Fault diagnosis system of intelligent bamboo cutting machine

Abstract

The invention provides a fault diagnosis system of an intelligent bamboo cutting machine, which relates to the technical field of mechanical engineering and comprises a feature extraction module used for collecting current and voltage of a motor, a spindle vibration frequency spectrum and dynamic parameters of a quadrilateral monitoring area formed by a spindle bearing, a transmission gear box, a cutter clamping end and a base supporting part in real time, and a current waveform characteristic value, a voltage harmonic component, a vibration frequency domain energy distribution parameter and an abnormal parameter combination in the quadrilateral region are extracted. Through cooperation of multiple modules, parameter acquisition, material-load analysis, dynamic threshold setting, graded early warning and accurate fault diagnosis are realized, and the operation and production continuity of the bamboo cutting machine can be efficiently guaranteed.

Description

Technical Field

[0001] The present invention relates to the technical field of mechanical engineering, in particular to a fault diagnosis system of an intelligent bamboo and wood cutting machine. Background Art

[0002] In the current bamboo and wood processing industry, traditional bamboo and wood cutting machine fault diagnosis technology mainly relies on single sensor monitoring with a fixed threshold, which makes it difficult to accurately identify equipment failures under complex working conditions.

[0003] For example, when cutting bamboo and wood materials of different hardness, the traditional diagnostic system only monitors the motor current through the current sensor and uses a fixed current threshold to determine the operating status of the equipment. When cutting high-hardness bamboo and wood materials, the normal current fluctuations of the motor are mistakenly judged as faults because they exceed the fixed threshold, resulting in frequent triggering of false alarms, which seriously interferes with normal production; and when early faults such as slight wear of the tool occur, the current changes are not obvious and cannot be detected in time. It is not discovered until the fault worsens and affects the cutting quality, which not only delays production progress but also increases maintenance costs. This traditional diagnostic method makes it difficult to establish a dynamic correlation between material properties, equipment operating parameters and faults, and cannot meet the requirements of intelligent bamboo and wood cutting machines for fault diagnosis accuracy and real-time performance. Summary of the Invention

[0004] The technical problem to be solved by the present invention is to provide an intelligent fault diagnosis system for bamboo and wood cutting machines, dynamically adjust the diagnosis standard, and realize accurate fault positioning.

[0005] In order to solve the above technical problems, the technical solutions of the present invention are as follows:

[0006] First, the fault diagnosis system of the intelligent bamboo wood cutting machine includes:

[0007] The feature extraction module is used to collect the motor current, voltage, spindle vibration spectrum, and dynamic parameters of the quadrilateral monitoring area composed of the spindle bearing, transmission gearbox, tool clamping end, and base support in real time, and extract the current waveform characteristic value, voltage harmonic components, vibration frequency domain energy distribution parameters, and abnormal parameter combinations within the quadrilateral area;

[0008] The load mapping module is used to analyze the hardness grade and specification parameters of the current cutting material based on the correlation between the vibration frequency domain energy distribution parameters and the changes in the quadrilateral geometric parameters, combined with the load impact component and the fluctuation value of the diagonal length of the quadrilateral, establish a mapping relationship between the material properties and the dynamic load of the motor, and generate an initial correction value based on the symmetry characteristics of the quadrilateral;

[0009] The dynamic diagnostic standard module is used to match the dynamic threshold calculation rules based on the material hardness grade and the angular displacement deviation within the quadrilateral area, and combines the material cutting resistance coefficient with the abnormal parameter gradient of the quadrilateral monitoring area to generate dynamic diagnostic standards for the current adaptive threshold interval and vibration energy threshold;

[0010] The early warning module analyzes the effective current value, harmonic distortion rate, vibration energy, and angular acceleration within the quadrilateral area based on dynamic diagnostic standards. When the current characteristic value is continuously greater than or equal to the dynamic threshold interval and the vibration energy of any two adjacent detection points in the quadrilateral exceeds the standard, a third-level early warning is triggered;

[0011] The fault diagnosis module is used to distinguish between instantaneous overload and real faults such as mechanical blocking and bearing wear based on a three-level early warning system through dynamic correlation analysis between the motor's electromagnetic torque and changes in quadrilateral area parameters, and generate a diagnostic report.

[0012] Furthermore, the motor current, voltage, and spindle vibration spectrum are collected in real time, as well as the dynamic parameters of the quadrilateral monitoring area composed of the spindle bearing, transmission gearbox, tool clamping end, and base support. The current waveform characteristic values, voltage harmonic components, vibration frequency domain energy distribution parameters, and abnormal parameter combinations within the quadrilateral area are extracted, including:

[0013] Wavelet packet decomposition is used to reduce noise on current, voltage signals, and vibration spectra. The effective value, harmonic distortion rate, and spectrum energy ratio of the current waveform are extracted as characteristic values. The spindle bearing displacement, gearbox meshing clearance, tool clamping end deflection angle, and pressure fluctuation parameters of the base support are collected within the quadrilateral monitoring area.

[0014] According to the radial displacement of the spindle bearing and the pressure fluctuation amplitude of the base support part, the real-time fluctuation value of the diagonal length of the quadrilateral is calculated. Combined with the difference between the axial deflection angle of the tool clamping end and the change rate of the meshing clearance of the transmission gearbox, the inner angular displacement deviation of the quadrilateral is generated.

[0015] The current characteristic value, the tool clamping end deflection angle and the base support pressure fluctuation amplitude, as well as the step diagonal length fluctuation value and the inner angle displacement deviation are combined into an abnormal parameter combination.

[0016] Furthermore, based on the correlation between the energy distribution parameters in the vibration frequency domain and the changes in the geometric parameters of the quadrilateral, combined with the load impact component and the fluctuation value of the diagonal length of the quadrilateral, the hardness grade and specification parameters of the current cutting material are analyzed, and a mapping relationship between the material properties and the dynamic load of the motor is established. The initial correction value based on the symmetry characteristics of the quadrilateral is generated, including:

[0017] The vibration frequency domain energy distribution parameters are divided into low-frequency, medium-frequency, and high-frequency bands. The energy proportion of each frequency band and the energy gradient change value in adjacent time windows are calculated. Based on the fluctuation value of the diagonal length of the quadrilateral, the maximum fluctuation amplitude and fluctuation frequency in the periodic change characteristics are extracted.

[0018] Determine the load impact component caused by the sudden change in material hardness based on the energy gradient change value and the maximum fluctuation amplitude in the high-frequency band;

[0019] Based on the axial deflection angle of the tool clamping end and the change rate of the meshing clearance of the transmission gearbox, combined with the fluctuation value of the diagonal length of the quadrilateral, the quadrilateral symmetry attenuation is calculated, and the ratio of the load impact component to the quadrilateral symmetry attenuation is used as the initial correction value.

[0020] Furthermore, according to the material hardness grade and the angular displacement deviation in the quadrilateral area, the dynamic threshold calculation rules are matched, and the material cutting resistance coefficient and the abnormal parameter gradient of the quadrilateral monitoring area are combined to generate dynamic diagnostic standards for the current adaptive threshold interval and vibration energy threshold, including:

[0021] According to the load impact component and the initial correction value, a mapping relationship table between the material hardness grade and the dynamic load of the motor is established;

[0022] Based on the mapping relationship table between material hardness grade and motor dynamic load, combined with the load impact component and initial correction value, the cutting resistance coefficient corresponding to the material hardness grade is generated;

[0023] According to the abnormal parameter gradient in the quadrilateral area, a linear mapping relationship between the cutting resistance coefficient and the abnormal parameter gradient is established, and the linear mapping coefficient is normalized by the initial correction value to obtain the normalized linear mapping coefficient;

[0024] Based on the fluctuation value of the diagonal length of the quadrilateral, the diagonal length change rate is calculated in real time and integrated with the normalized linear mapping coefficient to generate a dynamic adjustment factor;

[0025] According to the normalized linear mapping coefficient and the dynamic adjustment factor, the preset reference current threshold interval is adaptively scaled in sections to generate the current adaptive threshold interval;

[0026] Based on the vibration frequency domain energy distribution parameters, combined with the high-frequency energy proportion and dynamic adjustment factor, the maximum energy gradient value is calculated, and the vibration energy exceeding threshold is set according to the maximum energy gradient value;

[0027] The current adaptive threshold interval and the vibration energy exceeding threshold are combined into a dynamic diagnosis standard.

[0028] Furthermore, according to the normalized linear mapping coefficient and the dynamic adjustment factor, the preset reference current threshold interval is adaptively scaled in sections to generate the current adaptive threshold interval, including:

[0029] Based on the normalized linear mapping coefficient and the initial correction value, the preset reference current threshold interval is initially segmented and adjusted to generate a preliminary scaling threshold interval;

[0030] Based on the dynamic adjustment factor and the real-time matching relationship between the rate of change of the diagonal length of the quadrilateral and the material hardness grade, the boundary of the initial scaling threshold interval is dynamically adjusted. That is, when the dynamic adjustment factor is greater than 1, the upper limit value of the current threshold interval is proportionally expanded, and the upper limit adjustment range is the product of the dynamic adjustment factor and the initial correction value; when the dynamic adjustment factor is less than 1, the lower limit value of the current threshold interval is proportionally reduced, and the lower limit adjustment range is the absolute value of the difference between the normalized linear mapping coefficient and the dynamic adjustment factor;

[0031] According to the linear mapping relationship between the cutting resistance coefficient and the abnormal parameter gradient, combined with the dynamic boundary adjustment results, the current threshold interval is segmented and corrected to generate the current adaptive threshold interval matching the material hardness grade and quadrilateral deformation state.

[0032] Furthermore, based on dynamic diagnostic criteria, the current effective value, harmonic distortion rate, vibration energy, and angular acceleration within the quadrilateral area are analyzed. When the current characteristic value is continuously ≥ the dynamic threshold interval and the vibration energy of any two adjacent detection points in the quadrilateral exceeds the standard, a three-level warning is triggered, including:

[0033] Based on dynamic diagnostic standards, real-time monitoring of current effective value, harmonic distortion rate and quadrilateral area angular acceleration;

[0034] When the effective current value is continuously greater than or equal to the current adaptive threshold interval and the harmonic distortion rate increases synchronously, the current characteristic value is determined to be out of limit; the vibration energy values ​​of any two adjacent detection points in the quadrilateral monitoring area are synchronously detected. If the vibration energy is greater than or equal to the vibration energy exceeding the standard threshold, the vibration is determined to be abnormal;

[0035] Based on the current characteristic value exceeding the limit and vibration abnormality, the quadrilateral symmetry attenuation and the initial correction value are extracted to calculate the real-time ratio of the quadrilateral deformation variable to the initial correction value. When the real-time ratio exceeds the preset threshold, the third-level warning is triggered.

[0036] Furthermore, the three-level warning includes:

[0037] Level 1 warning: When the real-time ratio is ≥ 50% of the preset threshold for the first time, it indicates the risk of potential material hardness mutation;

[0038] Level 2 warning: When the real-time ratio is ≥ 80% of the preset threshold and the vibration energy continues to exceed the standard, it indicates the risk of mechanical jamming or bearing wear;

[0039] Level 3 warning: When the real-time ratio is continuously ≥ the preset threshold and the current characteristic value exceeds the limit time ≥ the set threshold, it is determined to be a real fault and the shutdown protection command is triggered.

