composed of multiple servers, or it can be a cloud server.

[0014] In one exemplary embodiment, such as Figure 2 As shown, a digital inspection method for key components of lifting equipment is provided. This method is executed by computer equipment, specifically by a terminal or server alone, or by both a terminal and a server. In this embodiment, the method is applied to... Figure 1 Taking server 104 as an example, the explanation includes the following steps 201 to 205. Wherein: Step 201: Obtain the daily maintenance and upkeep records of the lifting equipment.

[0015] Step 202: Determine the high-frequency failure points of the lifting equipment based on the daily maintenance and repair records of the lifting equipment; the components corresponding to the high-frequency failure points are regarded as critical components.

[0016] Step 203: Obtain the detection data collected by the sensors installed at each high-frequency fault point.

[0017] Step 204: Extract time-frequency domain features from the collected detection data and optimize the features to determine the core features.

[0018] Step 205: Determine the abnormal data in the detection data and the location of the component corresponding to the abnormal data based on the core features.

[0019] By implementing steps 201 to 205 above, this application analyzes the high-frequency fault points of the lifting equipment through daily maintenance and repair records, and collects the operating data of the components in real time by installing targeted sensors at the high-frequency fault points, thereby enabling the real-time detection of abnormal conditions of the components through digital detection.

[0020] In another exemplary embodiment of this application, step 202, determining the high-frequency failure points of the lifting equipment based on the daily maintenance and upkeep records of the lifting equipment, specifically includes: (a1) Standardize the daily maintenance and repair records of lifting equipment to obtain standardized data after preprocessing.

[0021] Collect raw data such as daily maintenance records, maintenance logs, fault reports, and component replacement records for high-risk lifting equipment. This raw data can originate from the equipment user's maintenance management system, paper archives, electronic documents, etc. The collected records undergo standardization processing, specifically including: 1) Unified data format: Convert electronic records (Excel, Word, PDF, CSV) and scanned copies of paper records of different formats into a unified structured data format to facilitate subsequent processing.

[0022] 2) Standardization of fault terminology: Refer to relevant industry standards for lifting equipment (such as GB / T 6067.1-2010 "Safety Regulations for Lifting Machinery Part 1: General Rules") to standardize the fault terminology in the records, so as to avoid data confusion caused by different expressions of the same fault. For example, "the weld is cracked" and "the weld is cracked" are standardized as "the weld is cracked".

[0023] 3) Missing data completion: For key information missing in the records (such as the time of the fault and the location of the fault), complete the data by checking related records and consulting maintenance personnel. Data that cannot be completed is marked as invalid data.

[0024] 4) Duplicate data removal: Remove duplicate entries or records with completely identical content to ensure data uniqueness.

[0025] After preprocessing, core information is extracted and organized into a standardized dataset. The core information includes: time of failure, location of failure, type of failure, cause of failure, maintenance measures, component model, equipment working status (lifting / luffing / slewing / braking), load weight, and ambient temperature.

[0026] (a2) Construct a typical failure mode library for lifting equipment based on the preprocessed standardized data; the typical failure mode library for lifting equipment is dynamically updated with the updates of daily maintenance and repair records.

[0027] Based on the preprocessed standardized dataset, fault information is categorized and organized into three main types: structural systems, mechanical systems, and electrical control systems, thus constructing a typical failure mode library. The specific process is as follows: 1) System Classification: Structural systems include metal structures (main beams, booms, outriggers, etc.) and connecting welds; Mechanism systems include hoisting mechanisms, luffing mechanisms, slewing mechanisms, and braking mechanisms, with core components including reducers, motors, bearings, gears, and wire ropes; Electrical control systems include controllers, sensors, frequency converters, contactors, and wiring.

[0028] 2) Failure Mode Analysis: For each type of system, extract the fault types and failure characteristics from the records, and summarize typical failure modes, such as "deformation of metal structure", "cracking of critical welds" and "corrosion of components" for structural systems, "wear of reducer shaft", "failure of gear meshing", "damage to bearings" and "fatigue fracture of wire rope" for mechanical systems, and "overheating of components", "aging of circuits", "failure of sensor" and "failure of controller" for electrical control systems.

[0029] 3) Improved Failure Mode Attributes: Supplement attribute information for each typical failure mode, including occurrence frequency (e.g., statistics on the number of occurrences in the past 3-5 years), associated components (e.g., "gearbox shaft wear" is associated with "gearbox input / output shaft"), applicable operating conditions (e.g., "metal structure deformation" often occurs in lifting and heavy-load operating conditions), and failure consequences (e.g., "critical weld cracking" may lead to structural collapse).

[0030] The failure mode library supports dynamic updates. When a new maintenance record is added, the new failure mode is automatically filtered and added to the library.

[0031] (a3) Construct a failure tree model for lifting equipment based on the failure causal relationships in the typical failure mode library of lifting equipment.

