A method and system for identifying abnormal operation of mine electromechanical equipment

Abstract

The present application relates to the technical field of equipment anomaly recognition, and particularly relates to a method and system for recognizing abnormal operation of mine electromechanical equipment, the method comprising: constructing a multi-source information synchronous acquisition architecture, obtaining bottom layer synchronous data with a unified timestamp, the bottom layer synchronous data including operation parameters, tail tension and original weighing signals; decoupling and reconstructing the original weighing signals through a digital signal processing algorithm to extract instantaneous linear load equivalent reflecting real material distribution; introducing a viscoelastic model of the conveyor belt, combining with self-adaptive updated equipment parameters to generate a dynamic speed difference threshold; executing multivariate coupled feature recognition logic to comprehensively consider multiple parameter features and distinguish between false skidding and real abnormal operation state; the present application actively suppresses system oscillation without interrupting production through dynamic speed difference threshold and multivariate collaborative recognition, ensuring continuous and stable operation of the conveyor belt.

Description

Technical Field

[0001] This invention relates to the field of equipment anomaly identification technology, specifically to a method and system for identifying operational anomalies in mining electromechanical equipment. Background Technology

[0002] As a widely used continuous transportation equipment in the industrial field, belt conveyors play a core role in material transportation in the mining industry due to their advantages such as large conveying capacity, simple structure, and convenient maintenance. Their stable operation is directly related to the efficiency and safety of the production line, and timely handling of equipment failures is the key to ensuring continuous production.

[0003] Existing mine belt conveyors operate at high speeds and under heavy loads, resulting in significant inertial impacts during braking. This causes the belts to experience extreme tensile stress, severely impacting their service life and frequently leading to belt breakage accidents.

[0004] Chinese patent application number CN202322265838.0 discloses an intelligent constant deceleration brake control system for mining belt conveyors, used for upgrading existing mine belt conveyor brake control systems or for supporting new production mine belt conveyors. The intelligent constant deceleration brake control system includes a brake control unit, a brake hydraulic unit, and a brake unit. The brake control unit is electrically connected to the main control unit and drive unit of the mine belt conveyor, as well as the brake hydraulic unit and brake unit of the intelligent constant deceleration brake control system. During normal operation of the mine belt conveyor, the main control unit controls the drive unit and brake hydraulic unit to operate and maintain constant deceleration braking of the belt conveyor. When the main control unit malfunctions or the brake unit data collected by the brake control unit is abnormal, the brake control unit takes over from the main control unit to control the brake hydraulic unit, maintaining constant deceleration braking of the mine belt conveyor to prevent impact during braking and thus reduce belt breakage accidents.

[0005] Although the aforementioned solution allows the brake control unit to automatically take over when the main control unit malfunctions or data becomes abnormal, ensuring constant deceleration during braking through hydraulic adjustment, thereby greatly reducing the risk of impact and tire breakage, some other problems may arise in practical applications. Specifically: In existing technologies, slippage protection is a key safety measure for belt conveyors. Its principle is usually achieved by monitoring the synchronization between the speed of the drive roller and the actual running speed of the conveyor belt. Currently, most existing methods involve installing sensors such as proximity switches next to the driven roller at the head of the conveyor to detect speed signals.

[0006] However, at the moment the equipment starts, since the belt itself is an elastic component, after the start command is issued, the drive roller begins to rotate. However, due to inertia and elastic elongation, the belt speed gradually increases from zero and cannot reach a stable value instantly. As a result, in the initial stage of startup, there is a brief, non-faulty speed difference between the drive roller speed and the belt speed.

[0007] Traditional slip detection devices have a fixed delay time to avoid interference during this stage; however, for long-distance, high-inertia conveyors, the startup process may far exceed the preset delay. If the protection device monitors too early, it may misjudge this normal physical acceleration process or the "false speed difference" caused by elastic elongation as a dangerous "slippage accident," thereby triggering an unnecessary emergency shutdown. Such malfunctions not only interrupt production, but frequent false alarms also reduce the staff's trust in the automatic protection system. Summary of the Invention

[0008] The purpose of this invention is to provide a method and system for identifying abnormal operation of mining electromechanical equipment, aiming to solve the problems mentioned in the background art.