[0040] Furthermore, based on the three-level warning, the system dynamically correlates the motor's electromagnetic torque with changes in quadrilateral area parameters to distinguish between instantaneous overload and actual faults such as mechanical jamming or bearing wear, and generates a diagnostic report including:

[0041] Based on the three-level warning results, the load impact component and the quadrilateral symmetry attenuation are extracted. The dynamic correlation analysis between the motor's electromagnetic torque and the quadrilateral's diagonal length fluctuations is then used to calculate the distinguishing factor between instantaneous overload and true faults.

[0042] If the instantaneous change rate of the load impact component is greater than or equal to the preset mutation threshold, and the quadrilateral symmetry attenuation returns to the initial fluctuation range within the preset time window after the warning, it is determined to be a transient overload caused by a sudden change in material hardness. If the load impact component change rate is continuously greater than or equal to the steady-state threshold, and the gearbox meshing clearance change rate and the quadrilateral diagonal length fluctuation value simultaneously exceed the historical average fluctuation range, it is determined to be a real fault of mechanical blocking or bearing wear.

[0043] The fault area is located based on the deflection angle of the tool clamping end and the pressure fluctuation amplitude of the base support part. If the deflection angle of the tool clamping end is ≥ the preset safety threshold of the quadrilateral inner angular displacement deviation, and the front side concentration of the pressure fluctuation amplitude of the base support part is ≥ 50% of the dynamic adjustment factor, the fault area is determined to be the tool clamping end. If the change rate of the transmission gearbox meshing clearance and the difference between the pressure fluctuation amplitude on both sides of the base support part is ≥ 1.2 times the vibration energy exceeding the threshold, the fault area is determined to be the transmission gearbox or the spindle bearing.

[0044] Based on the distinguishing factors and the fault occurrence area, a diagnostic report containing the fault type, location information and confidence rating is generated.

[0045] In a second aspect, a computing device includes:

[0046] one or more processors;

[0047] The storage device is used to store one or more programs. When the one or more programs are executed by the one or more processors, the one or more processors implement the system.

[0048] According to a third aspect, a computer-readable storage medium stores a program, which implements the system when executed by a processor.

[0049] The above solution of the present invention includes at least the following beneficial effects:

[0050] The feature extraction module collects multi-dimensional parameters and extracts key features, covering electrical signals and dynamic mechanical structure information, such as motor current and voltage characteristics and quadrilateral area mechanical parameters. This provides a rich and accurate data foundation for fault diagnosis, avoiding the limitations of single parameter monitoring and achieving a comprehensive understanding of the equipment's operating status. The load mapping module analyzes the relationship between vibration and geometric parameters, combines load impact and diagonal fluctuations, accurately analyzes material hardness and specifications, establishes a material-load mapping relationship, and generates initial correction values. This enables the equipment to automatically adjust its operating strategy based on material properties, improving cutting efficiency and quality, and reducing the risk of failure due to improper material adaptation.

[0051] The dynamic diagnostic standard module matches dynamic threshold calculation rules based on parameters such as material hardness and quadrilateral internal angular displacement, generating adaptive thresholds for current and vibration energy. Compared to fixed thresholds, this dynamic standard can be adjusted in real time as operating conditions change, accurately reflecting the normal operating range of the equipment, reducing misjudgments and missed judgments, and improving the accuracy and reliability of fault diagnosis. The early warning module is based on dynamic diagnostic standards and comprehensively analyzes multiple parameters, triggering a three-level early warning when specific conditions are met. The phased early warning mechanism can issue alarms in a timely manner at different stages of fault development, such as a first-level early warning to indicate potential risks, a second-level early warning to indicate potential fault hazards, and a third-level early warning to trigger shutdown protection, allowing operators to take appropriate measures based on the warning level to effectively prevent and control the occurrence and expansion of faults.

[0052] After a three-level warning, the fault diagnosis module accurately distinguishes between transient overload and actual faults through dynamic correlation analysis of the motor's electromagnetic torque and quadrilateral parameters, locates the fault area, and generates a diagnostic report. This process reduces the time and difficulty of manual troubleshooting. Maintenance personnel can quickly develop repair plans based on the report, improving equipment maintenance efficiency, reducing downtime, and ensuring production continuity. BRIEF DESCRIPTION OF THE DRAWINGS

[0053] Figure 1 The figure is a schematic diagram of a fault diagnosis system for an intelligent bamboo wood cutting machine provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0054] Exemplary embodiments of the present disclosure will be described in more detail below with reference to the accompanying drawings. Although exemplary embodiments of the present disclosure are shown in the accompanying drawings, it should be understood that the present disclosure can be implemented in various forms and should not be limited by the embodiments set forth herein. Rather, these embodiments are provided to enable a more thorough understanding of the present disclosure and to fully convey the scope of the present disclosure to those skilled in the art.

[0055] like Figure 1As shown, an embodiment of the present invention provides a fault diagnosis system for an intelligent bamboo wood cutting machine, comprising:

[0056] The feature extraction module is used to collect the motor current, voltage, spindle vibration spectrum, and dynamic parameters of the quadrilateral monitoring area composed of the spindle bearing, transmission gearbox, tool clamping end, and base support in real time, and extract the current waveform characteristic value, voltage harmonic components, vibration frequency domain energy distribution parameters, and abnormal parameter combinations within the quadrilateral area;

[0057] The load mapping module is used to analyze the hardness grade and specification parameters of the current cutting material based on the correlation between the vibration frequency domain energy distribution parameters and the changes in the quadrilateral geometric parameters, combined with the load impact component and the fluctuation value of the diagonal length of the quadrilateral, establish a mapping relationship between the material properties and the dynamic load of the motor, and generate an initial correction value based on the symmetry characteristics of the quadrilateral;

[0058] The dynamic diagnostic standard module is used to match the dynamic threshold calculation rules based on the material hardness grade and the angular displacement deviation within the quadrilateral area, and combines the material cutting resistance coefficient with the abnormal parameter gradient of the quadrilateral monitoring area to generate dynamic diagnostic standards for the current adaptive threshold interval and vibration energy threshold;

[0059] The early warning module analyzes the effective current value, harmonic distortion rate, vibration energy, and angular acceleration within the quadrilateral area based on dynamic diagnostic standards. When the current characteristic value is continuously greater than or equal to the dynamic threshold interval and the vibration energy of any two adjacent detection points in the quadrilateral exceeds the standard, a third-level early warning is triggered;

[0060] The fault diagnosis module is used to distinguish between instantaneous overload and real faults such as mechanical blocking and bearing wear based on a three-level early warning system through dynamic correlation analysis between the motor's electromagnetic torque and changes in quadrilateral area parameters, and generate a diagnostic report.

[0061] In an embodiment of the present invention, the feature extraction module collects multi-source data and extracts a combination of feature parameters, breaking through the limitations of the one-sided monitoring information of a traditional single sensor. For example, in the early stage of tool wear, the vibration spectrum and parameter changes in the quadrilateral area will be captured. Combined with the analysis of current and voltage signals, the fault can be accurately located to avoid missed judgments and misjudgments, thereby improving the diagnostic accuracy compared to traditional methods. The load mapping module establishes a dynamic mapping relationship between materials and loads, and the dynamic diagnostic standard module generates an adaptive threshold based on this. When cutting bamboo and wood materials of different hardness, the diagnostic threshold can be automatically adjusted according to the material characteristics, and normal current fluctuations will not be misjudged as faults, effectively reducing false alarms, ensuring production continuity, and solving the problem that traditional fixed threshold diagnosis is easily disturbed by working conditions.

[0062] The early warning module conducts multi-dimensional joint judgment analysis based on dynamic standards, and triggers a three-level early warning when multiple parameters such as current and vibration meet specific conditions at the same time. Compared with the traditional reliance on a single indicator to judge faults, this method can detect potential faults earlier, allowing operators to take timely measures to reduce losses caused by equipment failures. The fault diagnosis module is based on a three-level early warning. By analyzing the changes in the motor's electromagnetic torque and quadrilateral area parameters, it can accurately distinguish between instantaneous overload, mechanical jamming, bearing wear and other fault types, and provide positioning information and confidence ratings. In the past, technicians may make misjudgments when judging the cause of abnormal spindle noise based on experience. This system can directly locate the fault, help maintenance personnel quickly develop maintenance plans, shorten maintenance time, and improve equipment operation efficiency.

[0063] In a preferred embodiment of the present invention, the motor current, voltage, spindle vibration spectrum, and dynamic parameters of the quadrilateral monitoring area consisting of the spindle bearing, transmission gear box, tool clamping end, and base support are collected in real time, and the current waveform characteristic value, voltage harmonic component, vibration frequency domain energy distribution parameter, and abnormal parameter combination within the quadrilateral area are extracted, which may include:

[0064] Wavelet packet decomposition is used to reduce noise on current, voltage signals, and vibration spectra. The effective value, harmonic distortion rate, and spectrum energy ratio of the current waveform are extracted as characteristic values. The spindle bearing displacement, gearbox meshing clearance, tool clamping end deflection angle, and pressure fluctuation parameters of the base support are collected within the quadrilateral monitoring area.

[0065] According to the radial displacement of the spindle bearing and the pressure fluctuation amplitude of the base support part, the real-time fluctuation value of the diagonal length of the quadrilateral is calculated. Combined with the difference between the axial deflection angle of the tool clamping end and the change rate of the meshing clearance of the transmission gearbox, the inner angular displacement deviation of the quadrilateral is generated.

[0066] The current characteristic value, the tool clamping end deflection angle and the base support pressure fluctuation amplitude, as well as the step diagonal length fluctuation value and the inner angle displacement deviation are combined into an abnormal parameter combination.

[0067] In an embodiment of the present invention, when the bamboo wood cutting machine is in operation, the motor current, voltage signal, and main shaft vibration spectrum are susceptible to external electromagnetic interference, mechanical vibration, and other noise. The use of wavelet packet decomposition technology is like "finely screening" these signals, decomposing the signals into different frequency bands, removing the noise frequency band, and retaining the effective frequency band signal. For example, the high-frequency electromagnetic interference noise generated during the bamboo wood cutting process can be eliminated through wavelet packet decomposition. The effective value of the current waveform is then extracted, which reflects the average working capacity of the current; the harmonic distortion rate is used to measure the degree to which the current waveform deviates from the sine wave; and the spectrum energy ratio reflects the proportion of different frequency components in the signal energy. At the same time, displacement sensors, gap sensors, angle sensors, and pressure sensors are used to respectively collect the main shaft bearing displacement, gearbox meshing gap, tool clamping end deflection angle, and pressure fluctuation parameters of the base support.

[0068] In the quadrilateral monitoring area formed by the spindle bearing, transmission gearbox, tool clamping end and base support part, the radial displacement of the spindle bearing and the pressure fluctuation amplitude of the base support part will cause the shape of the quadrilateral to change. By calculating the real-time fluctuation value of the diagonal length of the quadrilateral caused by the change of these two parameters, the dimensional changes of the quadrilateral in the horizontal and vertical directions can be quantified. For example, when the spindle bearing is worn, the radial displacement increases, which causes the corresponding diagonal length to change. At the same time, the difference between the axial deflection angle of the tool clamping end and the rate of change of the meshing clearance of the transmission gearbox reflects the change in the internal angle of the quadrilateral, which is converted into the internal angle displacement deviation of the quadrilateral to measure the degree of distortion of the quadrilateral shape.