[0032] Using the top-level failure events of the structural system, mechanism system, and electrical control system as root nodes, and based on the fault causal relationships in the standardized dataset, the failure tree model of each system is constructed by deductive method into intermediate failure events and bottom events, and the logical relationships (AND gate, OR gate, NOT gate) between each failure event are clarified.

[0033] (a4) Screen high-frequency failure points of lifting equipment based on the typical failure mode library and failure tree model of lifting equipment.

[0034] In another exemplary embodiment of this application, step (a4), screening high-frequency failure points of the lifting equipment based on the typical failure mode library and the failure tree model of the lifting equipment, specifically includes: (a4-1) Based on the typical failure mode library of lifting equipment, count the number of occurrences of each type of failure mode in a preset time period, and calculate the occurrence frequency of each type of failure mode in the preset time period.

[0035] Failure mode occurrence frequency statistics: Based on the failure mode library, the occurrence frequency of various failure modes in the past 3-5 years is counted, and the occurrence frequency is calculated (occurrence frequency = number of occurrences / statistical years). High-frequency failure modes with an occurrence frequency of ≥5 times / year are selected.

[0036] (a4-2) Use the minimum cut set method or the probability importance method to calculate the degree of influence of each bottom failure event on the top failure event in the failure tree model.

[0037] Basic event importance analysis: The influence (importance) of each basic event in the failure tree on the top-level failure event is calculated using either the minimum cut set method or the probabilistic importance method. The minimum cut set method identifies all minimum cut sets (the minimum set of basic events that can cause the top-level failure event) that lead to the top-level failure event and counts the frequency of each basic event in these sets; the higher the frequency, the higher the importance. The probabilistic importance method quantifies the importance of basic events by calculating the influence of the probability of occurrence of a basic event on the probability of occurrence of a top-level event.

[0038] (a4-3) Determine the high-frequency failure points of the lifting equipment based on the occurrence frequency of various failure modes in the preset time period and the degree of influence of each bottom failure event on the top failure event in the failure tree model.

[0039] High-frequency failure point screening: Combining high-frequency failure modes and the importance ranking of low-level events, the failure locations corresponding to low-level events that occur ≥5 times / year and rank in the top 30% of importance are screened as high-frequency failure points. Specifically, these include: easily deformable areas of metal structures (middle section of main beam, boom head), critical easily cracked welds (welds at the root of boom, welds connecting the main beam and end beam), easily damaged shafts of reducers (input shaft, output shaft), and heat-concentrated components of the electrical control system (IGBT components of frequency converter, contactor contacts), etc.

[0040] This embodiment can establish typical failure modes and failure trees for the structure, mechanism, and electrical control system based on the daily maintenance and repair records of high-risk lifting equipment. It can accurately locate high-frequency failure points such as easily deformable areas of the metal structure, critical and easily cracked welds, vulnerable shafts of the reducer, and locations of heat-concentrated components in the electrical control system. It has the following beneficial effects: (1) Full life cycle data coverage makes failure pattern mining more comprehensive and avoids the one-sidedness caused by traditional methods relying on single maintenance data; make full use of historical maintenance data of high-risk lifting equipment to mine failure patterns, avoid the limitations of traditional methods relying on experience judgment, and improve the accuracy and scientific nature of fault location.

[0041] (2) By constructing a failure mode library and a failure tree model, the failure relationships of structure, mechanism and electrical control system are sorted out, which can quickly locate high-frequency fault points such as easily deformable areas of metal structures and key easily cracked welds, improving the location efficiency by more than 60%.

[0042] (3) It can realize the visualization and standardization of fault location, and mark the fault location in conjunction with equipment drawings, so that maintenance personnel can quickly find the fault location, carry out preventive maintenance, and reduce the downtime rate by more than 30%.

[0043] (4) It supports dynamic updates and model corrections, and can adapt to the fault characteristics of different types of high-risk lifting equipment. It has a wide range of applications and strong practicality.

[0044] In another exemplary embodiment of this application, step 203, acquiring detection data collected by sensors installed at each high-frequency fault point, specifically includes: (b1) Determine the type of sensor to be installed based on the fault mode at each high-frequency fault point.

[0045] Based on the fault type and operating conditions of the fault point, install targeted sensors: for example, install stress and strain sensors in easily deformable areas of metal structures, install eddy current crack sensors in easily cracked welds, install vibration sensors on easily damaged shafts of reducers, and install temperature imaging sensors in areas of concentrated heat in the electrical control system.

[0046] (b2) Construct a digital twin model of the lifting equipment, use the digital twin model of the lifting equipment to simulate the measurement angle and signal transmission path after the sensor is installed, and optimize the sensor installation angle and fixing method through finite element analysis.

