A method and system for predictive fault diagnosis of mine electromechanical equipment

Abstract

The application is a kind of predictive fault diagnosis method and system of mining electromechanical equipment, relates to the technical field of mining electromechanical equipment fault diagnosis, and comprises: natural excitation pickup and quantization module; response measurement module; synchronous acquisition and preprocessing module; frequency response function estimation module; modal parameter extraction module; stiffness degradation diagnosis module.In the application, the inherent non-stationary event during the operation of the mining electromechanical equipment is used as a wideband excitation, and an indirect quantization method of current to electromagnetic torque is combined to construct a digital twin mapping relationship between electrical quantity and mechanical excitation, so that accurate excitation measurement is realized under the condition of no mechanical force sensor, and the industry problems of difficult installation, high cost and easy damage of force sensors caused by high temperature, high humidity, high dust and strict explosion-proof in coal mine are solved.

Description

Technical Field

[0001] This invention relates to the field of fault diagnosis technology for mining electromechanical equipment, and in particular to a predictive fault diagnosis method and system for mining electromechanical equipment. Background Technology

[0002] The transmission chain of heavy electromechanical equipment such as coal mining machines, scraper conveyors, and transfer machines in underground coal mines is a core power transmission component. Faults caused by rigidity degradation, such as loose anchor bolts, slack bearing connections, and cracked bases, are significant contributing factors to equipment downtime and even underground safety accidents. Therefore, achieving early predictive diagnosis of such faults is crucial for ensuring coal mine production safety and improving equipment operation and maintenance efficiency.

[0003] Currently, fault diagnosis of mining electromechanical equipment mainly adopts three types of technical solutions: the first is vibration total value trend monitoring, which analyzes the trend by collecting vibration acceleration amplitude. However, this method relies on the absolute vibration amplitude and is easily affected by time-varying working conditions such as equipment load and operating speed, resulting in a high false alarm rate and failing to effectively identify early stiffness degradation faults. The second is fault classification methods based on neural networks, which require the collection of a large number of fault samples for model training. However, it is difficult to obtain fault samples underground, and the model is a black box structure, lacking physical interpretability, which is not convenient for on-site engineers to understand and make decisions. The third is offline periodic detection methods, such as manual tapping and listening, laser alignment detection, etc., which cannot achieve 24-hour online monitoring and are prone to missing sudden and gradual early loosening faults.

[0004] In addition, traditional operating modal analysis relies on manual hammering or vibrators to provide excitation, which cannot be adapted to the working conditions of explosion-proof, high dust and space-constrained underground operations; while directly installing force sensors to measure excitation force presents industry challenges such as high cost, easy damage and difficulty in explosion-proof wiring.

[0005] Therefore, a predictive fault diagnosis method and system for mining electromechanical equipment is proposed to address the aforementioned problems. Summary of the Invention

[0006] The purpose of this invention is to provide a predictive fault diagnosis method and system for mining electromechanical equipment in order to solve the above-mentioned problems.

[0007] To achieve the above objectives, the present invention adopts the following technical solution: A predictive fault diagnosis system for mining electromechanical equipment includes: The natural excitation pickup and quantization module is configured to use the inherent non-stationary events of equipment operation as excitations and quantify the excitation force through a current-to-electromagnetic torque or force sensor. The response measurement module is configured to install acceleration sensors at preset positions in the transmission chain to collect vibration responses; The synchronous acquisition and preprocessing module is configured to synchronously acquire, filter, denoise, and perform event-triggered filtering of excitation and response signals, and output data. The frequency response function estimation module is configured to use the H1 estimation algorithm combined with power spectrum and coherence function to calculate and verify the accuracy of the system frequency response function under harsh conditions online. The modal parameter extraction module is configured to extract natural frequencies and damping ratios, derive equivalent stiffness, and output parameters with clear physical meaning. The stiffness degradation diagnosis module is configured to compare real-time and baseline modal parameters, determine the degree of loosening of the transmission chain through two-level threshold warnings, and assist in locating the source of loosening.