[0009] To achieve the above objectives, the present invention provides the following technical solution: This invention provides a method for identifying abnormal operation of mining electromechanical equipment, comprising: A multi-source information synchronous acquisition architecture is constructed to obtain underlying synchronous data with a unified timestamp. The underlying synchronous data includes operating parameters, tail tension force, and original weighing signal. The original weighing signal is decoupled and reconstructed using digital signal processing algorithms to extract the instantaneous linear load equivalent that reflects the actual material distribution. A viscoelastic model of the conveyor belt is introduced, and combined with adaptively updated equipment parameters, a dynamic speed difference threshold is generated. The feature identification logic is executed in a multivariate coupling manner. It integrates the actual speed difference, dynamic speed difference threshold, static threshold and multiple parameter features in the underlying synchronization data to distinguish between false slippage and real abnormal operating state.

[0010] Preferably, obtaining underlying synchronization data with a unified timestamp specifically includes: The angular displacement fed back by the absolute encoder installed on the shaft end of the drive drum and the tail redirecting drum is acquired through high-speed industrial real-time Ethernet and converted into linear velocity. The operating parameters of the drive drum are obtained through the high-speed communication interface of the frequency converter, the tail tension force is obtained through the high-frequency tension sensor at the tail hydraulic tensioning device, and the original weighing signal is obtained through the dynamic weighing sensor array downstream of the receiving point. By using a timestamp alignment algorithm, the data sequence with standard intervals is reconstructed through interpolation compensation, thereby achieving alignment of multivariate time series.

[0011] Preferably, extracting the instantaneous linear load equivalent that reflects the actual material distribution specifically includes: A multi-level wavelet packet decomposition signal processing procedure is used to divide the original weighing signal into sub-signals of different frequency bands. High-frequency oscillation components identified as physical shocks and local jitter noise are filtered out by threshold denoising, and low-frequency components representing stable load changes are extracted. The extracted low-frequency components are subjected to inverse wavelet transform, and the reconstructed signal is synthesized and cleaned by zero-phase filtering. The amplitude integral of the reconstructed signal is divided by the belt-related parameters to obtain the true instantaneous line load equivalent.

[0012] Preferably, generating a dynamic speed difference threshold specifically includes: Establish a parameter adaptive matrix, and dynamically update the elastic modulus and viscous damping coefficient of the conveyor belt by reading ambient temperature sensor data and the cumulative running time of the equipment; Substitute the motor output torque, tail tension force, and instantaneous linear load equivalent into the viscoelastic model to solve for the maximum transient stress and the corresponding theoretical maximum allowable strain rate. Based on the theoretical maximum allowable strain rate and effective transmission length, the maximum theoretical speed difference allowed by the system is calculated, thereby generating the envelope of the dynamic speed difference threshold that converges smoothly with the system's operating state.

[0013] Preferably, distinguishing between false slippage and genuine abnormal operating conditions specifically includes: The underlying synchronous data is used to build a dynamically updated feature matrix, and the historical sequence of parameters is stored through a sliding time window. When the actual speed difference is between the upper limit of the static threshold and the dynamic speed difference threshold, the collaborative comparison program is triggered to analyze the state characteristics of the motor output torque and tail tension. If the motor output torque is within the preset rising or stable range, and the tail tension does not suddenly drop, it is determined to be a false slip caused by transient elastic elongation, and the stop signal is blocked. If the actual speed difference exceeds the upper limit of the dynamic speed difference threshold, or if there is a sudden and sharp drop in the motor output torque and tail tension, it is determined to be a real slippage or belt breakage accident, triggering emergency braking.

[0014] Preferably, it includes: When a false slip is detected, a collaborative disturbance rejection procedure is initiated, which includes parallel execution of drive adjustment and tension compensation. A flexible adjustment command is issued at the drive end, and the acceleration set value is instantly reduced based on the feedforward algorithm, and the torque output of the frequency converter is finely adjusted and compensated. A pressure superposition command is issued to the hydraulic tensioning device at the tensioning end, and a motion decision algorithm including proportional and differential actions is used to actively increase the dynamic tension to improve the local tensile stiffness.