[0047] Based on the digital twin model of the lifting equipment, the measurement angle and signal transmission path after the sensor is installed are simulated. The sensor installation angle and fixing method are optimized through finite element analysis to ensure that the sensor can accurately collect target parameters (such as the installation angle of the stress strain sensor is consistent with the direction of the force on the structure, with an error ≤ ±2°).

[0048] A customized shielding cover (metal shielding mesh + insulation and heat insulation layer) is configured for the sensor. Shielded cables are used for the sensor cables, and the wiring path is optimized to avoid parallel laying with power cables, thereby reducing electromagnetic coupling interference and improving the sensor data acquisition accuracy by more than 15%.

[0049] (b3) Obtain detection data collected by the sensor based on the optimized sensor installation angle and fixing method.

[0050] In addition, this application also sets up a sensor self-calibration mechanism. One optional implementation is that the sensor automatically triggers a self-calibration process every 72 hours, correcting measurement deviations by comparing with preset standard values. Another optional implementation is that, for critical sensors (such as crack detection sensors), a calibration deviation prediction model is established by combining historical calibration data from the cloud, predicting sensor drift trends in advance, and automatically reminding manual calibration.

[0051] In another exemplary embodiment of this application, different types of sensors for high-frequency fault points are connected to a data acquisition terminal. The functions of the data acquisition terminal are: (1) to realize real-time acquisition of sensor data; (2) to perform edge computing on real-time data and extract feature values; and (3) to store feature data according to time series. In step 204, the design is tailored to the high-frequency fault point monitoring scenario of the lifting machinery, and the edge computing power constraint of the data acquisition terminal is adapted (low power consumption and strong real-time performance). Combined with the core working scenarios of the lifting machinery (lifting, luffing, slewing, braking), various monitoring parameters such as structural stress concentration, weld cracks, reducer vibration, and temperature field of the electrical control system are specifically adapted to realize the linkage analysis of multiple parameters and working status, covering four core links: data preprocessing, time domain feature extraction, frequency domain feature extraction, and feature optimization. Specifically, in step 204, the time and frequency domain features of the collected detection data are extracted and the features are optimized to determine the core features, including: (c1) The collected detection data is denoised and normalized to obtain the preprocessed data.

[0052] (1) Perform wavelet transform denoising (adapt to multiple parameters: speed reducer vibration / structural stress / weld crack signal, resist industrial electromagnetic interference) Discrete wavelet transform is used to separate effective signals from interference signals. Parameters are optimized to address the differences between reducer vibration (high-frequency impact signal) and structural stress / weld cracks (low-frequency strain signal), adapting to the lightweight requirements of edge computing. The formula is as follows: in, These are wavelet transform coefficients (intermediate quantities output by edge calculation, used for subsequent signal reconstruction). The scaling factor (corresponding to the signal frequency, adapting to multiple parameters and linked to the operating state) Level-adaptive structural stress / weld crack low-frequency signal (strain fluctuation under lifting / braking conditions of crane machinery). High-frequency vibration signal of the speed reducer (gear meshing impact under slewing / amplitude conditions); The larger the value, the lower the frequency signal (the valid fault signal). The smaller the value, the higher the frequency of the signal (electromagnetic interference signal). This value can be dynamically adjusted according to the real-time working status of the lifting equipment to adapt to the fault characteristics of the lifting machinery. The shift factor (corresponding to the signal time dimension, matching the time-series data acquisition interval); Raw sampling data from multiple types of sensors (the first) The values ​​at each sampling point include: the vibration acceleration of the reducer, the strain value in the structural stress concentration area, the strain value of the weld crack monitoring, and the temperature value of the electrical control system. The wavelet basis function is preferred (db4 wavelet is preferred, which balances denoising effect and computational load, and can be implemented in a fixed manner on the edge computing terminal). For sampling point number ( , (This represents the total number of sampling points within a single time window).

[0053] Denoising signal reconstruction formula (edge ​​computing outputs valid original data): in, The effective sampled data after denoising (final output of edge computing preprocessing); Wavelet coefficients after thresholding (adapted to multi-parameter characteristics: for high-frequency scales) The coefficient of ) is set to 0 to filter electromagnetic interference; among which, structural stress / weld crack signal Vibration signal of speed reducer The temperature signal (gradually changing signal) of the electronic control system can be directly used. Simplify the process; It can be dynamically configured according to the real-time working status of the lifting machinery (such as lifting heavy load / no load). The minimum scale threshold; This is the maximum scale threshold.

[0054] (2) Data normalization (adapting to differences in the dimensions of multiple parameters and supporting the analysis of working conditions) The min-max normalization algorithm is used, which has low computational cost and meets the real-time requirements of edge computing. The formula is as follows: in, Normalized sampled data (value range [0,1], final output of edge computing preprocessing, used for subsequent feature extraction); This represents the minimum value of the original data of the target parameter within a single time window (adapting to multiple parameters and operating conditions: such as the minimum structural stress under lifting heavy load conditions, the minimum temperature of the electronic control system under braking conditions, and the minimum vibration acceleration of the reducer under slewing conditions). This represents the maximum value of the original data of the target parameter within a single time window (adapting to multiple parameters and operating conditions: such as the maximum structural stress under lifting and heavy load conditions, the maximum temperature of the electronic control system under braking conditions, and the maximum vibration acceleration of the reducer under slewing conditions).