[0008] Preferably, the natural stimulus picking and quantization module specifically includes: The non-stationary and transient operating events that naturally occur in the actual operation of downhole heavy machinery are collected, and mechanical signals that are approximately step excitation or pulse excitation are generated. The three-phase stator current of the motor is collected, and the excitation component and torque component are obtained through vector transformation and decoupling. Then, the transient electromagnetic torque time waveform of the motor is calculated in real time and is equivalent to the approximate value of the dynamic excitation force at the input end of the transmission system. Before use, the standard force sensor is calibrated to determine the proportional coefficient of current-torque-excitation force.

[0009] Preferably, the measurement module specifically includes: Based on the transmission path of the power chain dynamics, sensors are placed at key locations where vibration response is sensitive and structural loosening has a significant impact, including the bearing housing at the output end of the reducer, the housing near the coupling, the equipment anchor bolt mounting base, and key support parts of the frame. The sensor selection follows coal mine safety standards, and intrinsically safe IEPE accelerometers are used.

[0010] Preferably, the synchronous acquisition and preprocessing module specifically includes: A multi-channel synchronous parallel data acquisition card is adopted, with all channels using the same clock source, while ensuring that the excitation and response signals are aligned on the time axis; An 8th-order Bessel or Butterworth low-pass filter is configured at the analog signal input, with the cutoff frequency uniformly set to 1kHz, to filter out high-frequency noise above the analysis frequency band; an IIR dual second-order digital notch filter is used to filter out power frequency harmonics. The acquired signal is processed to remove DC components and linear trends.

[0011] Preferably, the method further includes: The system employs either signal energy triggering or slope change triggering strategies, monitors signal characteristics in real time, and automatically records valid data for k seconds before and after the trigger point only when a valid excitation event is detected.

[0012] Preferably, the frequency response function estimation module specifically includes: The H1 frequency response function estimation method is used, and the calculation formula is as follows: ; in The power spectral density of the excitation signal; The cross-power spectral density of the excitation and response signals; The effective time-domain data of the trigger record is segmented, and a Hanning window is added to each segment to reduce spectral leakage. The ensemble averaging of the multi-segment spectral results is then performed to obtain the frequency response function curve. Simultaneous calculation of coherence function Used for data quality assessment: .

[0013] Preferably, the modal parameter extraction module specifically includes: On the amplitude-frequency response curve of the frequency response function, the location of the resonance peak is identified by the peak detection algorithm, and the frequency corresponding to the peak is the natural frequency of the structure. ; For the complex value of the frequency response function near a single resonance peak, a circle fitting is performed on the Nyquist plane. The damping ratio is calculated by measuring the center, radius, and phase changes of the fitted circle. With correction of inherent frequency; Under the simplified model of a single-degree-of-freedom system, natural frequencies and stiffness ,quality Satisfying Relationship: ; Equipment structural quality Since the stiffness remains essentially constant during operation, the relative rate of change of stiffness can be directly approximated by the rate of change of natural frequency. ; in, As the reference stiffness; The reference natural frequency; This is the inherent frequency deviation; This refers to stiffness deviation.

[0014] Preferably, the stiffness degradation diagnosis module specifically includes: Under healthy conditions, the reference natural frequencies of each mode are calculated by collecting data from multiple effective excitation events. Compared with the reference damping ratio ; After each valid event is triggered, the frequency response function estimation and modal parameter extraction are repeatedly performed to obtain the current modal parameters. The deviation from the reference value is then calculated to obtain the natural frequency deviation. and damping ratio deviation Early warning thresholds are set based on engineering experience and test data.

[0015] A method for predictive fault diagnosis of mining electromechanical equipment includes: By using the inherent non-stationary events of equipment operation as excitation, the excitation force is quantified through current-to-electromagnetic torque or force sensors to obtain the input excitation signal required for frequency response function calculation; Intrinsically safe low-frequency IEPE accelerometers are installed at key locations in the transmission chain to collect vibration response signals of the mechanical structure; The excitation signal and vibration response signal are simultaneously acquired, filtered and denoised, trend terms are removed and event triggers are screened, and data is output for subsequent analysis. The frequency response function is obtained by combining the H1 estimation algorithm with the power spectrum and coherence function, and then the modal parameters are extracted and the equivalent stiffness is derived. By comparing real-time modal parameters with health baseline parameters, the degree of looseness of the transmission chain is determined through two-level threshold warnings, and the source of looseness is located by combining the changes in measurement points.

