An intelligent fault prediction method and system for rotary cutting machine based on multi-sensor analysis

Abstract

The present invention discloses an intelligent fault prediction method and system for a rotary cutter based on multi-sensor analysis, and relates to the technical field of fault prediction. The method comprises: collecting audio data of the rotary cutter position within a working cycle of the rotary cutter, wherein the audio data covers the audio signal generated by the rotary cutter during the process of rotary cutting wood, wherein the working cycle is defined as the time required for the card shaft to rotate one circle; mapping the audio data into a waveform graph and forming a waveform curve, wherein the horizontal axis of the waveform graph represents time. The present invention analyzes the characteristic points in the audio waveform curve, identifies abnormal vibration modes, and based on the temperature changes captured by thermal imaging technology and the operating state of the rotary cutter captured by visual imaging technology, the present invention completes comprehensive monitoring of the operating status of the rotary cutter. In addition, by simulating the operating state of the rotary cutter, the potential faults of the rotary cutter during operation are predicted, thereby extending the service life of the rotary cutter to a certain extent.

Description

Technical Field

[0001] The present invention relates to the technical field of fault prediction, and in particular to an intelligent fault prediction method and system for a rotary cutter based on multi-sensor analysis. Background Art

[0002] Peeling machine is a kind of mechanical equipment used for wood processing. In order to ensure the efficient and stable operation of the peeling machine, fault prediction is particularly important. Through regular maintenance and inspection, potential problems can be discovered and repaired in time.

[0003] After searching, the Chinese invention patent with publication number "CN114252103A" discloses a "fusion power station operation fault prediction method". This application cleans all the original data of operating parameters directly sampled by measuring sensors, eliminates blank and obviously abnormal data, and then uses the extended Kalman filter algorithm to estimate the state of the cleaned data for operating parameters with linear relationships that meet the Kalman filter conditions, corrects the parameters, and obtains all corrected operating parameter values, thereby solving the problem of inaccurate data caused by measurement errors of various sensors in the system.

[0004] In addition, the Chinese invention patent with publication number "CN119043430A" discloses "a comprehensive test bench for equipment fault prediction and health management". This application connects the hydraulic system, gear transmission system, control cabinet assembly and oil cooler device to each other; the hydraulic system is used to monitor the physical and chemical properties, pollution and other indicators of the passing oil online; the gear transmission system is used to detect the oil in the gearbox and the wear state of the gears; the control cabinet assembly is used to control the operation of the hydraulic system and the gear transmission system respectively; the oil cooler device is set on one side of the hydraulic system to cool the oil in the hydraulic system. This can be used to simulate the actual operation of mechanical equipment test verification platform, so as to study the monitoring effect verification of the oil online monitor and the oil online monitoring sensor on the actual mechanical equipment, and improve the accuracy, stability and reliability of the monitoring data of the oil online monitor and the oil online monitoring sensor.

[0005] However, in the actual operation process, the above two disclosed patents and similar patents are usually unable to accurately predict the faults caused by the rotary cutter during operation due to factors such as limitations on the types and quantities of sensors, as well as insufficient data processing and analysis methods. For this reason, the applicant proposes an intelligent fault prediction method and system for rotary cutters based on multi-sensor analysis to solve the problems raised in the above background technology. Summary of the invention

[0006] The purpose of the present invention is to provide an intelligent fault prediction method and system for rotary cutting machines based on multi-sensor analysis to solve the problems raised in the above-mentioned background technology.

[0007] To achieve the above object, the present invention provides the following technical solutions:

[0008] The first solution proposes an intelligent fault prediction method for rotary cutting machines based on multi-sensor analysis, which includes:

[0009] Collect audio data of the position of the rotary cutter during a working cycle of the rotary cutter, wherein the audio data covers the audio signal generated by the rotary cutter during the process of rotary cutting the wood, wherein the working cycle is defined as the time required for the clamping shaft to rotate one circle;

[0010] Mapping the audio data into a waveform graph to form a waveform curve, wherein the horizontal axis of the waveform graph represents time and the vertical axis represents amplitude;

[0011] Extracting characteristic points of the rotary cutter operation from the waveform curve, wherein the characteristic points include wave crests and wave troughs;

[0012] Based on visual imaging technology, the image data corresponding to the time stamp of the feature point when the rotary cutter is running is captured, and the image data includes the vibration image, the angle change image and the blade position image of the rotary cutter;

[0013] Based on thermal imaging technology, the heat source distribution data in the image data is obtained, and the heat source distribution data reflects the temperature change of the rotary cutter assembly;

[0014] Obtain threshold parameters and initial parameters of the rotary cutter;

[0015] The operating state of the rotary cutter of the rotary cutter is simulated by using the threshold parameters and the initial parameters, and the heat source distribution data based on the thermal imaging technology under the threshold parameter state and the initial parameter state are collected;

[0016] The heat source distribution data was compared and analyzed, and potential faults at the rotary cutter position in the rotary cutting machine were predicted based on the analysis results.