[0055] The above preprocessing calculations only involve subtraction and division, without the need for complex matrix operations. On edge computing platforms such as ARM Cortex-A series chips, the processing time for a single window is ≤1ms, which meets the real-time requirements.

[0056] (c2) Extract time-domain and frequency-domain features from the preprocessed data to obtain the extracted time-frequency domain features.

[0057] (1) Extraction of time-domain feature values ​​(adapting to multiple parameters and linking the working status of lifting equipment) Temporal features are the fundamental features for fault monitoring of lifting machinery. Combined with the core working states of lifting equipment (lifting, luffing, slewing, braking), differentiated extraction logic is designed for multiple parameters such as structural stress concentration, weld cracks, reducer vibration, and temperature field of the electrical control system. The following formulas cover the core statistics strongly correlated with faults and are all suitable for the lightweight edge computing requirements: 1) Average value (reflects the overall level of multiple parameters, and is relevant to coordinated operating conditions: such as the average structural stress under heavy lifting loads, and the average electronic control temperature under braking conditions). The specific calculation formula is as follows: in, : The mean of normalized data within a single time window (time domain feature value 1, edge calculation output); This represents the total number of sampling points within a single time window (adapting to multiple parameters and dynamic configurations of operating conditions: gearbox vibration (slewing / amplitude conditions)). Structural stress / weld cracks (lifting operation) Temperature field of electronic control system (braking condition) (Adaptable to different parameter sampling frequency requirements) This represents the normalized, valid sampled data.

[0058] 2) Variance (reflects the degree of dispersion of multiple parameters, and is used for early warning of operating conditions: such as the vibration variance of the reducer under variable luffing conditions and the structural stress variance under heavy lifting loads), the specific calculation formula is as follows: in, The variance of normalized data within a single time window (time domain eigenvalue 2, edge calculation output); using As the denominator (unbiased estimate), optimization is performed for different parameters and operating conditions: For parameters sensitive to small fluctuations such as structural stress concentration and weld cracks, the variance threshold can be adjusted through operating condition linkage (e.g., the stress variance warning threshold is higher under heavy-load lifting conditions than under no-load conditions), compared to The denominator can more accurately reflect the minute data fluctuations in the early stages of crane machinery failure, thus improving the sensitivity of fault early warning.

[0059] 3) Peak Factor (captures sudden fault peaks, adapts to multiple parameters: reducer gear impact, weld crack propagation, electrical control system overheating), the specific calculation formula is as follows: in, For peak factor (time domain eigenvalue 3, edge calculation output, dimensionless); The denominator is the maximum absolute value of the normalized data within a single time window; The root mean square (RMS) reflects the energy level of the data.

[0060] Application scenarios of peak factor: Achieving accurate early warning of the working status of lifting equipment: Under the slewing condition, the sudden change of the vibration peak factor of the reducer (exceeding the preset threshold) corresponds to gear meshing failure; under the lifting heavy load condition, the sudden change of the weld crack monitoring peak factor corresponds to crack propagation; under the braking condition, the sudden change of the temperature peak factor of the electronic control system corresponds to circuit overheating. It can provide early warning 1 to 3 minutes before the fault occurs, providing core basis for real-time protection.

[0061] 4) Trend change rate (reflects the fault development trend; linked operating conditions: structural stress trend in lifting operation and electronic control temperature trend in braking operation), the specific calculation formula is as follows: in, The mean trend change rate (time domain feature value 4, edge calculation output, percentage form); For the current time window ( The mean at time (time); For the front a time window ( The mean at time (time); The time window step size (adapts to dynamic configurations of multiple parameters and operating conditions: structural stress / weld cracks (slow fault development)). Gearbox vibration (rapid fault development) Temperature field of the electronic control system (rapid temperature rise under braking conditions) (Accurately match the fault development speed with different parameters).

[0062] Trend change rate can quantify the degree of failure deterioration and provide basic data for predicting the remaining life of equipment.

[0063] (2) Frequency domain feature extraction (focusing on the core vibration of the reducer, combined with slewing / amplitude conditions) To address the vibration faults (gear wear, bearing damage) of reducers under slewing / luffing conditions in lifting machinery, a Fast Fourier Transform (FFT) is used to extract frequency domain features. Frequency analysis is optimized by combining the reducer speed with the operating conditions. This can be quickly implemented at the edge terminal through hardware acceleration (FPGA).