[0016] In summary, due to the adoption of the above technical solution, the beneficial effects of the present invention are: 1. This invention utilizes inherent non-stationary events during the operation of mining electromechanical equipment as broadband excitations, and combines them with an indirect quantification method of converting current into electromagnetic torque to construct a digital twin mapping relationship between electrical quantities and mechanical excitations. This enables accurate measurement of excitations under conditions without mechanical force sensors, and solves the industry problem of high temperature, high humidity, high dust, and strict explosion-proof requirements in underground coal mines, which makes it difficult to install force sensors, costly, and prone to damage.

[0017] 2. This invention abandons the traditional black-box data-driven diagnostic model and, based on a mechanical dynamics physical model, directly derives physically meaningful parameters such as equivalent stiffness and natural frequency through frequency response function estimation and modal parameter extraction. The diagnostic results are traceable and verifiable, facilitating quick understanding and decision-making by on-site engineers, and significantly improving the reliability and engineering applicability of the diagnostic results. Attached Figure Description

[0018] Further details, features, and advantages of this application are disclosed in the following description of exemplary embodiments in conjunction with the accompanying drawings, in which: Figure 1 This is a system structure diagram of the present invention; Figure 2 This is a flowchart of the method of the present invention. Detailed Implementation

[0019] Several embodiments of this application will now be described in more detail with reference to the accompanying drawings to enable those skilled in the art to implement this application. This application may be embodied in many different forms and for various purposes and should not be limited to the embodiments set forth herein. These embodiments are provided to make this application thorough and complete, and to fully convey the scope of this application to those skilled in the art. The embodiments described do not limit this application.

[0020] Unless otherwise defined, all terms used herein (including technical and scientific terms) shall have the same meaning as commonly understood by one of ordinary skill in the art to which this application pertains. It will be further understood that terms such as those defined in commonly used dictionaries shall be interpreted as having a meaning consistent with their meaning in the relevant field and / or the context of this specification, and shall not be interpreted in an idealized or overly formal sense unless expressly defined herein.

[0021] Example 1 Its specific implementation method is combined with the appendix Figure 1 and attached Figure 2 Please provide a detailed explanation.

[0022] Appendix Figure 1 This invention provides a structural block diagram of a predictive fault diagnosis system for mining electromechanical equipment, showing the connection relationship between the natural excitation picking and quantization module and the stiffness degradation diagnosis module, and annotating the main functional interaction flow of each module.

[0023] Appendix Figure 2 The flowchart of a predictive fault diagnosis method for mining electromechanical equipment provided in this embodiment of the invention shows the complete steps from obtaining the input excitation signal required for frequency response function calculation to combining the measurement point change to assist in locating the loosening fault source.

[0024] In this embodiment, it includes: The natural excitation pickup and quantification module is configured to use the inherent non-stationary events of equipment operation as excitations, and quantify the excitation force through current-to-electromagnetic torque (indirect method) or force sensor (direct method) to solve the problem of sensorless excitation measurement in downhole. Specifically, it includes: Real-time acquisition and quantification of dynamic excitation force signals applied to large transmission machinery systems in underground coal mines provide accurate and synchronous input reference signals for subsequent frequency response function calculations, which is a prerequisite for realizing input-output dynamic modeling and structural health monitoring.