[0015] The present invention also provides a system for identifying abnormal operation of mining electromechanical equipment, the system comprising: The multi-source synchronous acquisition module is used to build a high-frequency synchronous acquisition architecture to acquire low-level synchronous data with a unified timestamp, including absolute displacement, tension force, operating parameters and weighing signals. The signal cleaning and reconstruction module is used to remove high-frequency interference from the weighing signal through wavelet packet decomposition and digital filtering algorithms, and to calculate the true instantaneous line load equivalent. The dynamic threshold generation module is used to introduce a parameter-adaptive viscoelastic model, combine it with the real-time stress state to derive the theoretical maximum allowable strain rate, and generate a dynamic velocity difference threshold that evolves with the working conditions. The multivariate feature identification module is used to identify normal elastic deformation and real operational faults by comparing the relationship between the actual speed difference and the dynamic speed difference threshold and static threshold based on the feature matrix and the multidimensional state trend within the sliding time window.

[0016] Preferably, in the multi-source synchronous acquisition module: The mechanical interface of the absolute encoder that acquires absolute displacement adopts a heavy-duty bearing structure, and the electrical output is a differential signal; The high-frequency tension sensor for collecting tension has an overload-resistant design to resist damage from instantaneous impacts; The weighing signal is acquired by a dynamic weighing sensor array covering the full width of the belt and uploaded after preliminary filtering by a local transmitter.

[0017] Preferably, the dynamic threshold generation module includes a model verification and correction mechanism: Offline calibration is achieved by executing a calibration program under no-load conditions, collecting strain response data, and then back-calculating and updating the actual elastic modulus and viscous damping coefficient. The relationship between the actual speed difference and the dynamic speed difference threshold is continuously monitored, and the online model self-check program is triggered when the actual value continuously approaches the upper limit of the dynamic speed difference threshold.

[0018] Preferably, the system further includes an active cooperative anti-interference module, which is used for: When the multivariate feature identification module determines that the slippage is false, an electromechanical-hydraulic active cooperative anti-disturbance command is sent to the drive end and the tensioning end. Continuously monitor the actual speed difference and tail tension status. Once the operation returns to stability, the additional control quantity is smoothly canceled through the ramp function, allowing the equipment to return to the normal closed-loop operation mode. When the speed difference exceeds the upper limit of the dynamic speed difference threshold, the disturbance rejection control is interrupted and emergency braking is triggered.

[0019] The technical effects and advantages of this invention are as follows: 1. This invention effectively overcomes the problem of poor adaptability of traditional fixed thresholds under transient conditions such as equipment startup and load changes by introducing a mechanism to dynamically generate speed difference thresholds using a viscoelastic model. This mechanism can adaptively adjust the threshold based on real-time load, ambient temperature, and equipment aging status. Especially during heavy-load startup, it allows the conveyor belt to produce necessary elastic elongation, thereby significantly reducing the misjudgment of normal physical phenomena as slippage faults and improving the system's fault tolerance under complex operating conditions.

[0020] 2. This invention introduces a viscoelastic model to dynamically generate a speed difference threshold. When false slippage is detected, the system can immediately activate the collaborative anti-disturbance mechanism between the drive end and the tension end. By flexibly adjusting the acceleration setpoint and dynamically superimposing the tension force, it actively suppresses elastic oscillations, avoids unnecessary shutdowns, and thus ensures the continuity and stability of the conveying process and improves the overall operating efficiency. Attached Figure Description

[0021] Figure 1 This is a schematic diagram of the belt conveyor of the present invention; Figure 2 This is a diagram of the overall system architecture of the present invention; Figure 3 This is a flowchart of the core identification method of the present invention; Figure 4 This is a flowchart of the multivariate feature identification logic judgment of the present invention.