[0069] The system integrates current characteristic values, tool clamping end deflection angle, base support pressure fluctuation amplitude, diagonal length fluctuation, and internal angular displacement deviation. These parameters comprehensively describe the operating status of the bamboo wood cutting machine from multiple dimensions, including motor electrical signals and mechanical structure deformation, forming an abnormal parameter combination. An abnormality in any of these parameters may indicate a potential equipment failure.

[0070] By performing signal noise reduction processing through wavelet packet decomposition, noise interference is effectively removed, allowing the collected current, voltage, and vibration signals to more accurately reflect the operating status of the equipment. The eigenvalues ​​extracted based on accurate signals provide a reliable data foundation for subsequent fault diagnosis, avoiding misjudgments due to noise and improving the accuracy of fault diagnosis. Comprehensive collection and in-depth analysis of the dynamic parameters of the quadrilateral monitoring area can capture subtle structural changes in the equipment. For example, even a slight change in the deflection angle of the tool clamping end or a slight anomaly in the gearbox meshing clearance can be reflected through the calculation of relevant parameters. These early abnormal signals can help the system detect potential faults earlier, improve the sensitivity of fault monitoring, and buy time for timely maintenance measures.

[0071] By combining the current characteristic values ​​with the mechanical parameters of the quadrilateral monitoring area to form an abnormal parameter combination, the device status is assessed from two key dimensions: electrical signals and mechanical structure. This multi-dimensional analysis approach provides a more comprehensive understanding of the device's operating conditions and accurately determines the type and location of the fault. For example, when the current characteristic values ​​are abnormal and the angular displacement deviation within the quadrilateral also exceeds the normal range, it can be more accurately determined that the fault is related to both mechanical structure deformation and abnormal motor load.

[0072] In a preferred embodiment of the present invention, based on the correlation between the vibration frequency domain energy distribution parameters and the changes in the quadrilateral geometric parameters, combined with the load impact component and the diagonal length fluctuation value of the quadrilateral, the hardness grade and specification parameters of the current cutting material are analyzed, a mapping relationship between the material properties and the dynamic load of the motor is established, and an initial correction value based on the quadrilateral symmetry characteristics is generated, which may include:

[0073] The vibration frequency domain energy distribution parameters are divided into low-frequency, medium-frequency, and high-frequency bands. The energy proportion of each frequency band and the energy gradient change value in adjacent time windows are calculated. Based on the fluctuation value of the diagonal length of the quadrilateral, the maximum fluctuation amplitude and fluctuation frequency in the periodic change characteristics are extracted.

[0074] Determine the load impact component caused by the sudden change in material hardness based on the energy gradient change value and the maximum fluctuation amplitude in the high-frequency band;

[0075] Based on the axial deflection angle of the tool clamping end and the change rate of the meshing clearance of the transmission gearbox, combined with the fluctuation value of the diagonal length of the quadrilateral, the quadrilateral symmetry attenuation is calculated, and the ratio of the load impact component to the quadrilateral symmetry attenuation is used as the initial correction value.

[0076] In an embodiment of the present invention, when the bamboo wood cutting machine is running, the vibration spectrum of the main shaft is a complex signal with multiple frequency components superimposed, and the distribution of its energy in different frequency bands contains rich information. The vibration frequency domain energy distribution parameters are divided into low-frequency, medium-frequency, and high-frequency bands, which are distinguished based on the causes and effects of vibrations of different frequencies. Low-frequency vibration is related to the low-frequency resonance of the overall mechanical structure of the equipment and the low-frequency disturbance of the transmission system; medium-frequency vibration originates from the normal working frequency of components such as gear meshing and bearing rotation; high-frequency vibration is more sensitive to changes in the material cutting process. When calculating the energy proportion of each frequency band, the energy size of each frequency band must be determined first. This requires analyzing the vibration spectrum, accumulating the energy values ​​of the corresponding frequency bands in the spectrum, and then dividing it by the total energy to obtain the proportion of the energy of the frequency band in the total energy. For example, if the accumulated value of the low-frequency band energy is A and the total energy is B, then the low-frequency band energy proportion is The calculation of energy gradient change value in adjacent time windows is to compare the energy change of the same frequency band in two adjacent time windows. Assuming that the energy of a frequency band in the previous time window is E1 and the energy of the frequency band in the next time window is E2, and the time interval is Δt, then the energy gradient change value of the frequency band is It reflects how quickly energy changes over time.

[0077] To measure the fluctuations in the diagonal length of the quadrilateral, displacement sensors installed in locations such as the spindle bearings and base supports collect real-time displacement data. As the forces acting on the material change during the cutting process, the quadrilateral structure deforms, causing the diagonal length to change. This collected diagonal length data is analyzed to observe its fluctuations over time, and statistical methods are used to identify periodic variations. Within a complete cycle, the maximum and minimum values ​​of the diagonal length fluctuation are determined, and the difference between the two is the maximum fluctuation amplitude. The frequency of the fluctuation is calculated by counting the number of fluctuation cycles per unit time.

[0078] Load impact component determination process:

[0079] Because high-frequency vibration energy is extremely sensitive to changes in material hardness, when a bamboo saw encounters a sudden change in hardness, such as a bamboo knot, the tool experiences an instantaneous increase in resistance. This change in resistance is quickly transmitted to the spindle, causing the high-frequency vibration to intensify. When analyzing the high-frequency energy gradient and maximum fluctuation amplitude, the first step is to observe the time series trend of the high-frequency energy. If the high-frequency energy gradient rises rapidly at a certain moment, it indicates a sharp increase in high-frequency vibration energy at that moment. Combined with the maximum fluctuation amplitude at that moment, the intensity of the load impact can be comprehensively judged.

[0080] For example, when the high-frequency energy gradient increases significantly over a short period of time, and the maximum fluctuation amplitude exceeds a certain threshold during normal cutting, it indicates a strong load impact. By establishing an empirical relationship model between high-frequency energy changes and load impact, the corresponding load impact magnitude is searched or calculated within the model based on the high-frequency energy gradient change and maximum fluctuation amplitude, thereby determining the load impact component.

[0081] Initial correction value calculation process:

[0082] The axial deflection angle of the tool holder and the rate of change of the gearbox meshing clearance are measured in real time using an angle sensor and a clearance sensor, respectively. During the cutting process, factors such as tool wear and uneven material stress can cause axial deflection of the tool holder. Simultaneously, the meshing state of the gearbox can also change, leading to changes in meshing clearance. These changes can cause deformation of the quadrilateral structure formed by the spindle bearing, gearbox, tool holder, and base support, affecting its symmetry.

[0083] Combined with the fluctuation value of the diagonal length of the quadrilateral, the symmetry attenuation of the quadrilateral is calculated. Specifically, the geometric parameters of the quadrilateral in the ideal symmetrical state, such as angles, side length ratios, etc., are first determined. Then, the actual measured parameters such as the lengths and angles of each side of the quadrilateral are compared with the ideal state parameters. By calculating the degree of difference between the actual parameters and the ideal parameters, the attenuation of the symmetry of the quadrilateral is quantified. For example, the symmetry attenuation is expressed by calculating the sum of the angle differences between the four internal angles of the quadrilateral and the internal angles of the ideal quadrilateral (such as all 90 degrees), or by calculating the sum of the deviation values ​​of the lengths of each side from the ideal side length ratio. Finally, the load impact component is divided by the quadrilateral symmetry attenuation, and the ratio obtained is used as the initial correction value. This calculation process is based on the relationship between the influence of material properties on the load (reflected by the load impact component) and the changes in equipment structure (reflected by the quadrilateral symmetry attenuation), so that the initial correction value can comprehensively reflect the dual influence of material and equipment status under the current cutting conditions.

[0084] By comprehensively analyzing vibration frequency-domain energy and quadrilateral geometric parameters, the system can accurately determine the hardness grade and specifications of the material being cut. For example, bamboo and wood materials of varying hardness produce different vibration spectra and quadrilateral parameter variations when cut. Based on this information, the system can quickly identify the material's characteristics, providing a basis for optimally adjusting cutting process parameters and avoiding poor cutting quality or equipment damage due to material misjudgment. By mapping material characteristics to the motor's dynamic load and combining this with initial correction values, the system can fully account for the impact of material factors on the equipment's operating status when diagnosing faults. When equipment anomalies occur, it can more accurately distinguish between normal load fluctuations caused by material changes and inherent equipment failures, reducing false positives and missed detections and improving the accuracy and reliability of fault diagnosis. The resulting initial correction values ​​can be used to dynamically adjust diagnostic criteria, automatically optimizing diagnostic parameters based on varying material characteristics and equipment operating conditions.

[0085] In a preferred embodiment of the present invention, according to the material hardness grade and the angular displacement deviation in the quadrilateral area, the dynamic threshold calculation rules are matched, and the material cutting resistance coefficient and the abnormal parameter gradient of the quadrilateral monitoring area are combined to generate the dynamic diagnostic criteria of the current adaptive threshold interval and the vibration energy threshold, which may include:

[0086] According to the load impact component and the initial correction value, a mapping relationship table between the material hardness grade and the dynamic load of the motor is established;

[0087] Based on the mapping relationship table between material hardness grade and motor dynamic load, combined with the load impact component and initial correction value, the cutting resistance coefficient corresponding to the material hardness grade is generated;

[0088] According to the abnormal parameter gradient in the quadrilateral area, a linear mapping relationship between the cutting resistance coefficient and the abnormal parameter gradient is established, and the linear mapping coefficient is normalized by the initial correction value to obtain the normalized linear mapping coefficient;

[0089] Based on the fluctuation value of the diagonal length of the quadrilateral, the diagonal length change rate is calculated in real time and integrated with the normalized linear mapping coefficient to generate a dynamic adjustment factor;

[0090] According to the normalized linear mapping coefficient and the dynamic adjustment factor, the preset reference current threshold interval is adaptively scaled in sections to generate the current adaptive threshold interval;

[0091] Specifically, based on the dynamic adjustment factor and the real-time matching relationship between the rate of change of the diagonal length of the quadrilateral and the hardness grade of the material, the initial scaling threshold interval is dynamically adjusted. That is, when the dynamic adjustment factor is greater than 1, the upper limit value of the current threshold interval is proportionally expanded, and the upper limit adjustment range is the product of the dynamic adjustment factor and the initial correction value; when the dynamic adjustment factor is less than 1, the lower limit value of the current threshold interval is proportionally reduced, and the lower limit adjustment range is the absolute value of the difference between the normalized linear mapping coefficient and the dynamic adjustment factor;

[0092] Based on the linear mapping relationship between the cutting resistance coefficient and the abnormal parameter gradient, combined with the dynamic boundary adjustment results, the current threshold interval is segmented and modified to generate a current adaptive threshold interval that matches the material hardness grade and quadrilateral deformation state;

[0093] Based on the vibration frequency domain energy distribution parameters, combined with the high-frequency energy proportion and dynamic adjustment factor, the maximum energy gradient value is calculated, and the vibration energy exceeding threshold is set according to the maximum energy gradient value;

[0094] The current adaptive threshold interval and the vibration energy exceeding threshold are combined into a dynamic diagnosis standard.

[0095] In the embodiments of the present invention, the load impact component is a key indicator for measuring the impact of sudden changes in material hardness on the equipment during the bamboo and wood cutting process. When cutting a sudden increase in hardness, such as a bamboo knot, the force on the cutter increases instantaneously, and the load impact component transmitted to the motor also increases. The initial correction value comprehensively considers the impact of material properties (such as hardness and texture) on the load, as well as parameter fluctuations caused by the equipment structure (geometric changes in the quadrilateral monitoring area).