[0025] Main excitation source: Collect non-stationary and transient operating events that naturally occur in the actual operation of heavy machinery and equipment underground, including heavy-load start-up, shutdown and power outage processes, and sudden load shocks (such as coal cutting by coal mining machines, chain jamming of scraper conveyors, and overload of transfer machines). These transient processes will generate mechanical signals that are approximately step excitations or pulse excitations. Their energy is distributed over a wide frequency range, which can effectively excite multiple natural modes of the mechanical structure. The excitation bandwidth requirements of modal analysis can be met without the need for additional artificial vibration equipment. Incentive quantification method: Indirect method (recommended): For motor-driven transmission systems commonly used in underground coal mines, based on the principles of motor dynamics, there is a strict linear mapping relationship between the output electromagnetic torque of the motor and the amplitude and phase of the stator current. High-precision, intrinsically safe current sensors are used to collect the three-phase stator current of the motor. After vector transformation (Clark transform, Park transform), the excitation component and torque component are obtained through decoupling. The transient electromagnetic torque time waveform of the motor is then calculated in real time and approximated as the dynamic excitation force at the input of the transmission system. Before use, a standard force sensor is calibrated to determine the proportional coefficient of current-torque-excitation force. This solution eliminates the need to install force sensors on the mechanical structure, avoids mechanical damage and explosion-proof hazards, and fully meets the intrinsic safety and space-constrained requirements of underground coal mines.

[0026] Direct method: Install high-precision quartz force sensors or impedance heads at key force transmission nodes in the transmission chain (such as motor base, reducer input shaft bearing seat, and frame connection) to directly measure the dynamic excitation force and response of the structure; however, force sensors are expensive, easily damaged in harsh underground working conditions, and difficult to wire and explosion-proof, and are only used for laboratory prototype verification or ground test bench calibration, and are not suitable for long-term online operation underground.

[0027] By converting electrical signals such as motor current and power into dynamic excitation forces of the mechanical system, and constructing a digital twin mapping relationship between electrical quantities and mechanical excitation, accurate excitation measurement can be achieved under conditions without mechanical force sensors. This fundamentally solves the industry problem of difficulty in installing force sensors and ensuring their long-term reliable operation due to the high temperature, high humidity, high dust, and strict explosion-proof requirements in underground coal mines.

[0028] The response measurement module is configured to install intrinsically safe low-frequency IEPE accelerometers at preset positions in the drive train to collect vibration responses and ensure reliable signal transmission. Specifically, it includes: Accurately collect vibration acceleration response signals generated by mechanical structures under natural excitation to obtain output information of structural dynamic behavior, which together with the excitation signal constitute a complete input-output dynamic dataset.

[0029] Measurement point layout: Based on the transmission path of the transmission chain dynamics, sensors are placed at key locations where vibration response is sensitive and structural loosening has a significant impact, including the bearing housing at the output end of the reducer, the housing near the coupling, the mounting base of the equipment anchor bolts, and key support parts of the frame, etc. The sensor is a low-frequency, high-sensitivity IEPE accelerometer with a range of ±5g, covering normal operation and impact conditions. Its frequency response is 0.1Hz-1kHz, balancing overall structural modal vibration (low frequency) and local component vibration (high frequency). In engineering applications, only 23 key measuring points need to be deployed, reducing hardware costs and installation complexity while ensuring monitoring effectiveness. The combination of mode amplitude and phase information from different measuring points enables precise location of loosening.

[0030] Sensor selection: Strictly adhering to coal mine safety standards, intrinsically safe IEPE accelerometers are used, possessing explosion-proof, dust-proof, and anti-interference capabilities. The sensor output signal is transmitted to the intrinsically safe data acquisition unit underground via a double-shielded cable, avoiding electromagnetic interference and signal attenuation caused by long-distance transmission.

[0031] The sensor mounting contact surface must be cleaned and polished to ensure flatness; magnetic mounting bases or threaded rigid fixing methods should be preferred to avoid loose installation that could cause high-frequency signal attenuation and ensure reliable acquisition of vibration response across the entire frequency band.

[0032] The synchronous acquisition and preprocessing module is configured to synchronously acquire, filter, denoise, and event-triggered filter excitation and response signals, and output high-quality data for analysis. Specifically, it includes: High-precision synchronous sampling, signal conditioning, noise suppression, and data cleaning are performed on the excitation signal (current / torque) and vibration response signal to eliminate the influence of downhole environmental interference, sensor drift, electrical noise, etc., and provide high-quality raw data for frequency response function estimation.