[0022] In the picture: 1. Drive roller; 2. Tail diverter roller; 3. Hydraulic tensioning device. Detailed Implementation

[0023] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present invention. Example 1

[0024] During the operation of long-distance belt conveyors, traditional monitoring methods are often limited by asynchronous sensor data acquisition or low sampling frequency, which makes it difficult for the system to accurately capture transient dynamic responses such as the propagation delay of elastic waves generated by the conveyor belt during operation, and makes it difficult to obtain high-quality basic input data.

[0025] In view of this, refer to Figures 1 to 4 As shown, this embodiment constructs a multi-source information high-frequency synchronous acquisition architecture, which provides deterministic latency and extremely low communication jitter through high-speed industrial real-time Ethernet, thereby achieving data synchronization.

[0026] The architecture is centered around a main control PLC. The main control PLC achieves global alignment through a hardware clock synchronization protocol, ensuring that all data points have a unified timestamp. Specifically, the sampling period of the underlying sensors is uniformly set to 10 milliseconds.

[0027] In this embodiment, absolute encoders with a resolution of not less than 12 bits are installed on the shaft ends of the drive roller 1 and the tail redirecting roller 2, respectively, to measure the angular displacement of the rollers in real time and convert it into linear velocity, specifically the linear velocity of the drive roller 1 and the linear velocity of the tail redirecting roller 2.

[0028] It should be noted that the mechanical interface of the encoder in this embodiment must adopt a heavy-duty bearing structure to resist the radial impact when the drum starts and stops; the electrical output must be a differential signal to avoid electromagnetic interference in the long-distance transmission environment underground; at the same time, the operating parameters of the active drum 1, including stator current, output frequency and torque, are obtained in real time through the high-speed communication interface of the frequency converter, and the sampling rate is consistent with that of the sensor.

[0029] At the tail section, the hydraulic tensioning device is equipped with three high-frequency tension sensors with a range of 0-500kN. These sensors must have a high sampling rate (usually above 500Hz) and be designed to withstand overload (such as dual-bridge redundancy) to prevent damage from the instantaneous impact of falling large pieces of gangue. After the tension signal is converted from an A / D converter, it is sent to the main control PLC via Ethernet by the local IO module.

[0030] A dynamic weighing sensor array is set downstream of the receiving point. The dynamic weighing sensor array can be composed of multiple piezoelectric sensors. Its array layout needs to cover the full width of the belt to measure the distribution of materials. It should be noted that in this implementation, the signal detected by the weighing sensor needs to be preliminarily filtered by the local transmitter (e.g., low-pass filter cutoff frequency 10Hz) before being uploaded to the PLC.

[0031] To achieve absolute alignment of multivariate time series, this architecture implements a timestamp alignment algorithm at the software level, assigning a precise timestamp based on GPS or network synchronization to each sensor's data packet. In this embodiment, the timestamp alignment algorithm employs interpolation compensation. If a sensor's data arrives late due to network jitter, the PLC performs linear interpolation based on the values ​​at the preceding and following time points to reconstruct a standard 10-millisecond interval sequence. The aligned multivariate time series (including linear velocity, torque, tension, and linear load) is stored in a circular buffer, providing continuous and synchronous physical quantity input for subsequent analysis. Example 2

[0032] Based on the synchronization data in Example 1, this example addresses the high-frequency interference problem caused by material impact. Specifically, when large pieces of gangue and other materials fall, they will cause severe physical impact on the conveying equipment. The resulting high-frequency interference signal is easily mixed with the real load signal, causing the traditional system to frequently trigger false overload alarms and making it impossible to accurately extract the real linear load equivalent that reflects the material distribution.

[0033] Reference Figures 1 to 4 As shown, this embodiment, based on the multi-source information high-frequency synchronous acquisition architecture established in Embodiment 1, uses digital signal processing algorithms to extract low-frequency components reflecting the actual material load from the composite broadband signal, providing reliable data input for subsequent adaptive control.

[0034] At the hardware level, the high-speed industrial real-time Ethernet and high-frequency synchronous acquisition system constructed in Example 1 ensure that the raw data from the dynamic weighing sensor array can be transmitted to the main control PLC in a period of 10 milliseconds with a unified timestamp.