[0017] As a further preferred embodiment of the present technical solution, the method of mapping audio data into a waveform diagram includes:

[0018] Acquiring acquisition points based on the frequency and amplitude of the audio data;

[0019] Arrange the collected points in time series to form a continuous curve;

[0020] Construct a waveform graph in a two-dimensional plane based on the starting point of the curve;

[0021] The convex points of the curve in the waveform graph are collected, the convex points are connected in series in sequence, and a waveform curve is formed.

[0022] As a further preferred embodiment of the present technical solution, a method for determining characteristic points of the rotary cutter running in a waveform curve includes:

[0023] Divide the waveform curve into several equally divided waveform segments based on the time series;

[0024] Obtaining the area of ​​a region formed between each waveform line segment and the abscissa of the waveform curve, wherein the area of ​​the region is used to represent the vibration amplitude of the rotary cutter;

[0025] The area differences between adjacent waveform segments are compared to identify the characteristic points of the rotary cutter operation.

[0026] As a further preferred embodiment of the present technical solution, the operating state of the rotary cutter includes: a vibration state, an angle state and a blade position state of the rotary cutter. The vibration state reflects the vibration amplitude and frequency of the rotary cutter during operation, the angle state involves the angle change when the rotary cutter contacts the wood, and the blade position state is directly related to the accuracy of the rotary cutter in cutting the wood.

[0027] As a further preferred embodiment of the present technical solution, the method for simulating the operating state of the rotary cutter of the rotary cutter using the threshold parameters and the initial parameters includes:

[0028] A simulation model of a rotary cutter is established based on computer simulation software and parameter values ​​of the rotary cutter, wherein the parameter values ​​of the rotary cutter include threshold parameters and initial parameters;

[0029] Set the physical parameters in the simulation model and record the motion trajectory and force images of the rotary cutter under different parameters;

[0030] The simulation results were analyzed to extract the simulation data of rotary cutter vibration, angle variation and blade position.

[0031] As a further preferred embodiment of the present technical solution, the method for establishing the simulation model includes:

[0032] A basic framework of physical simulation is constructed based on the initial parameters, and the basic framework is designed to store the physical parameters and working environment conditions of the peeling machine;

[0033] Capture the characteristic types of the rotary cutter during its motion, and use 3D modeling software to create a 3D model that is compatible with the characteristic types. The characteristic types include: the motion trajectory of the rotary cutter, the heat source effect, and the position information.

[0034] Using image acquisition technology, motion images within a working cycle are obtained from multiple three-dimensional models and output as continuous simulation images. The simulation images are displayed side by side on a unified display interface by a series of motion images.

[0035] As a further preferred embodiment of the present technical solution, a method for predicting a potential fault in the position of the rotary cutter in the rotary cutter based on the analysis results includes:

[0036] According to the acquisition specifications of the image data, a selection box including an n×n pixel acquisition box is constructed, and the selection box is used to identify the heat source color point;

[0037] Compare the threshold parameters, initial parameters and the heat source color points in the selection box in real-time state;

[0038] According to the comparison result, the heat source distribution at the rotary cutter position in real time is determined;

[0039] If the distribution of the heat source color points detected is significantly different from the distribution of the heat source color points under the threshold parameters or initial parameters, it is determined to be a potential fault. If there is no obvious difference, it is considered that the rotary cutter position is in a normal state.

[0040] As a further preferred embodiment of the present technical solution, a method for determining a potential fault includes:

[0041] Capturing the first heat source color point, the first heat source color point is a pixel point with the highest temperature chromaticity in the selection box in real time;

[0042] Capturing the second heat source color point, the second heat source color point is a pixel point with the lowest temperature chromaticity in the selection box in real time;

[0043] Calculate the average temperature chromaticity value of the heat source color point in the selection box based on the temperature chromaticity difference between the first heat source color point and the second heat source color point;

[0044] Compare the average temperature chromaticity value in real-time state with the average temperature chromaticity value under initial parameters to evaluate the change trend of heat source at the rotary cutter position;

[0045] If the average temperature chromaticity value exceeds a preset range, it is determined as a potential fault, and the preset range is obtained based on the average temperature chromaticity value under the initial parameters and the average temperature chromaticity value under the threshold parameters.