[0096] The specific calculation process is as follows: First, cutting experiments were conducted in a real-world production environment using bamboo and wood materials of varying hardness grades. During each cutting process, sensors were used to collect the load impact component and initial correction value in real time. For example, bamboo and wood materials were divided into hardness grades 1-5, from low to high. When cutting grade 1 material, the load impact component value (assuming X1) and initial correction value (assuming Y1), along with the corresponding dynamic load parameters such as motor current and voltage, were recorded. When cutting grade 2 material, X2, Y2, and the corresponding dynamic load parameters were recorded, and so on. After collecting a large amount of experimental data, statistical and analytical methods were used. The load impact component can be divided into several intervals (such as 0-10, 10-20, etc.), and the initial correction value is also divided into corresponding intervals. The average or typical value of the material hardness grade and motor dynamic load corresponding to each interval combination is calculated. Finally, a two-dimensional table is generated, with the load impact component intervals and initial correction value intervals as rows and columns. The table records the corresponding material hardness grade and motor dynamic load information, completing the establishment of a mapping relationship table.

[0097] In actual operation, determining the cutting resistance coefficient corresponding to the material hardness grade requires a data analysis process:

[0098] First, in a laboratory setting, bamboo and wood samples of varying hardness grades were selected and tested using a standard cutting process. During the experiments, a high-precision force sensor was used to collect real-time cutting resistance data, recording the corresponding material hardness grade. To ensure data accuracy and reliability, multiple cutting tests were repeated for each hardness grade, generating multiple sets of resistance data. For bamboo and wood materials with hardness grades 1-5, 10 cutting tests were conducted for each grade. The resulting tool resistance data was statistically analyzed to calculate the average cutting resistance for each hardness grade. For example, for material with hardness grade 1, the 10 cutting resistance data were 12N, 13N, 11N, and so on, with an average of 12.3N. For material with hardness grade 2, the average resistance was 15.7N, and so on. Plotting these hardness grades and the corresponding average cutting resistance data on a graph reveals a near linear relationship. To identify the specific expression for this linear relationship, the least squares method was used to fit the data. This process is similar to finding a straight line on a coordinate graph that minimizes the sum of the distances from all data points to it. Using data processing software (such as Excel's data analysis function or MATLAB), calculations ultimately yield the linear relationship constants a = 3.2 and b = 2.5. This means that the material hardness grade and cutting resistance coefficient satisfy the following relationship: cutting resistance coefficient = a × hardness grade + b.

[0099] In the actual cutting process, after determining that the hardness level of the material being cut is level 3 through the mapping table, substituting the hardness level 3 into the above relationship, the corresponding cutting resistance coefficient can be calculated: 3 × 3.2 + 2.5 = 12.1. This method achieves accurate calculation from material hardness level to cutting resistance coefficient.

[0100] The abnormal parameter gradient of the quadrilateral area reflects the speed of change of various parameters in the area (such as spindle bearing displacement, gearbox meshing clearance, tool clamping end deflection angle, base support pressure fluctuation) over time or space. When calculating, first perform a time series analysis on each parameter. Taking the spindle bearing displacement as an example, at two adjacent time points t1 and t2, the displacements D1 and D2 are obtained. Similarly, the change rates of other parameters are calculated, and then these change rates are comprehensively processed (such as averaging or weighted averaging) to obtain the abnormal parameter gradient in the quadrilateral area. Through the analysis of a large amount of historical data, it is found that there is an approximate linear relationship between the cutting resistance coefficient and the abnormal parameter gradient.

[0101] Prepare a two-dimensional rectangular coordinate system, with the cutting resistance coefficient as the ordinate and the abnormal parameter gradient as the abscissa. Mark each set of historical data collected as a point at the corresponding position in the coordinate system. For example, if the cutting resistance coefficient of a set of data is 12.3 and the abnormal parameter gradient is 0.8, mark a point at the intersection of the ordinate 12.3 and the abscissa 0.8. As more and more points are marked, it can be intuitively observed that these points are not randomly distributed, but are roughly concentrated near a certain straight line, which indicates that there is a linear correlation trend between the cutting resistance coefficient and the abnormal parameter gradient. The core goal of the least squares method is to find a straight line that minimizes the sum of the squares of the vertical distances of all data points to the straight line. In the specific operation, first assume that the equation of the straight line is "cutting resistance coefficient = k × abnormal parameter gradient + c", where k is the slope and c is the intercept, and these two parameters are unknown.

[0102] For each data point in the coordinate system (let its horizontal coordinate be x i , the vertical coordinate is y i, representing the i-th group of abnormal parameter gradients and cutting resistance coefficients), the vertical distance from this point to the hypothetical straight line can be calculated by certain mathematical methods. The vertical distances of all data points to the straight line are squared and summed to obtain a function about k and c. By using the extreme value method in mathematics, the partial derivatives of this function with respect to k and c are calculated, and the partial derivatives are set equal to 0. This will result in a system of equations containing two equations. By solving the system of equations, the specific values ​​of k and c that minimize the sum of the squares of the distances can be calculated. For example, after calculation, it is found that k = 5.2 and c = 1.8, then the linear mapping relationship between the cutting resistance coefficient and the abnormal parameter gradient is determined as "cutting resistance coefficient = 5.2 × abnormal parameter gradient + 1.8".

[0103] Since the numerical ranges of the cutting resistance coefficient and the abnormal parameter gradient vary greatly under different working conditions, in order to make the linear mapping coefficient universal and comparable, the linear mapping coefficient k needs to be normalized with the initial correction value. The normalized linear mapping coefficients obtained in this way can eliminate the influence of dimensional and numerical differences under different working conditions.

[0104] A bamboo cutting machine simulation experiment platform was built, and bamboo and wood materials of different hardness levels were selected for cutting tests, covering typical material types from soft to hard. In each cutting experiment, a high-precision displacement sensor was used to monitor the change in the diagonal length of the quadrilateral in real time. The diagonal lengths L1 and L2 corresponding to adjacent times t1 and t2 were recorded at intervals of 100 milliseconds, and the lengths were calculated according to the formula The rate of change of the diagonal length was calculated. Simultaneously, the normalized linear mapping coefficients were calculated based on the vibration spectrum, current, voltage, and other data collected during the cutting process. Repeated experiments were performed for each material, such as 20 cuts on soft bamboo wood. Each experiment recorded a set of data for the rate of change of the diagonal length and the normalized linear mapping coefficients, ultimately yielding multiple sets of raw data under different working conditions.

[0105] The orthogonal experimental method was used to preliminarily set the value combinations of α and β. The value ranges of α and β were divided into 5 levels (such as 0.1, 0.3, 0.5, 0.7, and 0.9), and 25 different weight combinations were constructed (such as α = 0.1, β = 0.9; α = 0.3, β = 0.7, etc.). The experimental data were divided into 25 groups according to different weight combinations. Each group of data was substituted into the dynamic adjustment factor calculation formula: dynamic adjustment factor α × diagonal length change rate + β × normalized linear mapping coefficient. Based on each group of dynamic adjustment factors, the corresponding current adaptive threshold range and vibration energy threshold were calculated, and these thresholds were applied to simulated fault diagnosis scenarios. In the simulated fault diagnosis scenario, different types of faults were set (such as slight tool wear, loose bearings, and sudden changes in material hardness). Fault diagnosis was performed using thresholds calculated using different weight combinations, and the diagnostic results were recorded (including whether the fault was successfully detected, whether the fault type was accurately judged, and whether there were false positives or missed positives).

[0106] The 25 groups of experimental results were quantitatively scored using diagnostic accuracy, false alarm rate, and missed alarm rate as evaluation indicators. For example, in a certain experiment, 3 faults were correctly diagnosed, 1 fault was missed, and there were no false alarms. The diagnostic accuracy of this group of experiments was 75%. By comparing the scores of different weight combinations, the weight combination with the best diagnostic performance was screened out. The final screened weight combination was applied to the actual production environment of the bamboo and wood cutting machine and a long-term operation test lasting one month was carried out. If deviations in the diagnostic effect were found during the actual verification process (such as a large number of false alarms), the weight coefficient was fine-tuned (such as adjusting α from 0.4 to 0.45) and verified again. After multiple rounds of optimization and verification, β = 0.3 and β = 0.7 were finally determined as the weighting coefficients of the dynamic adjustment factor. This combination can effectively balance the relationship between the quadrilateral structure change information and the cutting resistance parameters under different materials and different working conditions, and achieve precise adjustment of the dynamic threshold.

[0107] The specific analysis of the current adaptive threshold range is as follows:

[0108] Before the bamboo wood cutting machine is put into operation, the reference current threshold range has been set based on the ideal working conditions, such as [I min_b , I max_b ]. When the equipment is started, the dynamic adjustment factor, the rate of change of the diagonal length of the quadrilateral and the material hardness grade are obtained in real time. The dynamic adjustment factor comprehensively reflects the relationship between the equipment structure change and the cutting resistance, and its numerical change intuitively reflects the complexity of the working conditions. The rate of change of the diagonal length of the quadrilateral is calculated by the difference between the diagonal lengths at adjacent moments and the time interval, reflecting the real-time deformation speed of the quadrilateral structure; the material hardness grade is obtained by the correlation analysis of the energy distribution in the vibration frequency domain and the change of the quadrilateral geometric parameters. The system establishes a real-time matching relationship table of the three. For example, when the rate of change of the diagonal length is large and the material hardness grade is high, a higher dynamic adjustment factor corresponds.

[0109] When the dynamic adjustment factor is greater than 1, it indicates that the current operating conditions are severe and the equipment load may exceed normal levels. In this case, the upper limit of the current threshold range needs to be increased. The initial correction value is used as the basis for proportional adjustment, and the upper limit adjustment range is the product of the dynamic adjustment factor and the initial correction value. For example, if the upper limit of the base current threshold range is 100A, the dynamic adjustment factor is 1.2, and the initial correction value is 0.5, the adjusted upper limit is 100 + 1.2 × 0.5 = 100.6A.

[0110] When the dynamic adjustment factor is less than 1, it means that the working condition is relatively stable and the equipment load is at a low level, and the lower limit of the current threshold interval needs to be reduced. The lower limit adjustment range is the absolute value of the difference between the normalized linear mapping coefficient and the dynamic adjustment factor. Because the normalized linear mapping coefficient reflects the gradient relationship between the cutting resistance and the abnormal parameter, adjustment based on this can accurately match the current load. For example, if the normalized linear mapping coefficient is 0.8, the dynamic adjustment factor is 0.6, and the lower limit of the reference current threshold interval is 20A, then the adjusted lower limit value is 20-|0.8-0.6|=19.8A. The threshold interval [I min_1 , I max_1 ].

[0111] There is a linear mapping relationship between the cutting resistance coefficient and the abnormal parameter gradient, which is obtained by fitting a large amount of historical data using the least squares method. According to the current cutting resistance coefficient and the abnormal parameter gradient, the corresponding working condition is located in the established linear relationship. Combined with the dynamic boundary adjustment results [I min_1 , I max_1 ], the current threshold interval is segmented. For example, the cutting process is divided into three stages: light load, medium load, and heavy load. Each stage corresponds to a different cutting resistance coefficient and abnormal parameter gradient range. In the light load stage, if the cutting resistance coefficient is small and the abnormal parameter gradient is stable, the lower limit of the threshold interval is fine-tuned to make it more in line with the actual operating current; in the heavy load stage, if the cutting resistance coefficient is large and the abnormal parameter gradient fluctuates greatly, the upper limit of the threshold interval is further expanded. Through this segmented correction method, the final generated current adaptive threshold interval [I min_a , I max_a ]Precisely matches the current material hardness grade and quadrilateral deformation state.