[0064] 1) Discrete Fast Fourier Transform (FFT) formula (time-domain signal to frequency-domain signal) in, It is a frequency domain complex signal (intermediate output of edge computing, containing amplitude and phase information); For frequency point serial numbers (corresponding to different frequency components); The imaginary unit ( ).

[0065] Using the radix-2 FFT algorithm, the computational complexity is reduced from Reduce to ; Dynamic adjustment of reducer speed under combined slewing / amplitude conditions: FFT window length (at high speeds) When the speed is low On the edge terminal (equipped with an FPGA module), the processing time for a single window is ≤10ms, which meets the real-time requirements.

[0066] 2) Frequency domain amplitude spectrum formula (extracts the amplitude of fault characteristic frequencies and eliminates phase interference) in, For the first The amplitude at each frequency point (intermediate quantity of frequency domain characteristics); for The real part; for The imaginary part.

[0067] 3) Peak frequency (accurately identifies reducer fault type, linkage condition and speed), the specific calculation formula is as follows: in, Peak frequency (frequency domain eigenvalue 1, edge calculation output, unit: Hz); The frequency index corresponding to the maximum amplitude (i.e.) ); The sensor sampling frequency (dynamically configured for combined slewing / amplitude operation: when the speed is ≥100 r / min) is set. When the rotational speed is <100 r / min (Adaptable to different operating speeds of lifting machinery reducers).

[0068] The application scenario for peak frequency is: to achieve accurate fault identification based on the speed and operating conditions of the linkage reducer. Compared with the inherent failure frequency of the lifting machinery reducer (inner ring failure frequency) outer ring failure frequency ,in The real-time speed of the reducer (obtained from the working condition linkage) is consistent, which can directly determine the fault type; for example, when the speed of the reducer decreases under variable amplitude working conditions, the fault frequency decreases synchronously, and the algorithm can dynamically match the threshold to achieve accurate fault identification at the edge.

[0069] 4) Specific frequency band energy (reflects the wear degree of the reducer, and the frequency band is dynamically adjusted under linkage conditions) in, Signal energy within a specific frequency band (target frequency band) (frequency domain characteristic value 2, edge calculation output); , The starting and ending frequency points of the target frequency band (dynamic configuration of the linkage reducer operating conditions: 500Hz~2kHz frequency band for slewing condition (high speed), 100Hz~1kHz frequency band for variable amplitude condition (low speed), accurately matching the fault characteristic frequency band under different operating conditions, configurable).

[0070] It can preset dedicated frequency bands for different fault points (bearings / gears), which reduces the amount of calculation by more than 60% compared with full-band energy calculation, while improving fault identification.

[0071] (c3) Apply principal component analysis to extract core features from the extracted time-frequency domain features.

[0072] In another exemplary embodiment of this application, principal component analysis (PCA) is used to achieve dimensionality reduction of multi-parameter features, fusing features such as structural stress, weld cracks, reducer vibration, and electrical control temperature field, and combining them with the real-time working status of the lifting equipment to screen core features, adapting to edge computing power constraints. Specifically, in step (c3), principal component analysis is applied to the extracted time-frequency domain features to extract core features, including: (c3-1) Construct the feature covariance matrix based on the extracted time-frequency domain features.

[0073] The formula for calculating the characteristic covariance matrix is: in, for The covariance matrix of order ( This includes multiple parameter-based original feature dimensions, such as the mean / variance of structural stress, weld crack peak factor, gearbox vibration peak frequency / band energy, and the rate of change of electronic control temperature trend. ); The number of feature samples (number of continuous time windows under a single working condition; linkage working condition configuration: hoisting / braking condition (large parameter fluctuations)). No-load condition (parameters stable) ); for The original feature matrix of order (each row corresponds to a time window) (one eigenvalue) for A feature mean vector (the mean of each row of features).

[0074] (c3-2) Solve for the eigenvalues ​​and eigenvectors of the characteristic covariance matrix.

[0075] The solution formula is: in, , which are the eigenvalue vectors of the covariance matrix (sorted from largest to smallest). ); , which is the corresponding eigenvector matrix (each column is an eigenvector).

[0076] (c3-3) Sort the eigenvalues ​​of the characteristic covariance matrix.

[0077] (c3-4) Determine the core features based on the extracted time-frequency domain features and the feature vectors corresponding to the L largest feature values.

[0078] Principal component feature extraction (reducing feature dimensionality), the specific formula is as follows: in, for The principal component feature matrix (the optimized feature of the final output of edge computing, dynamically adjusted under linked working conditions; in complex working conditions (lifting + luffing composite working conditions), In simple operating conditions (no-load rotation), , ); For the front The matrix consisting of the eigenvectors corresponding to the largest eigenvalues ​​( (Rank).

[0079] Principal component analysis (PCA) is applied to fuse multi-parameter features and combine them with working condition dimensionality reduction to transform the original... Dimensionality reduction of 3D features to Dimensionality reduction eliminates redundant features (such as redundant features of structural stress and weld cracks under no-load conditions), reducing subsequent time-series storage capacity and cloud transmission bandwidth by more than 60%, while retaining more than 95% of fault information; the dimensionality reduction logic is differentiated under different operating conditions to improve the targeting of fault identification.