[0046] In the second aspect, in order to improve the above technical solution, the present invention also proposes an intelligent fault prediction system for a rotary cutter based on multi-sensor analysis. It should be noted that the intelligent fault prediction system for a rotary cutter based on multi-sensor analysis uses the above-mentioned intelligent fault prediction method for a rotary cutter based on multi-sensor analysis, and includes:

[0047] An audio sensor is used to collect audio signals generated by the rotary cutter during the wood peeling process;

[0048] A visual sensor is used to capture image data with time stamps corresponding to feature points when the rotary cutter is running;

[0049] Thermal imaging sensor, used to obtain temperature change information of the rotary cutter assembly, i.e., heat source distribution data;

[0050] A data acquisition module is used to collect real-time status information collected by audio sensors, visual sensors, and thermal imaging sensors;

[0051] A signal processing module is used to process the collected audio data, map it into a waveform diagram and form a waveform curve, and pre-process the image data and heat source distribution data, wherein the pre-processing includes filtering and denoising;

[0052] A feature extraction module is used to extract the feature points of the rotary cutter operation from the waveform curve and to extract the feature information of the rotary cutter vibration image, angle change image and blade position image from the image data;

[0053] The state simulation module establishes a simulation model of the rotary cutter based on computer simulation software and the numerical values ​​of the rotary cutter parameters;

[0054] The fault prediction module is used to compare and analyze the heat source distribution data in the simulation state and the real-time state, as well as the feature information extracted from the waveform curve and image data.

[0055] As a further preferred embodiment of the present technical solution, the fault prediction module further includes a fault diagnosis database, which is used to store a variety of fault modes and their corresponding characteristic information to facilitate matching and diagnosis of real-time status data. In addition, it also includes a user interaction interface, which is used to display characteristic information and control audio sensors, visual sensors, thermal imaging sensors, data acquisition modules, signal processing modules, feature extraction modules, status simulation modules and fault prediction modules. BRIEF DESCRIPTION OF THE DRAWINGS

[0056] Figure 1 This is a flowchart of the steps of an intelligent fault prediction method for a rotary cutter based on multi-sensor analysis of the present invention;

[0057] Figure 2 It is the waveform diagram proposed in step S200 of the present invention;

[0058] Figure 3 It is an auxiliary diagram of step S300 of the present invention.

[0059] Figure 4 It is an auxiliary diagram of step S801 of the present invention;

[0060] Figure 5 This is a composition diagram of an intelligent fault prediction system for rotary cutting machines based on multi-sensor analysis according to the present invention. DETAILED DESCRIPTION

[0061] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0062] Before understanding the technical solution proposed in the present invention, it should be clear that in the prior art, the rotary cutter part of the rotary cutter includes: a clamping shaft for supporting and driving the rotary cutter to rotate, a rotating tool holder for fixing the rotary cutter and allowing it to move radially, and a set of tool holders for installing the tool holder and ensuring its stability. During the operation of the rotary cutter, various faults may occur due to long-term wear, material unevenness, improper operation and other factors, such as tool wear, excessive spindle vibration, inaccurate tool holder positioning, etc. These will affect the processing accuracy and production efficiency of the rotary cutter. Therefore, it is of great significance to predict and diagnose faults of the rotary cutter and to discover and deal with potential problems in a timely manner to ensure production continuity and product quality.

[0063] like Figure 1 It can be seen that the present invention provides a technical solution: an intelligent fault prediction method for a rotary cutter based on multi-sensor analysis, comprising: steps S100 to S800.

[0064] Step S100: collecting audio data of the rotary cutter position during a working cycle of the rotary cutter.

[0065] It should be clear that in step S100, the audio data covers the audio signal generated by the rotary cutter during the process of rotary cutting wood, wherein the working cycle is defined as the time required for the clamping shaft to rotate one circle.

[0066] It should be added that, in the actual collection process, the audio data is collected through an audio sensor installed outside the rotary cutter near the rotary cutter. In the process of acquiring audio data, the audio sensor captures the sound generated when the rotary cutter contacts the wood through a built-in high-sensitivity microphone. These audio signals are then transmitted to the data acquisition module for preliminary signal processing and analysis. In step S100, the collection of audio data is continuous to ensure that all relevant sound events in the entire working cycle are covered. In addition, in order to improve the accuracy and reliability of the data, the audio sensor is usually calibrated regularly to adapt to the noise level and sound characteristics in different environments. In this way, it can be ensured that the collected audio data can truly reflect the working status of the rotary cutter and provide high-quality input information for subsequent fault prediction and diagnosis.