[0033] Synchronous acquisition: A multi-channel synchronous parallel data acquisition card is used, with all channels using the same clock source and a sampling rate of not less than 2.5kHz, satisfying the Nyquist sampling law for the 1kHz analysis frequency band. At the same time, it ensures that the excitation and response signals are strictly aligned on the time axis to avoid systematic deviations in the frequency response function calculation caused by phase errors.

[0034] Anti-aliasing filtering: Configure an 8th-order Bessel or Butterworth low-pass filter at the analog signal input end, with the cutoff frequency uniformly set to 1kHz, to filter out high-frequency noise above the analysis frequency band, avoid frequency aliasing during analog-to-digital conversion, and ensure that the effective low-frequency signal is not distorted; Power frequency notch filtering: Downhole equipment such as motors, cables, and frequency converters exist, resulting in extremely strong interference from the 50Hz power frequency and its harmonics. An IIR dual second-order digital notch filter is used to precisely filter out power frequency harmonics at 50Hz, 100Hz, and 150Hz, significantly improving the signal-to-noise ratio without damaging the effective signal in the mechanical mode frequency band. Trend removal: The acquired signal is processed to remove DC components and linear trends, eliminating signal baseline shift caused by sensor temperature drift, zero drift, and integration error, ensuring accurate frequency domain analysis results.

[0035] Event triggering: The system adopts signal energy triggering or slope change triggering strategies. The system monitors signal characteristics in real time and automatically records valid data for k seconds (k is 5) before and after the trigger point only when a valid excitation event such as start-up, shutdown, or load change is detected. This avoids storing meaningless data for a long time under stable operating conditions and reduces storage pressure and subsequent computing load.

[0036] The frequency response function estimation module is configured to use the H1 estimation algorithm combined with power spectrum and coherence function to calculate and verify the accuracy of the system frequency response function under harsh conditions online. Specifically, it includes: Based on synchronously acquired excitation and response time-domain data, power spectral density and cross-power spectral density are estimated in the frequency domain, and then the frequency response function that characterizes the dynamic properties of the system is calculated. Establish the transmission relationship from input stimulus to output response.

[0037] Specific implementation: Algorithm Selection: The H1 frequency response function estimation method, the most mature and noise-resistant algorithm in engineering, is adopted. Its core assumptions are that noise is mainly concentrated at the response signal end, the excitation signal has a high signal-to-noise ratio, and it is highly compatible with downhole operating conditions. Calculation Formula: ; in The power spectral density of the excitation signal reflects the energy distribution of the excitation at various frequencies; The cross-power spectral density of the excitation and response signals reflects the degree of correlation between them in the frequency domain.

[0038] H1 estimation can effectively suppress unrelated random noise in the response and improve the smoothness and reliability of the frequency response function.

[0039] Spectrum estimation details: The effective time-domain data of the trigger record is segmented, and a Hanning window is added to each segment to reduce spectral leakage. A 50% overlap is set between segments to make full use of the data. The ensemble averaging of the multi-segment spectrum results is performed to finally obtain a smooth and stable frequency response function curve. Coherence function: Simultaneous calculation of coherence function Used for data quality assessment: ; The coherence function ranges from 0 to 1. The closer the value is to 1, the higher the linearity of the excitation and response and the higher the signal-to-noise ratio. If the coherence value is too low, it indicates that the data is severely affected by noise or nonlinear interference, and the frequency response function results for the corresponding frequency band are invalid.

[0040] Abandoning traditional offline manual excitation methods such as laboratory hammering and vibrator methods, this method achieves online, real-time, and adaptive frequency response function calculation using natural equipment excitation. Combined with the H1 estimation algorithm and coherence function quality verification, it ensures high-precision frequency response function acquisition even in harsh downhole environments with strong noise, interference, and vibration.