[0080] In another exemplary embodiment of this application, the data acquisition terminals of each high-frequency fault point are connected to a wireless Bluetooth and 5G dual-mode transmission device to transmit the acquired data to a portable detector and / or a cloud platform, thereby achieving secure and efficient data transmission.

[0081] The data acquisition terminal supports both Bluetooth (short-range, low-power) and 5G (long-range, high-bandwidth) transmission. It uses the AES-256 encryption algorithm to encrypt the transmitted data and assigns a unique identity identifier (ID) to each data acquisition terminal. The identity authentication mechanism prevents data from being tampered with or stolen.

[0082] Maintenance personnel carry portable testing devices as they move through areas with high-risk lifting equipment. These portable rapid testing devices connect to and collect data from various high-frequency fault points via a dual-mode connection device, recording the data. If a data warning is triggered, the portable rapid testing device issues an audible and visual alarm and displays the location of the alarm data, the fault type, severity, and emergency handling suggestions. During off-site periods, data is uploaded to a cloud platform in real-time via 5G, enabling remote real-time monitoring of the equipment status.

[0083] Maintenance personnel carry portable detectors and move around in the high-risk lifting equipment area. The portable rapid detectors connect via Bluetooth and collect and record data from various high-frequency fault points. If a data warning occurs, the detector will issue an audible and visual alarm and display the location of the alarm data. Maintenance personnel can use portable rapid testing instruments to view real-time characteristic values, historical data, and digital twin models of equipment for each high-frequency fault point. Clicking on the fault warning location will display the associated components, maintenance records, and 3D structural drawings of that fault point. At the same time, standardized maintenance work orders can be automatically generated based on the fault information and uploaded to the enterprise operation and maintenance management system.

[0084] The cloud platform constructs a high-precision digital twin of equipment based on full lifecycle data and real-time transmission data, enabling real-time simulation and historical backtracking of equipment operating status; it integrates gray prediction model and gradient boosting tree algorithm to establish component remaining life prediction model, accurately predicting the remaining life of components corresponding to high-frequency and high-risk failure points with an error of ≤±10%; and it provides personalized operation and maintenance optimization suggestions to equipment users based on big data analysis, including optimal maintenance cycle adjustment, spare parts inventory optimization, and operator training suggestions.

[0085] Based on the above, the digital inspection method for key components of lifting equipment further includes: (d1) Upload the core features collected in real time to the cloud platform.

[0086] (d2) Utilize cloud platforms to construct digital twins of lifting equipment.

[0087] (d3) Using the digital twin of the lifting equipment, the real-time simulation of the equipment's operating status is carried out based on the core features collected in real time. This allows for the display of the fault warning location during fault warning and the prediction of potential fault risks.

[0088] Based on the above, the digital inspection method for key components of lifting equipment further includes: (e1) Based on the core characteristics of historical periods, a gray prediction model and / or gradient boosting tree algorithm are applied to construct a prediction model for the remaining life of key components of lifting equipment.

[0089] (e2) The remaining service life of key components corresponding to high-frequency failure points is predicted using the remaining service life prediction model of key components of lifting equipment.

[0090] Compared with the prior art, this application has the following advantages: (1) The high-frequency fault points established based on daily maintenance records have the advantage of being "tailored to local conditions" and provide targeted analysis for each high-frequency fault point of each device.

[0091] (2) Install targeted sensors at high-frequency fault points according to different fault types, use multiple data sources to diagnose and analyze high-risk lifting equipment, and use digital detection to avoid the risk of fault delays caused by the uneven experience or level of maintenance personnel, making daily maintenance more comprehensive and accurate.

[0092] (3) Data is connected to the wireless Bluetooth device at each high-frequency fault point. Data can be transmitted wirelessly, and characteristic data can be obtained without maintenance personnel having to go to the risk area.

[0093] (4) Algorithm innovation: Breaking through the limitations of general algorithms, it realizes deep linkage between multiple parameters and the working status of lifting equipment: ① Optimize core parameters such as wavelet scale and FFT window for the parameter characteristics of structural stress, weld cracks, reducer vibration and electric control temperature field; ② Dynamically adjust the sampling frequency and dimensionality reduction according to working conditions such as lifting, luffing, slewing and braking; ③ Integrate multi-parameter features to achieve working condition adaptability dimensionality reduction.

[0094] (5) Edge computing adaptability: All formulas have low computational complexity (maximum is 1000). The lightweight design of multi-parameter fusion and working condition linkage logic does not require high-performance computing power and can be implemented in real time on the data acquisition terminal (ARM / FPGA), solving the technical problems of high latency and high bandwidth consumption in traditional cloud computing.