[0112] Analyze the energy distribution parameters of the vibration frequency domain, divide the vibration spectrum into low frequency, medium frequency and high frequency bands, and calculate the proportion of high frequency band energy to the total energy. For example, the high frequency band energy is E h , the total energy is E t , Combined with the dynamic adjustment factor, consider the varying characteristics of the equipment's vibration energy under different operating conditions. A large dynamic adjustment factor indicates poor operating conditions and potentially increased vibration, requiring a stricter threshold for exceeding the vibration energy limit. By analyzing historical data on vibration energy variations under different operating conditions, we can identify the relationship between the high-frequency energy percentage, the dynamic adjustment factor, and the maximum energy gradient.

[0113] During the normal operation of the bamboo and wood cutting machine, a variety of data are collected through multi-source sensors. For the energy distribution in the vibration frequency domain, it is divided into low-frequency, medium-frequency, and high-frequency bands, and the energy value of each frequency band is monitored in real time. For example, the energy proportion of the high-frequency band is obtained by calculating the ratio of the energy of the high-frequency band to the total energy. At the same time, according to the previous method of calculating the dynamic adjustment factor, the value of the dynamic adjustment factor is continuously obtained. Under different cutting conditions (such as cutting bamboo and wood materials of different hardness, different degrees of tool wear, etc.), the data of the high-frequency band energy proportion and the dynamic adjustment factor are recorded. Multiple experiments are carried out under each working condition. For example, when cutting harder bamboo and wood materials, 20 cutting experiments are carried out, and the energy proportion of the high-frequency band and the dynamic adjustment factor are recorded each time.

[0114] Preliminary setting of the value range of γ and δ:

[0115] To determine the values ​​of γ and δ, we first preliminarily set their ranges. Assume that the range of γ is [0.1, 0.9], the range of δ is [0.1, 0.9], and that γ + δ = 1. Within this range, we select several different combinations, such as (γ = 0.1, δ = 0.9) and (γ = 0.3, δ = 0.7), for a total of 10 different γ and δ combinations. For each γ and δ combination, we calculate the collected experimental data using the formula: "Maximum energy gradient = γ × high-frequency energy fraction + δ × dynamic adjustment factor." For a specific cutting experiment, for example, if the high-frequency energy fraction is 0.4 and the dynamic adjustment factor is 0.6, when γ = 0.3 and δ = 0.7, the maximum energy gradient = 0.3 × 0.4 + 0.7 × 0.6 = 0.12 + 0.42 = 0.54.

[0116] After calculating the maximum energy gradient values ​​for different γ and δ combinations, an evaluation is conducted based on actual equipment operating conditions and fault conditions. For example, observe whether the equipment experiences abnormal vibration or faults when the maximum energy gradient exceeds a certain empirical value. If, for a certain γ and δ combination, the maximum energy gradient exceeds the empirical value, the equipment actually experiences abnormal vibration, this combination effectively reflects the equipment's operating status. Through multiple experiments and evaluation of different combinations, determine the γ and δ values ​​that accurately reflect the equipment's operating status. Assuming that evaluation shows that the maximum energy gradient value when γ = 0.4 and δ = 0.6 more accurately indicates abnormal equipment vibration, then determine γ = 0.4 and δ = 0.6 as the final coefficients. After determining γ = 0.4 and δ = 0.6, analyze a large amount of experimental data to determine the relationship between the maximum energy gradient value and vibration energy. For example, statistically analyze data from 100 cutting experiments to observe the changes in the maximum energy gradient value and vibration energy. Based on actual experience and equipment safety requirements, determine an empirical value for the maximum energy gradient value. For example, when the maximum energy gradient exceeds 0.8, the corresponding vibration energy value is the vibration energy exceeding threshold. If, during multiple experiments, it is found that when the maximum energy gradient exceeds 0.8, the vibration energy value of the equipment increases significantly and abnormal vibration occurs, then the vibration energy value at this time is set as the vibration energy exceeding threshold, thereby achieving effective monitoring and early warning of abnormal equipment vibration.

[0117] When calculating the current adaptive threshold interval, the dynamic threshold calculation rules are first matched according to the hardness level of the cutting material and the angular displacement deviation within the quadrilateral area. For example, when it is detected that the hardness of the cutting material is high and the angular displacement deviation within the quadrilateral is large, the calculation rules applicable to high-load conditions will be called first. Combining the material cutting resistance coefficient and the abnormal parameter gradient of the quadrilateral monitoring area, the normalized linear mapping coefficient and the dynamic adjustment factor are used to segmentally scale the preset reference current threshold interval. If the current working condition shows that the equipment load is large, the dynamic adjustment factor will increase the upper limit of the reference current threshold interval by 10%-20% and the lower limit by 5%-10%, thereby generating a current adaptive threshold interval that meets the current working condition, such as [I min_a , I max_a ].

[0118] For the vibration energy exceeding threshold, the system first analyzes the vibration frequency domain energy distribution parameters, focusing on the energy proportion of the high frequency band. When the energy proportion of the high frequency band exceeds the normal range (such as more than 30%) and the dynamic adjustment factor shows that the equipment is in an unstable state, the vibration energy exceeding threshold is set by calculating the maximum energy gradient value (obtained by combining the energy proportion of the high frequency band and the dynamic adjustment factor), combined with the equipment's historical operating data and safety standards. For example, if a fault has occurred when the equipment vibration energy exceeds 50 units under similar working conditions, the system will set the vibration energy exceeding threshold to 45 units, leaving a certain safety margin. When the calculated current adaptive threshold interval and the vibration energy exceeding threshold are integrated, a multi-dimensional judgment matrix is ​​created. The matrix is ​​divided into four areas with the current threshold interval as the horizontal axis and the vibration energy threshold as the vertical axis: normal operating area, current abnormal area, vibration abnormal area and fault warning area. The normal operating area requires the current value to be between [I min_a , I max_a ], and the vibration energy value is lower than 45 units; the current abnormal area corresponds to the current value exceeding the threshold range, and the vibration energy is normal; the vibration abnormal area corresponds to the vibration energy exceeding the standard, and the current value is normal; the fault warning area is when the current value exceeds the threshold range and the vibration energy exceeds the standard.

[0119] During operation, sensors collect motor current and vibration energy data in real time at millisecond speeds. For example, the current sensor collects current data 100 times per second, and the vibration sensor collects vibration energy data 50 times per second. This data is immediately transmitted to the diagnostic system for comparison with integrated dynamic diagnostic criteria. The system compares the latest current value with the upper and lower limits of the adaptive current threshold range, and also compares the vibration energy value with the vibration energy exceeding threshold.

[0120] When the data collected at a certain moment shows that the current value is greater than I max_a And when the vibration energy value is higher than 45 units, the system will trigger the fault judgment logic. To avoid misjudgment, the system will continuously monitor multiple data points (such as monitoring data 5 times in a row). If these 5 data meet the conditions that the current and vibration energy exceed the standard at the same time, the system will immediately trigger a three-level warning. At this time, the red alarm light on the cutting machine operation panel will light up, and at the same time, an alarm message containing the fault type (such as overload or mechanical abnormality) and the specific values ​​of the abnormal parameters (such as the current current value, vibration energy value) will be sent to the remote monitoring center to remind the operator that the equipment has an abnormality. If only the current or vibration energy single indicator exceeds the standard, the system will record the abnormal data, continuously observe the operating status of the equipment, and will not trigger a fault warning, so as to distinguish between short-term fluctuations and real faults.

[0121] By mapping material hardness levels to motor dynamic loads and combining multiple parameters to generate dynamic diagnostic criteria, the system can more accurately reflect the operating status of the machine under different materials and operating conditions. This avoids the misjudgments and missed detections that can occur when using fixed thresholds, improving the accuracy of fault diagnosis. Dynamic diagnostic criteria adaptively adjust based on the machine's real-time operating status and the characteristics of the material being cut. Whether cutting bamboo or wood materials of varying hardness or undergoing structural changes, the thresholds can be adjusted promptly to ensure the diagnostic criteria consistently match the actual operating conditions, enhancing the machine's adaptability to diverse operating environments. Timely and accurate diagnosis of equipment anomalies enables operators to take prompt repair and adjustment measures, preventing further escalation of the fault, ensuring stable operation of the bamboo and wood cutting machine and improving production efficiency and product quality.

[0122] In a preferred embodiment of the present invention, based on dynamic diagnostic criteria, the effective value of current, harmonic distortion rate, vibration energy, and angular acceleration within the quadrilateral area are analyzed. When the current characteristic value is continuously greater than or equal to the dynamic threshold interval and the vibration energy of any two adjacent detection points in the quadrilateral exceeds the standard, a three-level warning is triggered, which may include:

[0123] Based on dynamic diagnostic standards, real-time monitoring of current effective value, harmonic distortion rate and quadrilateral area angular acceleration;

[0124] When the effective current value is continuously greater than or equal to the current adaptive threshold interval and the harmonic distortion rate increases synchronously, the current characteristic value is determined to be out of limit; the vibration energy values ​​of any two adjacent detection points in the quadrilateral monitoring area are synchronously detected. If the vibration energy is greater than or equal to the vibration energy exceeding the standard threshold, the vibration is determined to be abnormal;

[0125] Based on the current characteristic value exceeding the limit and the vibration abnormality, the quadrilateral symmetry attenuation and the initial correction value are extracted to calculate the real-time ratio of the quadrilateral deformation to the initial correction value. When the real-time ratio exceeds the preset threshold, a three-level warning is triggered; the three-level warning includes:

[0126] Level 1 warning: When the real-time ratio is ≥ 50% of the preset threshold for the first time, it indicates the risk of potential material hardness mutation;

[0127] Level 2 warning: When the real-time ratio is ≥ 80% of the preset threshold and the vibration energy continues to exceed the standard, it indicates the risk of mechanical jamming or bearing wear;

[0128] Level 3 warning: When the real-time ratio is continuously ≥ the preset threshold and the current characteristic value exceeds the limit time ≥ the set threshold, it is determined to be a real fault and the shutdown protection command is triggered.

[0129] In an embodiment of the present invention, when the bamboo wood cutting machine is running, the high-precision sensor acts as the "sensory organ" of the equipment, continuously collecting key data. The current sensor collects motor current data at a frequency of 10 times per second, and the current value collected each time will be processed inside the sensor. By accumulating the current data over a period of time (such as 0.1 seconds) and dividing it by the number of collections, the effective value of the current is obtained. This value represents the average ability of the motor to do work during that time period. For the monitoring of the harmonic distortion rate, the collected current signal is deeply analyzed. Under normal circumstances, the ideal current waveform is a sine wave, but when the motor load changes, the circuit components are abnormal, etc., the current waveform will be distorted. By comparing the difference between the actual current waveform and the standard sine wave, the changes in the proportion of different frequency components in the current signal are analyzed, and then the harmonic distortion rate is obtained to measure the degree to which the current waveform deviates from the sine wave.