[0041] The modal parameter extraction module is configured to extract natural frequencies and damping ratios, derive equivalent stiffness, and output parameters with clear physical meanings through methods such as peak picking and circle fitting. Specifically, it includes: From the smooth and reliable frequency response function curve, core modal parameters that can directly reflect the structural stiffness and constraint state are extracted, including natural frequency, damping ratio, and equivalent stiffness, providing quantifiable physical indicators for structural loosening diagnosis.

[0042] Specific implementation: Peak Picking (PP): On the amplitude-frequency response curve of the frequency response function, the position of the resonance peak is identified by a peak detection algorithm, and the frequency corresponding to the peak is the natural frequency of the structure. For low-damping mechanical systems such as coal mine transmission chains, the peak picking method has low computational complexity, high speed, and high stability, making it very suitable for online rapid identification in embedded systems.

[0043] Circular Fitting Mode Identification: For the complex value (real part + imaginary part) of the frequency response function near a single resonance peak, circular fitting is performed on the Nyquist plane. By measuring the center, radius, and phase changes of the fitted circle, the damping ratio is accurately calculated. The method corrects the natural frequency; it is insensitive to noise and can significantly improve the parameter identification accuracy under low damping and strong interference conditions.

[0044] Equivalent stiffness calculation: Under a simplified model of a single-degree-of-freedom system, natural frequencies and stiffness ,quality Satisfying Relationship: ; Equipment structural quality Since the stiffness remains essentially constant during operation, the relative rate of change of stiffness can be directly approximated by the rate of change of natural frequency. ; in, As the reference stiffness; The reference natural frequency; This is the inherent frequency deviation; For stiffness deviation; Meanwhile, structural loosening can lead to increased friction on the connecting surfaces and a deterioration of the contact state, manifested as a significant increase in the damping ratio, which can be used as an auxiliary criterion for judging loosening.

[0045] The traditional complex, offline modal analysis algorithm is made lightweight and real-time, adapted to embedded hardware platforms, and directly outputs parameters with clear physical meanings such as stiffness and damping. It eliminates unexplainable black box features and ensures that the monitoring results are traceable and verifiable.

[0046] The stiffness degradation diagnosis module is configured to compare real-time and reference modal parameters, determine the degree of looseness of the transmission chain through two-level threshold warnings, and assist in locating the source of looseness; Specifically, it includes: By comparing the changing trends of real-time modal parameters and health benchmark parameters, the degree of stiffness degradation of the transmission chain structure is quantitatively assessed, faults such as loose bolts, cracked bases, and loose connections are identified, and graded early warning and positioning information are output.

[0047] Specific implementation: Benchmark modeling: After a new installation, major overhaul, or confirmed healthy state of the equipment, data is collected through multiple effective excitation events to calculate the reference natural frequencies of each mode. Compared with the reference damping ratio This creates a device-specific dynamic health fingerprint database; Trend monitoring: After each valid event is triggered, frequency response function estimation and modal parameter extraction are repeatedly performed to obtain the current modal parameters. The deviation from the benchmark value is calculated to obtain the natural frequency deviation. and damping ratio deviation : , ; The system automatically updates historical trend curves to achieve long-term continuous tracking; Threshold warning: Two levels of warning thresholds are set based on engineering experience and test data: Note the threshold: a 2% decrease in natural frequency or a 20% increase in damping ratio indicates that the structure may be slightly loose, requiring attention and prompting increased inspection.

[0048] Alarm threshold: A 5% drop in natural frequency or a 50% increase in damping ratio, accompanied by a significant decrease in coherence function (enhanced structural nonlinearity), indicates a serious loosening fault. An alarm signal will be output immediately, and it is recommended to stop the machine for inspection in a timely manner.

[0049] Location assistance: If multiple measuring points are set up, the relative changes in the frequency of the same mode at each measuring point can be compared. The location of the measuring point with the largest change is the source of the loosening fault (e.g., loose anchor bolts mainly affect local modes, and the frequency change at the corresponding measuring point is the most significant).