[0067] Furthermore, it should be added that, in the present invention, since the specific model of the peeling machine changes with the specifications of the processed wood, the audio sensor, visual sensor and thermal imaging sensor mentioned in this application are not specifically limited.

[0068] Step S200: Map the audio data into a waveform diagram and form a waveform curve.

[0069] It should be noted that, in step S200 , the horizontal axis of the waveform graph represents time, and the vertical axis represents amplitude.

[0070] It should be supplemented that the method of mapping the audio data into the waveform diagram in step S200 includes: steps S201 to S204.

[0071] Specifically, step S201: acquiring acquisition points based on the frequency and amplitude of audio data.

[0072] It should be noted that in step S201, the frequency and amplitude of the audio data are analyzed by means of a fast Fourier transform (FFT) algorithm. The Fourier transform (FFT) algorithm can convert a time domain signal into a frequency domain signal, thereby identifying the frequency components in the audio signal. Since the Fourier transform (FFT) algorithm is a common signal processing method in the prior art, its specific principle will not be described herein.

[0073] In addition, it should be explained that for the welding wire, the acquisition points in step S201 are specifically the points with the highest and lowest frequencies and the largest and smallest amplitudes.

[0074] Step S202: Arrange the acquisition points in time series to form a continuous curve.

[0075] It should be noted that, in the present invention, step S202 is used to form the continuity of the waveform curve, ensuring that the waveform can accurately reflect the trend of the audio data changing over time.

[0076] Step S203: construct a waveform graph in a two-dimensional plane based on the starting point of the curve.

[0077] Step S204: collecting convex points of the curve in the waveform graph, connecting the convex points in series in sequence, and forming a waveform curve.

[0078] refer to Figure 2 It can be seen that in the present invention, the convex points of the waveform curve refer to the local maximum points on the curve, which appear as peaks in the waveform graph. By connecting these convex points, a continuous and representative waveform curve is obtained, which can intuitively show the characteristics of the audio signal.

[0079] Step S300: extracting characteristic points of the rotary cutter operation from the waveform curve.

[0080] It should be noted that, in step S300, the characteristic points in the waveform curve include crests and troughs. It should be supplemented that the method for determining the characteristic points of the rotary cutter running in the waveform curve includes steps S301 to S303.

[0081] Step 301: Segment the waveform curve into a number of equally divided waveform line segments based on the time series.

[0082] Step S302: Obtain the area of ​​a region formed between each waveform line segment and the abscissa of the waveform curve, where the area of ​​the region is used to represent the vibration amplitude of the rotary cutter.

[0083] Step S303: comparing the area differences between adjacent waveform line segments to identify the characteristic points of the rotary cutter operation.

[0084] refer to Figure 3 It can be seen that it should be supplemented that in step S303, the determination of the peak and trough feature points is determined by obtaining the maximum and minimum values ​​within the area of ​​the region, where the waveform segment corresponding to the maximum value of the area is the peak, and the waveform segment corresponding to the minimum value of the area is the trough.

[0085] Step S400: Based on visual imaging technology, image data with timestamps corresponding to feature points when the rotary cutter is running is captured.

[0086] It should be noted that, in the present invention, the image data in step S400 includes the vibration image of the rotary cutter, the angle change image and the blade position image. In addition, it should be supplemented that the visual imaging technology in step S400 is captured by a visual sensor. In the prior art, the visual sensor is a high-speed camera, which is set at a key position of the rotary cutter to ensure that the various dynamic changes of the rotary cutter during operation can be accurately recorded. Through these image data, the operating state of the rotary cutter can be further analyzed, thereby providing intuitive visual evidence for fault prediction. In step S400, by analyzing the image data, abnormal vibration of the rotary cutter, abnormal change of the angle and deviation of the blade position can be identified.

[0087] Step S500: Based on thermal imaging technology, heat source distribution data in the image data is obtained.