[0035] At the software level, this embodiment embeds a digital filtering algorithm into the main control PLC of Embodiment 1. This algorithm is based on a three-layer wavelet packet decomposition signal processing flow to decompose, analyze and reconstruct the original weighing signal containing broadband components.

[0036] The first step in implementing the algorithm is to perform wavelet packet decomposition on the original weighing signal. The system pre-selects the Db4 (Daubechies 4th order) wavelet as the basis function. Due to its good compact support and regularity, it can effectively balance the localization analysis capabilities in the time and frequency domains and is suitable for capturing transient features in the weighing signal.

[0037] During decomposition, the original weighing signal is divided into sub-signals of different frequency bands by passing through high-pass and low-pass filter banks. Specifically, since this embodiment adopts a three-layer wavelet packet decomposition signal processing flow, the first layer divides the overall frequency band of the original signal in half; the second layer further divides the low-frequency part obtained in the first layer; and the third layer further subdivides the low-frequency part obtained in the second layer. In this way, the original signal is decomposed into several sub-band signals with different frequency ranges, and each sub-band contains component information of a specific frequency band in the original signal.

[0038] After each layer of decomposition, the coefficients are denoised using a threshold (the threshold rule adopts the Minimax criterion) to filter out components with a frequency greater than 20 Hz. This means that oscillation components with a frequency greater than 20 Hz are determined to be noise generated by simple physical impacts (such as large pieces of gangue falling) and local vibrations of the belt. At the same time, low-frequency components in the range of 0 to 5 Hz are extracted to determine that this frequency band represents the stable load changes of the material distribution on the belt, i.e., the real instantaneous linear load information.

[0039] During the reconstruction phase, the system performs inverse wavelet transform on the coefficients of the 0-5Hz sub-band to synthesize the cleaned weighing signal. To ensure that the reconstructed signal is not offset on the time axis and to avoid introducing phase distortion that could affect the real-time performance of subsequent control, the algorithm uses zero-phase filtering to ensure that the output signal is aligned with the original signal in time. The reconstructed signal represents the true instantaneous linear load equivalent, and its calculation requires combining the belt speed signal from the encoder and the spatial calibration of the sensor array. Specifically, the system integrates the amplitude of the reconstructed signal and divides it by the product of the belt length covered by the weighing sensor array and the real-time linear velocity of the tail-end redirecting roller 2 to obtain the instantaneous linear load equivalent. The main control PLC uses this equivalent value as the input parameter for subsequent models and updates it to the shared memory area in real time.

[0040] The entire wavelet packet decomposition and reconstruction algorithm runs in the dedicated FPGA module of the main control PLC, with the calculation delay controlled within 2 milliseconds to avoid affecting the 10-millisecond sampling period. In addition, the system is designed with an adaptive correction mechanism. When an abnormal belt running status is detected (such as a sudden speed change), it will temporarily switch to the alternative wavelet basis function to enhance the robustness of the algorithm. Example 3

[0041] Due to the inherent viscoelastic properties of conveyor belts, traditional static speed difference thresholds lack flexibility and cannot be adjusted in real time according to changes in ambient temperature, equipment aging, and actual working conditions. This makes it difficult for the system to accurately distinguish between normal physical elastic deformation and actual belt slippage in complex operating environments.

[0042] Therefore, this embodiment introduces the Kelvin-Voigt viscoelastic model of the conveyor belt to generate a dynamic speed difference threshold that evolves in real time with the working conditions, thereby replacing the fixed static threshold used in traditional control.

[0043] Reference Figures 1 to 4 As shown, specifically, the Kelvin-Voigt viscoelastic model is pre-set in the main control PLC, which simplifies the conveyor belt as a material that simultaneously possesses the properties of an elastic solid and a viscous fluid.