[0095] (6) Fault Targeting: The formula and characteristic value are closely matched to the core fault scenarios of lifting machinery: ① Structural stress concentration and weld crack fault (linked lifting heavy load condition); ② Reducer vibration fault (linked slewing / luffing condition); ③ Electrical control system overheating fault (linked braking condition); Multi-parameter collaborative analysis improves the fault identification accuracy rate ≥98%, solves the problem of missed and false alarms in single parameter monitoring, and has clear engineering application value.

[0096] (7) Configurability: Key parameters in the formula (scale factor) Number of time windows Frequency band range Dimensionality Reduction All of these technologies can be flexibly configured according to different types of lifting machinery (bridge cranes, tower cranes, etc.) and different working conditions (lifting, luffing, slewing, etc.), which can not only adapt to single working condition monitoring, but also support collaborative analysis of composite working conditions, greatly expanding the scope of patent protection.

[0097] This application integrates four core technology systems: full lifecycle data mining for lifting equipment, multimodal sensor collaborative monitoring, condition-adaptive intelligent algorithms, and cloud-edge collaborative decision-making. By constructing a dynamically updated failure mode library and intelligent failure tree model, it accurately locates high-frequency failure points; employs a customized multimodal sensor array to collect multi-dimensional parameters; achieves real-time feature extraction and optimization at the edge based on condition-adaptive algorithms; and combines cloud-based big data analysis to establish a digital twin of the equipment, forming a closed-loop detection system encompassing "data acquisition - feature analysis - fault early warning - lifespan prediction - maintenance decision-making." This not only achieves integrated, collaborative, intelligent, digital, rapid, and periodic detection of multi-source damage, but also simulates equipment operating conditions through the digital twin, predicting potential failure risks in advance. It provides equipment users with accurate operating status data and maintenance decision support, significantly improving the accuracy, timeliness, and safety of high-risk lifting equipment maintenance, and significantly reducing enterprise maintenance and operation costs and the incidence of safety accidents.

[0098] This application also provides an application scenario in which the aforementioned digital detection method for key components of lifting equipment is applied. Specifically, the digital detection method for key components of lifting equipment provided in this embodiment can be applied to high-risk lifting equipment component detection scenarios. This scenario includes a data processing stage and a data organization stage; the data organization stage is used to organize the daily maintenance and repair records of the lifting equipment; the data processing stage is used to analyze high-frequency fault points based on the organized data and to analyze abnormal data by obtaining the data detected from high-frequency fault points. The digital detection method for key components of lifting equipment provided in this embodiment belongs to the data processing stage.

[0099] In one exemplary embodiment, a computer device is provided, which may be a server or a terminal, and its internal structure diagram may be as follows. Figure 3 As shown, the computer device includes a processor, memory, input / output (I / O) interfaces, and a communication interface. The processor, memory, and I / O interfaces are connected via a system bus, and the communication interface is also connected to the system bus via the I / O interfaces. The processor provides computational and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system, computer programs, and a database. The internal memory provides the environment for the operation of the operating system and computer programs stored in the non-volatile storage media. The database stores digital inspection data for key components of the lifting equipment. The I / O interfaces are used for information exchange between the processor and external devices. The communication interface is used for communication with external terminals via a network connection. When the computer program is executed by the processor, it implements a method for digital inspection of key components of lifting equipment.

[0100] Those skilled in the art will understand that Figure 3 The structures shown are merely block diagrams of some structures related to the present application and do not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than shown in the figures, or combine certain components, or have different component arrangements. In an exemplary embodiment, a computer device is provided, including a memory and a processor. The memory stores a computer program, and the processor executes the computer program to implement the steps in the above-described method embodiments.

[0101] In one exemplary embodiment, a computer-readable storage medium is provided storing a computer program that, when executed by a processor, implements the steps in the above-described method embodiments.

[0102] In one exemplary embodiment, a computer program product is provided, including a computer program that, when executed by a processor, implements the steps in the above-described method embodiments.

[0103] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, data stored, data displayed, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties. Moreover, the collection, use and processing of the relevant data are carried out in compliance with the relevant data protection laws and policies of the country where the location is located, and with the authorization granted by the owner of the corresponding device.

[0104] Those skilled in the art will understand that all or part of the processes in the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments described above. Any references to memory, databases, or other media used in the embodiments provided in this application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can take many forms, such as Static Random Access Memory (SRAM) or Dynamic Random Access Memory (DRAM).

[0105] The databases involved in the embodiments provided in this application may include at least one type of relational database and non-relational database. Non-relational databases may include, but are not limited to, blockchain-based distributed databases. The processors involved in the embodiments provided in this application may be general-purpose processors, central processing units, graphics processing units, digital signal processors, programmable logic devices, quantum computing-based data processing logic devices, etc., and are not limited to these.