[0088] It should be noted that in step S500, the heat source distribution data reflects the temperature change of the rotary cutter assembly. It should be noted that in the present invention, the thermal imaging technology corresponds to a thermal imaging sensor, which is installed in the rotary cutter near the high-speed camera to monitor and record the temperature change of the rotary cutter assembly during operation. The data obtained through thermal imaging technology can be used to monitor the thermal state of the rotary cutter assembly in real time, thereby detecting abnormal conditions such as overheating that may lead to failures. In step S500, by analyzing the heat source distribution data, the abnormal temperature increase of the rotary cutter assembly can be identified, which may indicate potential overheating problems or wear. Combined with the image data captured in step S400, the operating condition of the rotary cutter can be more comprehensively evaluated. The application of thermal imaging technology is not limited to monitoring temperature, but can also assist in determining the degree of wear of the rotary cutter, because wear usually leads to an increase in local temperature. In addition, the analysis results of the thermal imaging data can be combined with the analysis results of the peak and trough feature points in step S303 to provide a more accurate basis for the prediction of rotary cutter failures. Through this multi-dimensional data analysis, comprehensive monitoring of the operating status of the rotary cutter can be achieved, thereby effectively preventing the occurrence of failures and improving the stability of the equipment and production efficiency.

[0089] Step S600: obtaining the threshold parameters and initial parameters of the rotary cutter of the rotary cutter.

[0090] It should be noted that in step S600, the threshold parameters and initial parameters of the rotary cutter blade are obtained through the product nameplate of the rotary cutter. The specific parameters include the rated speed, rated power, and operating temperature range of the rotary cutter. In actual applications, the real-time data collected by the thermal imaging sensor, audio sensor, and force sensor are compared with these threshold parameters to promptly detect whether the rotary cutter exceeds the normal operating range, so as to take corresponding maintenance measures. In addition, the initial parameters are the performance indicators of the rotary cutter in a new state, which provide a benchmark for subsequent performance degradation analysis and fault prediction. By comparing the initial parameters and real-time monitoring data, the wear degree and remaining service life of the rotary cutter assembly can be evaluated, providing a scientific basis for equipment maintenance and replacement.

[0091] Step S700: simulating the operating state of the rotary cutter of the rotary cutter by using the threshold parameters and the initial parameters, and collecting the heat source distribution data based on the thermal imaging technology under the threshold parameter state and the initial parameter state.

[0092] It should be clear that in step S700, the operating state of the rotary cutter includes: the vibration state, angle state and blade position state of the rotary cutter. The vibration state reflects the vibration amplitude and frequency of the rotary cutter during operation. The angle state involves the angle change when the rotary cutter contacts the wood. The blade position state is directly related to the accuracy of the rotary cutter in cutting the wood.

[0093] Furthermore, it should be supplemented that the method for simulating the operating state of the rotary cutter of the rotary cutter using the threshold parameters and the initial parameters includes: Step S701 to Step S703.

[0094] Step S701: Based on computer simulation software and the parameter values ​​of the peeling machine, a simulation model of the peeling machine blade is established.

[0095] It should be noted that in step S701, the parameter values ​​of the rotary cutter include threshold parameters and initial parameters. In addition, the simulation model includes: a three-dimensional rotary cutter model, a kinematic simulation model, and a vibration and mechanical simulation model. Based on the geometric characteristics of the rotary cutter, a three-dimensional model of the rotary cutter is constructed using three-dimensional modeling software (such as Blender, AutoCAD, and SolidWorks). The three-dimensional rotary cutter model should be able to accurately reflect the shape, size, blade position and other information of the rotary cutter, and provide a basis for subsequent motion and mechanical simulation. The rotary cutter motion trajectory modeling uses Euler angles or quaternions to describe the posture changes and motion trajectory of the rotary cutter, simulating the rotation, tilt and contact state of the rotary cutter with the wood. The vibration and mechanical simulation model focuses on analyzing the dynamic response of the rotary cutter under different working conditions, including vibration frequency, amplitude, and generated force and torque.

[0096] It should be added that the models constructed in step S701 are all realized by computer simulation technology in the prior art, which ensures the accuracy and reliability of the models. Through these models, various situations of the rotary cutter in actual work can be simulated, thereby providing a scientific basis for fault prediction. After the model is established, a parametric analysis is performed to determine the key factors affecting the performance of the rotary cutter. The parametric analysis includes adjusting the geometric parameters, material properties, working environment, etc. of the rotary cutter, observing the impact of these changes on the performance of the rotary cutter, and then optimizing the rotary cutter design and working parameters. In this way, potential fault points can be discovered in advance, providing data support for the maintenance and fault prevention of the rotary cutter.