[0050] Example 2 Please see Figure 2 A predictive fault diagnosis method for mining electromechanical equipment includes the following parts: By using the inherent non-stationary events of equipment operation as excitation, the excitation force is quantified through current-to-electromagnetic torque (indirect method) or force sensor (direct method) to obtain the input excitation signal required for frequency response function calculation; Intrinsically safe low-frequency IEPE accelerometers are installed at key locations in the transmission chain to collect vibration response signals of the mechanical structure and ensure reliable signal transmission. The excitation signal and vibration response signal are simultaneously acquired, filtered and denoised, trend term removed and event triggers are screened, and high-quality data is output for subsequent analysis. The H1 estimation algorithm is used in combination with power spectrum and coherence function to obtain accurate frequency response function. Then, modal parameters are extracted and equivalent stiffness is derived by peak picking, circle fitting and other methods. By comparing real-time modal parameters with health baseline parameters, the degree of looseness of the transmission chain is determined through two-level threshold warnings, and the source of looseness is located by combining the changes in measurement points.

[0051] The above formulas are all dimensionless calculations. The formulas are derived from software simulations based on a large amount of collected data to obtain the most recent real-world results. The preset parameters in the formulas are set by those skilled in the art according to the actual situation.

[0052] The foregoing has only described certain exemplary embodiments of the present invention by way of illustration. Undoubtedly, those skilled in the art can modify the described embodiments in various ways without departing from the spirit and scope of the present invention. Therefore, the foregoing drawings and descriptions are illustrative in nature and should not be construed as limiting the scope of protection of the claims of the present invention.

[0053] It should be noted that, in this document, the use of relational terms such as "first" and "second" is merely for distinguishing one entity or operation from another, and does not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes the element.

[0054] It should be understood that in the various embodiments of this application, the order of the above-mentioned processes does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of this application.

[0055] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0056] Those skilled in the art will understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0057] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0058] In addition, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit.

[0059] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

[0060] The foregoing has only described certain exemplary embodiments of the present invention by way of illustration. Undoubtedly, those skilled in the art can modify the described embodiments in various ways without departing from the spirit and scope of the present invention. Therefore, the foregoing drawings and descriptions are illustrative in nature and should not be construed as limiting the scope of protection of the claims of the present invention.

Claims

1. A predictive fault diagnosis system for mining electromechanical equipment, characterized in that, include: The natural excitation pickup and quantification module is configured to use the inherent non-stationary events of equipment operation as excitations and quantify the excitation force through a current-to-electromagnetic torque or force sensor. The response measurement module is configured to install acceleration sensors at preset positions in the transmission chain to collect vibration responses; The synchronous acquisition and preprocessing module is configured to synchronously acquire, filter, denoise, and perform event-triggered filtering of excitation and response signals, and output data. The frequency response function estimation module is configured to use the H1 estimation algorithm combined with power spectrum and coherence function to calculate and verify the accuracy of the system frequency response function under harsh conditions online. The modal parameter extraction module is configured to extract natural frequencies and damping ratios, derive equivalent stiffness, and output parameters with clear physical meaning. The stiffness degradation diagnosis module is configured to compare real-time and baseline modal parameters, determine the degree of looseness of the transmission chain through two-level threshold warnings, and assist in locating the source of looseness.

2. The predictive fault diagnosis system for mining electromechanical equipment according to claim 1, characterized in that, The natural stimulus picking and quantization module specifically includes: The system collects non-stationary and transient operating events that naturally occur in downhole heavy machinery during actual operation, generating mechanical signals that are approximately step-excited or pulse-excited. The three-phase stator current of the motor is collected, and the excitation component and torque component are obtained through vector transformation and decoupling. Then, the transient electromagnetic torque time waveform of the motor is calculated in real time and is equivalent to the approximate value of the dynamic excitation force at the input end of the transmission system. Before use, the standard force sensor is calibrated to determine the proportional coefficient of current-torque-excitation force.

3. The predictive fault diagnosis system for mining electromechanical equipment according to claim 1, characterized in that, The response measurement module specifically includes: Based on the transmission path of the power chain dynamics, sensors are placed at key locations where vibration response is sensitive and structural loosening has a significant impact, including the bearing housing at the output end of the reducer, the housing near the coupling, the equipment anchor bolt mounting base, and key support parts of the frame. The sensor selection follows coal mine safety standards, and intrinsically safe IEPE accelerometers are used.