[0130] To monitor angular acceleration within the quadrilateral area, multiple sensors installed on the spindle bearing, transmission gearbox, tool holder, and base support work together. Each sensor acquires real-time motion parameters such as displacement and angle of the corresponding component. For example, the sensor at the spindle bearing monitors changes in bearing rotation angle, while the sensor at the transmission gearbox monitors changes in gear meshing displacement. By integrating these parameters and calculating the change in motion parameters of each component at adjacent moments, combined with the time interval, the change in angular acceleration within the quadrilateral area can be determined. This data is updated at a fixed frequency. The real-time monitored RMS current is carefully compared with the current adaptive threshold range specified in the dynamic diagnostic criteria. The current adaptive threshold range is dynamically adjusted based on various factors, such as the hardness of the cutting material and the operating status of the equipment. If the RMS current value is greater than or equal to the lower limit of the current adaptive threshold range for five consecutive monitoring sessions, it indicates that the motor's current output is consistently high.

[0131] At the same time, pay close attention to the changing trend of the harmonic distortion rate. Within a 10-second time window, record the numerical changes in the harmonic distortion rate. By calculating the difference between the starting value and the ending value of the harmonic distortion rate in this time period, the growth rate is obtained. If this growth rate exceeds the pre-set normal variation range (this range is obtained through statistics of normal equipment operation data), it means that the degree of distortion of the current waveform is abnormally increasing. Only when the two conditions of the effective current value continuously being higher than the lower threshold value and the harmonic distortion rate rising synchronously are met at the same time, will it be determined that the current characteristic value exceeds the limit, which means that the current operating state of the motor is abnormal, and there may be electrical problems such as overload and circuit component failure.

[0132] While monitoring the current, the vibration energy within the quadrilateral monitoring area is detected. This area is composed of the spindle bearing, transmission gearbox, tool clamping end, and base support. Sensors are installed at key locations in these parts to collect vibration energy data. The area is divided into multiple pairs of adjacent detection points (such as the spindle bearing and transmission gearbox, the transmission gearbox and the tool clamping end, etc.). For each pair of adjacent detection points, the sensor collects their vibration energy values ​​in real time. These vibration energy values ​​are then compared with the vibration energy exceeding threshold in the dynamic diagnostic standard. This vibration energy exceeding threshold is determined based on a combination of factors such as the equipment's design parameters, vibration data during normal operation, and the maximum vibration intensity that the equipment can withstand. As long as the vibration energy values ​​of any two adjacent detection points are greater than or equal to the vibration energy exceeding threshold, it means that the vibration intensity in the area has exceeded the normal range, and it is determined that vibration abnormality has occurred in the area. This may be caused by mechanical structural instability factors such as loose components, wear, and improper installation.

[0133] When the two conditions of current characteristic value exceeding the limit and vibration abnormality are met at the same time, the changes in the quadrilateral structure are further analyzed in depth. The quadrilateral symmetry attenuation is calculated by comparing the difference between the current geometric shape of the quadrilateral and the ideal symmetrical shape. Specifically, the four internal angles of the quadrilateral are measured, compared with the internal angles of the ideal quadrilateral (such as all 90 degrees), the angle difference is calculated, and then the four angle differences are added together; or the length of each side of the quadrilateral is measured, compared with the ideal side length, and the deviation value of the ratio of the length of each side to the ideal side length is calculated, and then these deviation values ​​are added together. In these ways, a numerical value that can reflect the degree of change in the symmetry of the quadrilateral is obtained, that is, the quadrilateral symmetry attenuation.

[0134] The initial correction value is a comprehensive parameter generated during analysis of the relationship between material properties and device structural changes. It incorporates information about the impact of various factors, such as material hardness and device load, on the device's operating state. The system divides the quadrilateral symmetry attenuation by the initial correction value. The result is a real-time ratio of the quadrilateral's deformation to the initial correction value. This ratio intuitively reflects the degree of change in the quadrilateral's current structure relative to its initial state.

[0135] Triggering the three-level warning includes:

[0136] Level 1 Warning: The calculated real-time ratio is compared with a preset threshold. This threshold is a critical value determined based on factors such as the equipment's structural strength, design parameters, and long-term operating experience. It represents the safe critical value for changes in the quadrilateral structure. When the real-time ratio reaches 50% of the preset threshold for the first time, the system triggers a Level 1 Warning. This occurs because a sudden change in material hardness can cause a transient change in cutting resistance, which in turn alters the equipment's operating parameters and causes deformation of the quadrilateral structure. The Level 1 Warning alerts the operator to the potential risk of a sudden change in material hardness, allowing them to adjust cutting parameters or take appropriate measures to prevent further damage to the equipment caused by material problems.

[0137] Level 2 Warning: If the real-time ratio continues to rise, reaching 80% of the preset threshold, and the vibration energy remains excessive (i.e., vibration energy is detected to be greater than or equal to the vibration energy excess threshold multiple times in a row), the system triggers a Level 2 warning. Faults such as mechanical jamming or bearing wear can cause increased vibration in the equipment, placing greater stress on the quadrilateral structure and causing significant deformation. The Level 2 warning can more accurately alert operators to potential mechanical failure risks, helping maintenance personnel promptly identify the fault point and prevent further deterioration.

[0138] Level 3 Warning: When the real-time ratio remains above the preset threshold and the current characteristic value exceeds the set threshold for a period of time (e.g., the current characteristic value exceeds the limit for more than 1 minute), it is determined to be a true fault and triggers a Level 3 warning. At this time, a shutdown protection command is immediately issued, stopping the cutting machine. This is to quickly cut off power when the equipment is at risk of serious failure, prevent further escalation of the fault, protect core components from damage, reduce maintenance costs and downtime, and ensure production safety and continuity.

[0139] The first-level warning alerts operators of potential material hardness changes when the real-time ratio reaches a preset threshold of 50%. This allows operators to take proactive action, such as adjusting cutting parameters or replacing cutting tools, to prevent damage to the equipment caused by sudden changes in material hardness and reduce the likelihood of equipment failure. The second-level warning combines the real-time ratio with persistent vibration anomalies to accurately indicate fault risks such as machine seizure or bearing wear. Compared to traditional diagnostic methods, this multi-parameter correlation analysis approach more accurately determines fault type and location, improving diagnostic accuracy and shortening troubleshooting time. The third-level warning's strict triggering conditions ensure that shutdown protection is triggered only when a true fault is confirmed. This avoids unnecessary downtime caused by misdiagnosis and ensures timely shutdown when a serious fault occurs, preventing further damage to the equipment, reducing repair costs, and ensuring safe and continuous production. The entire warning process is based on dynamic diagnostic criteria, performing multi-dimensional analysis of current, vibration, and quadrilateral area parameters, fully utilizing a variety of data and information during equipment operation. This comprehensive diagnostic approach provides a comprehensive picture of the equipment's operating status, overcoming the limitations of single-parameter diagnosis and improving the reliability and effectiveness of the fault diagnosis system.

[0140] In a preferred embodiment of the present invention, based on a three-level warning, dynamic correlation analysis of the motor's electromagnetic torque and changes in quadrilateral area parameters is performed to distinguish between instantaneous overload and actual faults such as mechanical jamming and bearing wear, and generate a diagnostic report that may include:

[0141] Based on the three-level warning results, the load impact component and the quadrilateral symmetry attenuation are extracted. The dynamic correlation analysis between the motor's electromagnetic torque and the quadrilateral's diagonal length fluctuations is then used to calculate the distinguishing factor between instantaneous overload and true faults.

[0142] If the instantaneous change rate of the load impact component is greater than or equal to the preset mutation threshold, and the quadrilateral symmetry attenuation returns to the initial fluctuation range within the preset time window after the warning, it is determined to be a transient overload caused by a sudden change in material hardness. If the load impact component change rate is continuously greater than or equal to the steady-state threshold, and the gearbox meshing clearance change rate and the quadrilateral diagonal length fluctuation value simultaneously exceed the historical average fluctuation range, it is determined to be a real fault of mechanical blocking or bearing wear.

[0143] The fault area is located based on the deflection angle of the tool clamping end and the pressure fluctuation amplitude of the base support part. If the deflection angle of the tool clamping end is ≥ the preset safety threshold of the quadrilateral inner angular displacement deviation, and the front side concentration of the pressure fluctuation amplitude of the base support part is ≥ 50% of the dynamic adjustment factor, the fault area is determined to be the tool clamping end. If the change rate of the transmission gearbox meshing clearance and the difference between the pressure fluctuation amplitude on both sides of the base support part is ≥ 1.2 times the vibration energy exceeding the threshold, the fault area is determined to be the transmission gearbox or the spindle bearing.

[0144] Based on the distinguishing factors and the fault occurrence area, a diagnostic report containing the fault type, location information and confidence rating is generated.

[0145] In an embodiment of the present invention, when the third-level warning is triggered, the historical data of the load impact component and the quadrilateral symmetry attenuation are immediately retrieved from the historical monitoring database. The load impact component is a parameter generated during the cutting process when the hardness of the material suddenly changes (such as encountering bamboo or wood knots), and the equipment is instantly subjected to additional load. Its numerical value reflects the intensity of the impact. The quadrilateral symmetry attenuation is obtained by comparing the current shape of the quadrilateral monitoring area (composed of the spindle bearing, transmission gear box, tool clamping end and base support) with the ideal symmetrical shape. For example, the sum of the deviations of the four internal angles from 90 degrees, or the sum of the proportional deviations of the lengths of each side from the ideal side length, is measured. The larger the value, the more serious the structural deformation.

[0146] At the same time, the motor's electromagnetic torque and the fluctuations in the length of the diagonal lines of the quadrilateral are collected in real time. The motor's electromagnetic torque reflects the changes in the motor's output power, while the fluctuations in the length of the diagonal lines of the quadrilateral reflect the stress and deformation of the equipment's mechanical structure. Within a 10-second time window, the motor's electromagnetic torque is recorded at a fixed frequency (e.g., 10 times per second), with the maximum (peak) and minimum (valley) values ​​marked. The fluctuations in the length of the diagonal lines of the quadrilateral are also recorded at the same frequency, and their peak and valley values ​​are obtained.

[0147] Subsequently, the two sets of data were analyzed. The overall data changes were first observed to determine whether they were relatively stable or fluctuating wildly. The direction of change was then compared to determine whether they increased, decreased, or both. The magnitude of the change in the data points was also calculated, such as the difference between adjacent data points in the motor's electromagnetic torque and the change in the fluctuation value of the diagonal length of a quadrilateral. Using logic similar to that used to calculate the correlation coefficient, the similarity between the two changes was assessed. If the change characteristics of the two across multiple dimensions were highly similar, the resulting discriminant factor value would tend to indicate a transient overload. Conversely, if the difference was significant, the discriminant factor value would tend to indicate a true fault, thereby quantifying the degree of difference between the two.

[0148] The instantaneous overload determination process specifically includes:

[0149] Samples of bamboo and wood materials with different hardness levels were selected, covering various types of materials from soft to hard, and situations where the hardness of the materials suddenly changed were artificially set. For example, normal bamboo and wood materials were combined with special bamboo and wood materials with knots and obvious density differences to simulate the scenario where the hardness of the materials suddenly changed during actual cutting. During each experiment, it was ensured that the bamboo and wood cutting machine was in normal operation, and that key components such as cutters and bearings were free of faults or obvious wear. During the cutting process, high-precision sensors were used to collect load impact component data at an extremely high frequency (such as 100 times per second), and each experiment was repeated multiple times (such as 20 cutting experiments for each material combination) to obtain enough data samples.