[0097] It should be further emphasized that, since computer simulation technology is a mature technology at this stage, its application in simulating complex mechanical systems has been very extensive. Therefore, the present invention does not make specific limitations on the construction methods of the three-dimensional rotary cutter model, kinematic simulation model, and vibration and mechanical simulation model.

[0098] Furthermore, it should be supplemented that the method for establishing the simulation model in step S701 includes: steps S7011 to S7013.

[0099] Step S7011: constructing a basic framework of physical simulation based on initial parameters, wherein the basic framework is intended to store the physical parameters and working environment conditions of the rotary cutting machine;

[0100] Step S7012: Capturing the characteristic types of the rotary cutter during the motion process, and using 3D modeling software to create a 3D model that is compatible with the characteristic types, the characteristic types including: the motion trajectory of the rotary cutter, the heat source effect, and the position information;

[0101] Step S7013: Using image acquisition technology, obtain motion images within a working cycle from multiple three-dimensional models, and output them as continuous simulation images. The simulation images are displayed side by side on a unified display interface by a series of motion images.

[0102] Step S702: setting the physical parameters in the simulation model, and recording the motion trajectory images and force images of the rotary cutter under different parameters.

[0103] It should be noted that in step S702, the setting of physical parameters is based on the simulation of the actual working environment and material properties of the rotary cutter, including but not limited to temperature, humidity, hardness and density of wood. By precisely controlling these parameters, the actual performance of the rotary cutter under different working conditions can be simulated. The recorded motion trajectory images and force condition images will be used for subsequent analysis to ensure the stability and reliability of the rotary cutter under various conditions.

[0104] Step S703: Analyze the simulation results and extract the simulation data of the rotary cutter vibration, angle change and blade position.

[0105] It should be noted that in step S703, by comparing the simulation data with the actually collected heat source distribution data, the accuracy of the simulation model can be verified and necessary adjustments can be made to the model. If there is a large deviation between the simulation data and the actual data, it may be necessary to readjust the parameters of the simulation model to ensure that the simulation results can truly reflect the actual working status of the rotary cutter. In addition, the analysis results of the simulation data can also be used to predict the wear trend and remaining service life of the rotary cutter, providing a scientific basis for the maintenance and replacement of the rotary cutter. Through this analysis method that combines simulation with actual data, the working status of the rotary cutter can be more comprehensively understood, thereby achieving more accurate fault prediction and diagnosis.

[0106] Step S800: Compare and analyze the heat source distribution data, and predict potential faults at the rotary cutter position in the rotary cutting machine based on the analysis results.

[0107] It should be noted that, in the present invention, the method for predicting a potential fault at the position of the rotary cutter in the rotary cutter based on the analysis result includes: steps S801 to S804.

[0108] refer to Figure 4 It can be seen that step S801: construct a selection box including an n×n pixel acquisition box according to the acquisition specification of the image data, and the selection box is used to identify the heat source color point.

[0109] Step S802: compare the threshold parameters, initial parameters and the heat source color points in the selection box in real time.

[0110] Step S803: determining the heat source distribution at the rotary cutter position in real time according to the comparison result;

[0111] Step S804: If the distribution of the heat source color points detected is significantly different from the distribution of the heat source color points under the threshold parameters or the initial parameters, it is determined to be a potential fault. If there is no significant difference, it is considered that the rotary cutter position is in a normal state.

[0112] In addition, it should be noted that the method for determining potential faults in step S804 includes: step S8041-step S8045.

[0113] Step S8041: Capturing a first heat source color point, the first heat source color point is a pixel point with the highest temperature chromaticity in a selection box in real time;

[0114] Step S8042: capturing a second heat source color point, where the second heat source color point is a pixel point with the lowest temperature chromaticity in the selection box in real time;

[0115] Step S8043: Calculating the average temperature chromaticity value of the heat source color points in the selection box based on the temperature chromaticity difference between the first heat source color point and the second heat source color point;

[0116] Step S8044: comparing the average temperature chromaticity value in the real-time state with the average temperature chromaticity value under the initial parameters to evaluate the change trend of the heat source at the rotary cutter position;

[0117] Step S8045: If the average temperature chromaticity value exceeds a preset range, it is determined to be a potential fault, and the preset range is obtained based on the average temperature chromaticity value under the initial parameters and the average temperature chromaticity value under the threshold parameters.