[0106] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0107] This document uses specific examples to illustrate the principles and implementation methods of this application. The descriptions of the above embodiments are only for the purpose of helping to understand the methods and core ideas of this application. Furthermore, those skilled in the art will recognize that, based on the ideas of this application, there will be changes in the specific implementation methods and application scope. Therefore, the content of this specification should not be construed as a limitation of this application.

Claims

1. A digital inspection method for key components of lifting equipment, characterized in that, include: Obtain daily maintenance and repair records of lifting equipment; Identify the high-frequency failure points of the lifting equipment based on its daily maintenance and repair records. Components that experience high-frequency failures are considered critical components. Acquire detection data collected by sensors installed at each high-frequency fault point; The collected detection data are subjected to time-frequency domain feature extraction and feature optimization to determine the core features; Based on the core features, identify the abnormal data in the detection data and the location of the corresponding components.

2. The digital inspection method for key components of lifting equipment according to claim 1, characterized in that, Based on the daily maintenance and upkeep records of the lifting equipment, identify the high-frequency failure points of the lifting equipment, specifically including: Standardized preprocessing is performed on the daily maintenance and repair records of lifting equipment to obtain standardized data after preprocessing; A typical failure mode library for lifting equipment is constructed based on the preprocessed standardized data; the typical failure mode library for lifting equipment is dynamically updated as daily maintenance and repair records are updated. Based on the failure causal relationships in the typical failure mode library of lifting equipment, a failure tree model of lifting equipment is constructed. High-frequency failure points of lifting equipment are screened based on a library of typical failure modes and a failure tree model of lifting equipment.

3. The digital inspection method for key components of lifting equipment according to claim 2, characterized in that, Based on a library of typical failure modes and a failure tree model of lifting equipment, high-frequency failure points of lifting equipment are screened, specifically including: Based on the typical failure mode library of lifting equipment, the number of occurrences of each type of failure mode in a preset time period is counted, and the occurrence frequency of each type of failure mode in the preset time period is calculated. The minimum cut set method or the probability importance method are used to calculate the degree of influence of each bottom-level failure event on the top-level failure event in the failure tree model. The high-frequency failure points of the lifting equipment are determined based on the occurrence frequency of various failure modes in the preset time period and the degree of influence of each bottom failure event on the top failure event in the failure tree model.

4. The digital inspection method for key components of lifting equipment according to claim 1, characterized in that, Acquire detection data collected by sensors installed at each high-frequency fault point, specifically including: Determine the type of sensor to be installed based on the fault mode at each high-frequency fault point; A digital twin model of the lifting equipment is constructed, and the measurement angle and signal transmission path after sensor installation are simulated using the digital twin model of the lifting equipment. The sensor installation angle and fixing method are optimized through finite element analysis. Acquire detection data from sensors installed based on optimized sensor mounting angles and fixing methods.

5. The digital inspection method for key components of lifting equipment according to claim 1, characterized in that, The collected detection data undergoes time-frequency domain feature extraction and optimization to determine core features, specifically including: The collected detection data is denoised and normalized to obtain preprocessed data; Time-domain and frequency-domain feature values ​​are extracted from the preprocessed data to obtain the extracted time-frequency domain features; Principal component analysis was applied to extract the core features from the extracted time-frequency domain features.

6. The digital inspection method for key components of lifting equipment according to claim 5, characterized in that, Principal component analysis is applied to extract core features from the extracted time-frequency domain features, specifically including: Construct a feature covariance matrix based on the extracted time-frequency domain features; Find the eigenvalues ​​and eigenvectors of the characteristic covariance matrix; Sort the eigenvalues ​​of the feature covariance matrix; The core features are determined based on the extracted time-frequency domain features and the feature vectors corresponding to the L largest feature values.

7. The digital inspection method for key components of lifting equipment according to claim 1, characterized in that, The digital inspection method for key components of lifting equipment also includes: The core features collected in real time are uploaded to the cloud platform; Utilize cloud platforms to construct digital twins of lifting equipment; By utilizing a digital twin of the lifting equipment and real-time simulating the equipment's operating status based on real-time collected core features, the location of fault warnings can be displayed during fault warnings.

8. The digital inspection method for key components of lifting equipment according to claim 1, characterized in that, The digital inspection method for key components of lifting equipment also includes: Based on the core characteristics of historical periods, a gray prediction model and / or gradient boosting tree algorithm are applied to construct a prediction model for the remaining life of key components of lifting equipment. The remaining service life of critical components corresponding to high-frequency failure points is predicted using a critical component remaining service life prediction model for lifting equipment.

9. A computer device, comprising: A memory, a processor, and a computer program stored in the memory and capable of running on the processor, characterized in that the processor executes the computer program to implement the digital detection method for key components of lifting equipment as described in any one of claims 1-8.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When executed by a processor, the computer program implements the digital detection method for key components of lifting equipment as described in any one of claims 1-8.

Images