4. The predictive fault diagnosis system for mining electromechanical equipment according to claim 1, characterized in that, The synchronous acquisition and preprocessing module specifically includes: A multi-channel synchronous parallel data acquisition card is adopted, with all channels using the same clock source, while ensuring that the excitation and response signals are aligned on the time axis; An 8th-order Bessel or Butterworth low-pass filter is configured at the analog signal input, with the cutoff frequency uniformly set to 1kHz, to filter out high-frequency noise above the analysis frequency band; an IIR dual second-order digital notch filter is used to filter out power frequency harmonics. The acquired signal is processed to remove DC components and linear trends.

5. A predictive fault diagnosis system for mining electromechanical equipment according to claim 4, characterized in that, Also includes: The system employs either signal energy triggering or slope change triggering strategies, monitors signal characteristics in real time, and automatically records valid data for k seconds before and after the trigger point only when a valid excitation event is detected.

6. The predictive fault diagnosis system for mining electromechanical equipment according to claim 1, characterized in that, The frequency response function estimation module specifically includes: The H1 frequency response function estimation method is used, and the calculation formula is as follows: ; in The power spectral density of the excitation signal; The cross-power spectral density of the excitation and response signals; The effective time-domain data of the trigger record is segmented, and a Hanning window is added to each segment to reduce spectral leakage. The ensemble averaging of the multi-segment spectral results is then performed to obtain the frequency response function curve. Simultaneous calculation of coherence function Used for data quality assessment: .

7. A predictive fault diagnosis system for mining electromechanical equipment according to claim 1, characterized in that, The modal parameter extraction module specifically includes: On the amplitude-frequency response curve of the frequency response function, the location of the resonance peak is identified by the peak detection algorithm, and the frequency corresponding to the peak is the natural frequency of the structure. ; For the complex value of the frequency response function near a single resonance peak, a circle fitting is performed on the Nyquist plane. The damping ratio is calculated by measuring the center, radius, and phase changes of the fitted circle. With correction of inherent frequency; Under the simplified model of a single-degree-of-freedom system, natural frequencies and stiffness ,quality Satisfying Relationship: ; Equipment structural quality Since the stiffness remains essentially constant during operation, the relative rate of change of stiffness can be directly derived from the rate of change of natural frequency. ; in, As the reference stiffness; The reference natural frequency; This is the inherent frequency deviation; This refers to stiffness deviation.

8. A predictive fault diagnosis system for mining electromechanical equipment according to claim 1, characterized in that, The stiffness degradation diagnosis module specifically includes: Under healthy conditions, the reference natural frequencies of each mode are calculated by collecting data from multiple effective excitation events. Compared with the reference damping ratio ; After each valid event is triggered, the frequency response function estimation and modal parameter extraction are repeatedly performed to obtain the current modal parameters. The deviation from the reference value is then calculated to obtain the natural frequency deviation. and damping ratio deviation Early warning thresholds are set based on engineering experience and test data.

9. A method for predictive fault diagnosis of mining electromechanical equipment, wherein the predictive fault diagnosis system for mining electromechanical equipment according to claim 1 is characterized in that, include: By using the inherent non-stationary events of equipment operation as excitation, the excitation force is quantified through current-to-electromagnetic torque or force sensors to obtain the input excitation signal required for frequency response function calculation; Intrinsically safe low-frequency IEPE accelerometers are installed at key locations in the transmission chain to collect vibration response signals of the mechanical structure. The excitation signal and vibration response signal are simultaneously acquired, filtered and denoised, trend terms are removed and event triggers are screened, and data is output for subsequent analysis. The frequency response function is obtained by combining the H1 estimation algorithm with the power spectrum and coherence function, and then the modal parameters are extracted and the equivalent stiffness is derived. By comparing real-time modal parameters with health baseline parameters, the degree of looseness of the transmission chain is determined through two-level threshold warnings, and the source of looseness is located by combining the changes in measurement points.

Images