[0150] After collecting the load impact component data obtained from each experiment, the data is preliminarily sorted. The data of each experiment are arranged in chronological order, marking the time point when the material hardness mutation occurs, as well as the load impact component value within a period of time before and after the time point (such as 0.5 seconds before the mutation to 2 seconds after the mutation). For each set of experimental data, calculate the rate of change of the load impact component in different time intervals (such as 0.01 seconds, 0.05 seconds, and 0.1 seconds). Taking the 0.1 second interval as an example, for two adjacent data points, subtract the load impact component value of the latter data point from the value of the previous data point, and then divide it by 0.1 seconds to obtain the load impact component change rate within the time interval.

[0151] For all experimental data, the range of load impact component change rates within each time interval under normal cutting conditions (no sudden hardness changes) was calculated. Statistical quantities such as the mean and standard deviation of these change rates were calculated to understand the general pattern and degree of fluctuation of the load impact component under normal conditions. Based on the statistically analyzed load impact component change rate data during normal cutting, combined with actual engineering experience and equipment safety requirements, the upper limit of normal fluctuation was determined.

[0152] Specifically, the judgment is made through the mean value and standard deviation method: the average value of the load impact component change rate under normal circumstances is added with several times the standard deviation (such as 2 times or 3 times the standard deviation) to obtain a value. This value can cover the fluctuation range under most normal circumstances and can be used as a reference for the upper limit of normal fluctuations. For example, if the average value is Q and the standard deviation is S, then the upper limit of normal fluctuations can be set to Q+2S. After determining the upper limit of normal fluctuations, the value is appropriately increased on this basis and set as the preset mutation threshold. The extent of the increase needs to comprehensively consider the equipment's bearing capacity, the risk of misjudgment, and the actual application requirements. If the increase is too small, the load impact component change rate may easily exceed the threshold when the material hardness suddenly changes, resulting in frequent misjudgments; if the increase is too large, the actual material hardness mutation may not be detected in time.

[0153] The preset mutation threshold can be set at 1.1-1.3 times the upper limit of normal fluctuation. For example, if the upper limit of normal fluctuation is Y, the preset mutation threshold can be set to 1.2Y. This threshold setting ensures that when the material hardness suddenly changes, the rate of change of the load impact component has a high probability of exceeding the threshold, triggering the corresponding judgment mechanism, while minimizing misjudgments caused by accidental fluctuations during normal operation, ensuring the accuracy and reliability of instantaneous overload judgments.

[0154] In the actual production scenario of bamboo and wood cutting machines, bamboo and wood materials of different batches and different hardness grades are selected for cutting operations. During each cutting process, load impact component data is collected at a fixed frequency (such as once per second) to fully cover the entire cutting cycle. For each material, cutting experiments are carried out for multiple days and multiple periods. For example, for bamboo and wood materials of a certain hardness grade, cutting operations are carried out for 8 hours every day for 7 consecutive days to obtain a large amount of raw data under different working conditions. During the experiment, other key information of the equipment operation, such as cutting speed, tool wear, ambient temperature, etc., is recorded synchronously. For the collected load impact component data, the load impact component change rate at each time point is calculated in the same way as the instantaneous change rate. That is, the data at adjacent time points are selected, the load impact component value at the next time point is subtracted from the value at the previous time point, and then divided by the time interval (such as 1 second) to obtain the load impact component change rate at that time point.

[0155] For each cutting test data, the load impact component change rate at each time point in the entire cutting process is fully calculated to form a series of continuous change rate data sequences. All experimental data are classified according to different working conditions. For example, according to the hardness of the material, it is divided into soft material group, medium-hard material group, and hard material group; according to the cutting speed, it is divided into low-speed cutting group, medium-speed cutting group, high-speed cutting group, etc. A detailed statistical analysis is performed on the load impact component change rate data under each type of working condition. Calculate the statistical quantities such as the mean, median, and standard deviation of the data, draw the distribution histogram of the data, observe the distribution form of the data, and understand the central trend and discrete degree of the load impact component change rate under different working conditions. Determine the normal load fluctuation range by combining the statistical analysis results under various working conditions. Different methods can be used for different working condition classifications:

[0156] Based on the mean and standard deviation: Under certain operating conditions, add a certain multiple of the standard deviation (e.g., 2 or 3 times the standard deviation) to the mean of the load impact component's rate of change to obtain the upper limit of normal load fluctuation under that condition. Subtract a certain multiple of the standard deviation from the mean to obtain the lower limit. For example, if the mean under certain operating conditions is W and the standard deviation is R, the normal fluctuation range is [A-2R, A+2R].

[0157] Based on practical experience and engineering judgment: Appropriate adjustments are made to the ranges obtained from statistical calculations, taking into account the design parameters of the bamboo and wood cutting machine, long-term operating experience, and the load variations during normal operation of the equipment. For example, considering that the equipment may experience brief load fluctuations during startup and shutdown, data from these special periods can be specially processed to avoid including them in the calculation of the normal load fluctuation range.

[0158] After determining the normal load fluctuation range, the upper limit of this range is selected as a baseline for the steady-state threshold. However, to ensure that the steady-state threshold accurately distinguishes normal load fluctuations from sustained abnormal load changes and to avoid misjudgments due to large normal fluctuations, this upper limit needs to be adjusted appropriately. During this adjustment process, the equipment's safe operation requirements, failure risk tolerance, and actual production needs are comprehensively considered. For example, if the equipment is sensitive to abnormal load fluctuations, the steady-state threshold can be set at 90%-95% of the upper limit of normal load fluctuations to provide early warning of potential failures. If the equipment has a strong overload capacity and wishes to reduce unnecessary warning interference, the steady-state threshold can be set at 105%-110% of the upper limit of normal load fluctuations. The resulting steady-state threshold effectively identifies sustained abnormal load changes while ensuring safe equipment operation, providing a reliable basis for identifying true faults such as mechanical jamming or bearing wear.

[0159] Monitor the rate of change of the transmission gearbox meshing clearance and the fluctuation value of the diagonal length of the quadrilateral. The rate of change of the transmission gearbox meshing clearance is obtained by calculating the ratio of the difference in the meshing clearance between adjacent time points to the time interval; the change in the fluctuation value of the diagonal length of the quadrilateral is measured by comparing the difference between the current value and the historical average value. The historical average fluctuation range is the average value and upper and lower floating range of the fluctuation range of these two parameters when the equipment is operating normally, which is obtained from the statistics of the normal operation data of the equipment accumulated over a long period of time. When the rate of change of the load impact component is continuously greater than or equal to the steady-state threshold, and the rate of change of the transmission gearbox meshing clearance and the fluctuation value of the diagonal length of the quadrilateral simultaneously exceed the historical average fluctuation range, it indicates that the equipment has a continuous abnormal load, and it is determined to be a real fault such as mechanical jamming or bearing wear.

[0160] Fault determination of tool clamping end:

[0161] Obtain the tool clamping end deflection angle and base support part pressure fluctuation amplitude data. The preset safety threshold is set based on the maximum allowable deflection angle of the equipment structure design and the angle fluctuation range during normal operation, and is used to determine whether the deflection of the tool clamping end is within the safe range. When calculating the front side concentration of the base support part pressure fluctuation amplitude, first divide the base support part into the front and rear areas, and count the pressure fluctuation amplitudes of the front and rear sides respectively. Divide the front side pressure fluctuation amplitude by the total pressure fluctuation amplitude to obtain the front side concentration ratio. When the tool clamping end deflection angle ≥ the preset safety threshold, and the base support part pressure fluctuation amplitude front side concentration ≥ 50% of the dynamic adjustment factor, since the problem with the tool clamping end will affect the cutting force distribution, resulting in abnormal pressure concentration on the front side of the base support part, the fault area is determined to be the tool clamping end.

[0162] Transmission gearbox or main shaft bearing fault determination:

[0163] Calculate the gearbox mesh clearance change rate and the difference between the pressure fluctuation amplitudes on both sides of the base support. First, calculate the pressure fluctuation amplitudes on the left and right sides of the base support separately, then calculate the difference between them. The gearbox mesh clearance change rate is calculated using the same method as above. The vibration energy threshold is determined when setting the dynamic diagnostic criteria, based on the equipment's normal operating vibration energy range and safety requirements. When the difference between the gearbox mesh clearance change rate and the pressure fluctuation amplitudes on both sides of the base support is greater than or equal to 1.2 times the vibration energy threshold, it indicates a fault in the gearbox or main shaft bearing, resulting in uneven force distribution on the equipment and abnormal changes in related parameters. The fault is then determined to be in the gearbox or main shaft bearing. A diagnostic report is generated based on the calculated discrimination factor, the determined fault type, and the located fault location. The discrimination factor is a quantitative indicator of fault diagnosis accuracy. A value closer to 1 indicates higher fault diagnosis accuracy; a value further away from 1 indicates lower accuracy.

[0164] The fault type is clearly marked as a real fault such as instantaneous overload, mechanical jamming, bearing wear, etc. The fault occurrence area is specific to the tool clamping end, transmission gearbox or spindle bearing. When making confidence ratings, multiple factors are taken into consideration. If the various parameters involved change significantly during the fault judgment process, and the data collection is complete and reliable, and the degree of match with the preset standards and historical data is high, the confidence rating will be "high"; if there is a certain fluctuation in some parameters but it does not affect the overall judgment, the data reliability is general, and the confidence rating is "medium"; if the parameter changes are not obvious, the data is missing or the uncertainty is high, the confidence rating is "low". Finally, a complete diagnostic report containing the fault type, location information and confidence rating is formed, providing a clear and accurate basis for equipment maintenance and production decisions.

[0165] By dynamically analyzing the correlation between the motor's electromagnetic torque and the parameters of the quadrilateral region, and comprehensively assessing multiple parameters, this system can accurately distinguish between transient overloads and actual faults such as mechanical jamming and bearing wear. This avoids the misdiagnosis and omissions that are common with traditional diagnostic methods, improving the accuracy and reliability of fault diagnosis. The fault area is located based on parameters such as the tool clamping end deflection angle and the pressure fluctuation amplitude of the base support, enabling maintenance personnel to quickly pinpoint the specific fault location. This reduces troubleshooting time, improves equipment maintenance efficiency, shortens equipment downtime, and ensures production continuity. The generated diagnostic report includes fault type, location information, and confidence rating, providing comprehensive and accurate decision-making support for maintenance personnel and production management. Maintenance personnel can use the report to develop targeted maintenance plans, while production management can rationally adjust production schedules based on the fault situation, optimize equipment maintenance strategies, and reduce maintenance costs. Timely and accurate fault diagnosis and location help companies identify and resolve potential equipment issues, prevent further escalation and deterioration, and effectively extend equipment life.

[0166] An embodiment of the present invention further provides a computing device comprising: a processor and a memory storing a computer program, wherein the computer program, when executed by the processor, executes the system described above. All implementations in the above system embodiments are applicable to this embodiment and can achieve the same technical effects.

[0167] The embodiment of the present invention further provides a computer-readable storage medium storing instructions, which, when executed on a computer, causes the computer to execute the system described above. All implementations in the above system embodiments are applicable to this embodiment and can achieve the same technical effects.

[0168] The above is a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principles of the present invention. These improvements and modifications should also be regarded as within the scope of protection of the present invention.