[0118] Second, as Figure 5 As shown, the present invention proposes an intelligent fault prediction system for peeling machine based on multi-sensor analysis. The system mainly includes:

[0119] The audio sensor is responsible for collecting the audio signal generated by the rotary cutter during the wood peeling process;

[0120] A visual sensor is used to capture image data with time stamps corresponding to feature points when the rotary cutter is running;

[0121] Thermal imaging sensor, used to obtain temperature change information of the rotary cutter assembly, i.e., heat source distribution data;

[0122] The data acquisition module is responsible for collecting real-time status information collected by audio sensors, visual sensors, and thermal imaging sensors;

[0123] The signal processing module is used to process the collected audio data, map it into a waveform diagram and form a waveform curve, and pre-process the image data and heat source distribution data, including filtering and denoising;

[0124] A feature extraction module is used to extract the feature points of the rotary cutter operation from the waveform curve, and to extract the feature information of the rotary cutter vibration image, angle change image and blade position image from the image data;

[0125] The state simulation module establishes a simulation model of the rotary cutter based on computer simulation software and the numerical values ​​of the rotary cutter parameters;

[0126] The fault prediction module is used to compare and analyze the heat source distribution data in the simulation state and the real-time state, as well as the feature information extracted from the waveform curve and image data.

[0127] In addition, the fault prediction module of the present invention further includes a fault diagnosis database, which is used to store a variety of fault modes and their corresponding characteristic information to facilitate matching and diagnosis of real-time status data. At the same time, it also includes a user interaction interface, which is used to display characteristic information and control audio sensors, visual sensors, thermal imaging sensors, data acquisition modules, signal processing modules, feature extraction modules, state simulation modules and fault prediction modules.

[0128] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is limited by the attached embodiments and their equivalents.

Claims

1. An intelligent fault prediction method for peeling machine based on multi-sensor analysis, characterized in that: include: Collect audio data of the position of the rotary cutter during a working cycle of the rotary cutter, wherein the audio data covers the audio signal generated by the rotary cutter during the process of rotary cutting the wood, wherein the working cycle is defined as the time required for the clamping shaft to rotate one circle; Mapping the audio data into a waveform graph to form a waveform curve, wherein the horizontal axis of the waveform graph represents time and the vertical axis represents amplitude; Extracting characteristic points of the rotary cutter operation from the waveform curve, wherein the characteristic points include wave crests and wave troughs; Based on visual imaging technology, the image data corresponding to the time stamp of the feature point when the rotary cutter is running is captured, and the image data includes the vibration image, the angle change image and the blade position image of the rotary cutter; Based on thermal imaging technology, the heat source distribution data in the image data is obtained, and the heat source distribution data reflects the temperature change of the rotary cutter assembly; Obtain threshold parameters and initial parameters of the rotary cutter; The operating state of the rotary cutter of the rotary cutter is simulated by using the threshold parameters and the initial parameters, and the heat source distribution data based on the thermal imaging technology under the threshold parameter state and the initial parameter state are collected; The heat source distribution data was compared and analyzed, and potential faults at the rotary cutter position in the rotary cutting machine were predicted based on the analysis results.

2. The intelligent fault prediction method for peeling machine based on multi-sensor analysis according to claim 1 is characterized by: Methods for mapping audio data into waveforms include: Acquiring acquisition points based on the frequency and amplitude of the audio data; Arrange the collected points in time series to form a continuous curve; Construct a waveform graph in a two-dimensional plane based on the starting point of the curve; The convex points of the curve in the waveform graph are collected, the convex points are connected in series in sequence, and a waveform curve is formed.

3. The intelligent fault prediction method for rotary cutting machine based on multi-sensor analysis according to claim 1 is characterized by: The method for determining the characteristic points of the rotary cutter running in the waveform curve includes: Divide the waveform curve into several equally divided waveform segments based on the time series; Obtaining the area of ​​a region formed between each waveform line segment and the abscissa of the waveform curve, wherein the area of ​​the region is used to represent the vibration amplitude of the rotary cutter; The area differences between adjacent waveform segments are compared to identify the characteristic points of the rotary cutter operation.

4. The intelligent fault prediction method for rotary cutting machine based on multi-sensor analysis according to claim 1 is characterized by: The operating state of the rotary cutter includes: the vibration state, angle state and blade position state of the rotary cutter. The vibration state reflects the vibration amplitude and frequency of the rotary cutter during operation, the angle state involves the angle change when the rotary cutter contacts the wood, and the blade position state is directly related to the accuracy of the rotary cutter in cutting the wood.