Claims

1. Intelligent bamboo wood cutting machine fault diagnosis system, characterized by: include: The feature extraction module is used to collect the motor current, voltage, spindle vibration spectrum, and dynamic parameters of the quadrilateral monitoring area composed of the spindle bearing, transmission gearbox, tool clamping end, and base support in real time, and extract the current waveform characteristic value, voltage harmonic components, vibration frequency domain energy distribution parameters, and abnormal parameter combinations within the quadrilateral area; The load mapping module is used to analyze the hardness grade and specification parameters of the current cutting material based on the correlation between the vibration frequency domain energy distribution parameters and the changes in the quadrilateral geometric parameters, combined with the load impact component and the fluctuation value of the diagonal length of the quadrilateral, establish a mapping relationship between the material properties and the dynamic load of the motor, and generate an initial correction value based on the symmetry characteristics of the quadrilateral; The dynamic diagnostic standard module is used to match the dynamic threshold calculation rules based on the material hardness grade and the angular displacement deviation within the quadrilateral area, and combines the material cutting resistance coefficient with the abnormal parameter gradient of the quadrilateral monitoring area to generate dynamic diagnostic standards for the current adaptive threshold interval and vibration energy threshold; The early warning module analyzes the effective current value, harmonic distortion rate, vibration energy, and angular acceleration within the quadrilateral area based on dynamic diagnostic standards. When the current characteristic value is continuously greater than or equal to the dynamic threshold interval and the vibration energy of any two adjacent detection points in the quadrilateral exceeds the standard, a third-level early warning is triggered; The fault diagnosis module is used to distinguish between instantaneous overload and real faults such as mechanical blocking and bearing wear based on a three-level early warning system through dynamic correlation analysis between the motor's electromagnetic torque and changes in quadrilateral area parameters, and generate a diagnostic report.

2. The fault diagnosis system for an intelligent bamboo wood cutting machine according to claim 1, characterized in that: The system collects motor current, voltage, spindle vibration spectrum, and dynamic parameters of the quadrilateral monitoring area consisting of the spindle bearing, transmission gearbox, tool holder, and base support in real time. It also extracts current waveform eigenvalues, voltage harmonic components, vibration frequency domain energy distribution parameters, and abnormal parameter combinations within the quadrilateral area, including: Wavelet packet decomposition is used to reduce noise on current, voltage signals, and vibration spectra. The effective value, harmonic distortion rate, and spectrum energy ratio of the current waveform are extracted as characteristic values. The spindle bearing displacement, gearbox meshing clearance, tool clamping end deflection angle, and pressure fluctuation parameters of the base support are collected within the quadrilateral monitoring area. According to the radial displacement of the spindle bearing and the pressure fluctuation amplitude of the base support part, the real-time fluctuation value of the diagonal length of the quadrilateral is calculated. Combined with the difference between the axial deflection angle of the tool clamping end and the change rate of the meshing clearance of the transmission gearbox, the inner angular displacement deviation of the quadrilateral is generated. The current characteristic value, the tool clamping end deflection angle and the base support pressure fluctuation amplitude, as well as the step diagonal length fluctuation value and the inner angle displacement deviation are combined into an abnormal parameter combination.

3. The fault diagnosis system for the intelligent bamboo wood cutting machine according to claim 2, characterized in that: Based on the correlation between the energy distribution parameters in the vibration frequency domain and the changes in the quadrilateral geometric parameters, combined with the load impact component and the fluctuation value of the diagonal length of the quadrilateral, the hardness grade and specification parameters of the current cutting material are analyzed, and a mapping relationship between the material properties and the dynamic load of the motor is established. The initial correction value based on the symmetry characteristics of the quadrilateral is generated, including: The vibration frequency domain energy distribution parameters are divided into low-frequency, medium-frequency, and high-frequency bands. The energy proportion of each frequency band and the energy gradient change value in adjacent time windows are calculated. Based on the fluctuation value of the diagonal length of the quadrilateral, the maximum fluctuation amplitude and fluctuation frequency in the periodic change characteristics are extracted. Determine the load impact component caused by the sudden change in material hardness based on the energy gradient change value and the maximum fluctuation amplitude in the high-frequency band; Based on the axial deflection angle of the tool clamping end and the change rate of the meshing clearance of the transmission gearbox, combined with the fluctuation value of the diagonal length of the quadrilateral, the quadrilateral symmetry attenuation is calculated, and the ratio of the load impact component to the quadrilateral symmetry attenuation is used as the initial correction value.

4. The fault diagnosis system for an intelligent bamboo wood cutting machine according to claim 3, characterized in that: Based on the material hardness grade and the angular displacement deviation within the quadrilateral area, the dynamic threshold calculation rules are matched. In combination with the material cutting resistance coefficient and the abnormal parameter gradient of the quadrilateral monitoring area, dynamic diagnostic standards for the current adaptive threshold interval and vibration energy threshold are generated, including: According to the load impact component and the initial correction value, a mapping relationship table between the material hardness grade and the dynamic load of the motor is established; Based on the mapping relationship table between material hardness grade and motor dynamic load, combined with the load impact component and initial correction value, the cutting resistance coefficient corresponding to the material hardness grade is generated; According to the abnormal parameter gradient in the quadrilateral area, a linear mapping relationship between the cutting resistance coefficient and the abnormal parameter gradient is established, and the linear mapping coefficient is normalized by the initial correction value to obtain the normalized linear mapping coefficient; Based on the fluctuation value of the diagonal length of the quadrilateral, the diagonal length change rate is calculated in real time and integrated with the normalized linear mapping coefficient to generate a dynamic adjustment factor; According to the normalized linear mapping coefficient and the dynamic adjustment factor, the preset reference current threshold interval is adaptively scaled in sections to generate the current adaptive threshold interval; Based on the vibration frequency domain energy distribution parameters, combined with the high-frequency energy proportion and dynamic adjustment factor, the maximum energy gradient value is calculated, and the vibration energy exceeding threshold is set according to the maximum energy gradient value; The current adaptive threshold interval and the vibration energy exceeding threshold are combined into a dynamic diagnosis standard.

5. The fault diagnosis system for the intelligent bamboo wood cutting machine according to claim 4, characterized in that: According to the normalized linear mapping coefficient and the dynamic adjustment factor, the preset reference current threshold interval is adaptively scaled in sections to generate the current adaptive threshold interval, including: Based on the normalized linear mapping coefficient and the initial correction value, the preset reference current threshold interval is initially segmented and adjusted to generate a preliminary scaling threshold interval; Based on the dynamic adjustment factor and the real-time matching relationship between the rate of change of the diagonal length of the quadrilateral and the material hardness grade, the boundary of the initial scaling threshold interval is dynamically adjusted. That is, when the dynamic adjustment factor is greater than 1, the upper limit value of the current threshold interval is proportionally expanded, and the upper limit adjustment range is the product of the dynamic adjustment factor and the initial correction value; when the dynamic adjustment factor is less than 1, the lower limit value of the current threshold interval is proportionally reduced, and the lower limit adjustment range is the absolute value of the difference between the normalized linear mapping coefficient and the dynamic adjustment factor; According to the linear mapping relationship between the cutting resistance coefficient and the abnormal parameter gradient, combined with the dynamic boundary adjustment results, the current threshold interval is segmented and corrected to generate the current adaptive threshold interval matching the material hardness grade and quadrilateral deformation state.

6. The fault diagnosis system for the intelligent bamboo wood cutting machine according to claim 5, characterized in that: Based on dynamic diagnostic criteria, the system analyzes the effective current value, harmonic distortion rate, vibration energy, and angular acceleration within the quadrilateral area. When the current characteristic value is continuously ≥ the dynamic threshold interval and the vibration energy of any two adjacent detection points in the quadrilateral exceeds the standard, a three-level warning is triggered, including: Based on dynamic diagnostic standards, real-time monitoring of current effective value, harmonic distortion rate and quadrilateral area angular acceleration; When the effective current value is continuously greater than or equal to the current adaptive threshold interval and the harmonic distortion rate increases synchronously, the current characteristic value is determined to be out of limit; the vibration energy values ​​of any two adjacent detection points in the quadrilateral monitoring area are synchronously detected. If the vibration energy is greater than or equal to the vibration energy exceeding the standard threshold, the vibration is determined to be abnormal; Based on the current characteristic value exceeding the limit and vibration abnormality, the quadrilateral symmetry attenuation and the initial correction value are extracted to calculate the real-time ratio of the quadrilateral deformation variable to the initial correction value. When the real-time ratio exceeds the preset threshold, the third-level warning is triggered.

7. The fault diagnosis system for an intelligent bamboo wood cutting machine according to claim 6, characterized in that: The three-level warning includes: Level 1 warning: When the real-time ratio is ≥ 50% of the preset threshold for the first time, it indicates the risk of potential material hardness mutation; Level 2 warning: When the real-time ratio is ≥ 80% of the preset threshold and the vibration energy continues to exceed the standard, it indicates the risk of mechanical jamming or bearing wear; Level 3 warning: When the real-time ratio is continuously ≥ the preset threshold and the current characteristic value exceeds the limit time ≥ the set threshold, it is determined to be a real fault and the shutdown protection command is triggered.

8. The fault diagnosis system for an intelligent bamboo wood cutting machine according to claim 7, characterized in that: Based on the three-level warning system, the system dynamically analyzes the correlation between the motor's electromagnetic torque and the changes in quadrilateral area parameters to distinguish between instantaneous overload, mechanical jamming, and bearing wear. It then generates a diagnostic report, including: Based on the three-level warning results, the load impact component and the quadrilateral symmetry attenuation are extracted. The dynamic correlation analysis between the motor's electromagnetic torque and the quadrilateral's diagonal length fluctuations is then used to calculate the distinguishing factor between instantaneous overload and true faults. If the instantaneous change rate of the load impact component is greater than or equal to the preset mutation threshold, and the quadrilateral symmetry attenuation returns to the initial fluctuation range within the preset time window after the warning, it is determined to be a transient overload caused by a sudden change in material hardness. If the load impact component change rate is continuously greater than or equal to the steady-state threshold, and the gearbox meshing clearance change rate and the quadrilateral diagonal length fluctuation value simultaneously exceed the historical average fluctuation range, it is determined to be a real fault of mechanical blocking or bearing wear. The fault area is located based on the deflection angle of the tool clamping end and the pressure fluctuation amplitude of the base support part. If the deflection angle of the tool clamping end is ≥ the preset safety threshold of the quadrilateral inner angular displacement deviation, and the front side concentration of the pressure fluctuation amplitude of the base support part is ≥ 50% of the dynamic adjustment factor, the fault area is determined to be the tool clamping end. If the change rate of the transmission gearbox meshing clearance and the difference between the pressure fluctuation amplitude on both sides of the base support part is ≥ 1.2 times the vibration energy exceeding the threshold, the fault area is determined to be the transmission gearbox or the spindle bearing. Based on the distinguishing factors and the fault occurrence area, a diagnostic report containing the fault type, location information and confidence rating is generated.

9. A computing device, characterized in that include: one or more processors; A storage device for storing one or more programs, wherein when the one or more programs are executed by the one or more processors, the one or more processors implement the system according to any one of claims 1 to 8.

10. A computer-readable storage medium, characterized in that The computer-readable storage medium stores a program, which, when executed by a processor, implements the system according to any one of claims 1 to 8.

Images