5. The intelligent fault prediction method for rotary cutting machine based on multi-sensor analysis according to claim 4 is characterized by: The method of simulating the operating state of the rotary cutter of the rotary cutter by using the threshold parameters and the initial parameters includes: A simulation model of a rotary cutter is established based on computer simulation software and parameter values ​​of the rotary cutter, wherein the parameter values ​​of the rotary cutter include threshold parameters and initial parameters; Set the physical parameters in the simulation model and record the motion trajectory and force images of the rotary cutter under different parameters; The simulation results were analyzed to extract the simulation data of rotary cutter vibration, angle variation and blade position.

6. The intelligent fault prediction method for peeling machine based on multi-sensor analysis according to claim 5 is characterized by: The simulation model building method includes: A basic framework of physical simulation is constructed based on the initial parameters, and the basic framework is designed to store the physical parameters and working environment conditions of the peeling machine; Capture the characteristic types of the rotary cutter during its motion, and use 3D modeling software to create a 3D model that is compatible with the characteristic types. The characteristic types include: the motion trajectory of the rotary cutter, the heat source effect, and the position information. Using image acquisition technology, motion images within a working cycle are obtained from multiple three-dimensional models and output as continuous simulation images. The simulation images are displayed side by side on a unified display interface by a series of motion images.

7. The intelligent fault prediction method for peeling machine based on multi-sensor analysis according to claim 1 is characterized by: Methods for predicting potential failures at the rotary cutter position in the rotary cutter based on the analysis results include: According to the acquisition specifications of the image data, a selection box including an n×n pixel acquisition box is constructed, and the selection box is used to identify the heat source color point; Compare the threshold parameters, initial parameters and the heat source color points in the selection box in real-time state; According to the comparison result, the heat source distribution at the rotary cutter position in real time is determined; If the distribution of the heat source color points detected is significantly different from the distribution of the heat source color points under the threshold parameters or initial parameters, it is determined to be a potential fault. If there is no obvious difference, it is considered that the rotary cutter position is in a normal state.

8. The intelligent fault prediction method for peeling machine based on multi-sensor analysis according to claim 7 is characterized by: Methods for determining potential faults include: Capturing the first heat source color point, which is a pixel point with the highest temperature chromaticity in the selection box in real time; Capturing the second heat source color point, the second heat source color point is a pixel point with the lowest temperature chromaticity in the selection box in real time; Calculate the average temperature chromaticity value of the heat source color point in the selection box based on the temperature chromaticity difference between the first heat source color point and the second heat source color point; Compare the average temperature chromaticity value in real time with the average temperature chromaticity value under initial parameters to evaluate the change trend of heat source at the rotary cutter position; If the average temperature chromaticity value exceeds a preset range, it is determined as a potential fault, and the preset range is obtained based on the average temperature chromaticity value under the initial parameters and the average temperature chromaticity value under the threshold parameters.

9. An intelligent fault prediction system for peeling machine based on multi-sensor analysis, using an intelligent fault prediction method for peeling machine based on multi-sensor analysis as claimed in any one of claims 1 to 8, characterized in that: include: An audio sensor is used to collect audio signals generated by the rotary cutter during the wood peeling process; A visual sensor is used to capture image data with time stamps corresponding to feature points when the rotary cutter is running; Thermal imaging sensor, used to obtain temperature change information of the rotary cutter assembly, i.e., heat source distribution data; A data acquisition module is used to collect real-time status information collected by audio sensors, visual sensors, and thermal imaging sensors; A signal processing module is used to process the collected audio data, map it into a waveform diagram and form a waveform curve, and pre-process the image data and heat source distribution data, wherein the pre-processing includes filtering and denoising; A feature extraction module is used to extract the feature points of the rotary cutter operation from the waveform curve and to extract the feature information of the rotary cutter vibration image, angle change image and blade position image from the image data; The state simulation module establishes a simulation model of the rotary cutter based on computer simulation software and the numerical values ​​of the rotary cutter parameters; The fault prediction module is used to compare and analyze the heat source distribution data in the simulation state and the real-time state, as well as the feature information extracted from the waveform curve and image data.

10. The intelligent fault prediction system for rotary cutting machine based on multi-sensor analysis according to claim 9, characterized in that: The fault prediction module further includes a fault diagnosis database, which is used to store multiple fault modes and their corresponding characteristic information to facilitate matching and diagnosis of real-time status data. In addition, it also includes a user interaction interface, which is used to display characteristic information and control audio sensors, visual sensors, thermal imaging sensors, data acquisition modules, signal processing modules, feature extraction modules, status simulation modules and fault prediction modules.

Images