[0044] In this model, the actual stress at a certain point inside the conveyor belt is the sum of two parts. One part is proportional to the instantaneous strain at that point, reflecting the elastic response of the material; the other part is proportional to the instantaneous rate of change of strain at that point, reflecting the viscous response of the material. Its damping characteristics are determined by the viscous damping coefficient.

[0045] To ensure that this theoretical model accurately reflects the actual behavior of a specific conveyor belt throughout its entire lifecycle, an adaptive parameter matrix is ​​constructed within the system. This matrix dynamically outputs and updates the optimal elastic modulus and viscous damping coefficient values ​​for the conveyor belt under the current condition by real-time reading of ambient temperature sensor data deployed in the tunnel and the accumulated equipment operating time recorded in the PLC. When the ambient temperature rises, the polymer material of the conveyor belt softens, and the elastic modulus decreases accordingly. Furthermore, as the equipment operating time accumulates, the material may experience changes in viscous damping characteristics due to fatigue aging. These effects are all corrected in real-time through this matrix, ensuring that the model closely matches the actual state of the equipment.

[0046] The calculation process of the dynamic speed difference threshold is as follows: In each sampling period, the system will input the real-time motor output torque, the actual tail tension force provided in Example 1, and the real instantaneous line load equivalent obtained after wavelet packet decomposition and reconstruction in Example 2 into the Kelvin-Voigt model that has been updated with the latest parameters, in order to solve for the maximum transient stress that the conveyor belt can theoretically withstand under the current working conditions.

[0047] Since the strain rate directly characterizes the relative stretching speed per unit length of belt, the system can further derive the corresponding theoretical maximum allowable strain rate from the solved maximum transient stress. After determining this theoretical maximum allowable strain rate, the maximum theoretical speed difference allowed by the system at the current moment is obtained by multiplying this maximum allowable strain rate by the effective transmission length of the belt.

[0048] Based on this principle, the system can generate a dynamic speed difference threshold upper limit envelope during device startup. This envelope allows for a relatively large transient speed difference in the early stages of startup, such as 0.3-0.5 m / s, to accommodate the large elastic elongation that inevitably occurs during heavy-load startup. Subsequently, as the system gradually approaches a stable operating state, the envelope will smoothly decrease according to an exponential decay law, and finally converge smoothly to a small conventional static threshold, such as 0.1 m / s, at the end of the startup process, thus achieving a smooth switch from the startup transient to the steady-state operation.

[0049] To ensure the long-term accuracy of the model, the system also has a verification and calibration mechanism that combines offline and online methods. Periodically, such as once a month, the system will execute a special calibration program under no-load conditions. This program applies a known, smoothly changing torque step by controlling the main drive frequency converter, while simultaneously acquiring high-frequency strain response data of the conveyor belt. Then, it uses these measured data to inversely calculate the most accurate elastic modulus and viscous damping coefficient of the conveyor belt under the current condition, and updates the corresponding values ​​in the adaptive matrix with these new parameters, thereby achieving self-calibration of the model.

[0050] During online operation, the system continuously monitors the relationship between the actual speed difference and the dynamic speed difference threshold. If the actual value is very close to but does not exceed the upper limit of the dynamic speed difference threshold multiple times, the model self-check program will be triggered. The system will record the deviation log and prompt the maintenance personnel to pay attention. All calculated dynamic speed difference threshold data, as well as model parameters, environmental variables, etc., are uploaded to the monitoring system through the communication interface of the main control PLC for real-time graphical display, providing operators with an intuitive perception of the equipment status. Example 4

[0051] In terms of fault diagnosis logic, relying solely on a single speed difference criterion is often insufficient to support reliable decision-making. This makes it easy for the system to misjudge the transient elastic elongation caused by heavy-load startup as belt slippage or belt breakage, lacking the deep identification capability of multivariate coupling.

[0052] In view of this, this embodiment introduces feature identification logic with multivariable coupling, and by integrating the inherent correlation and trend of various information such as speed, force, and current, it can distinguish between normal elastic deformation of the conveyor belt and real dangerous slippage or belt breakage accidents, thereby reducing the false alarm rate.

[0053] Reference Figures 1 to 4As shown, specifically, the system continuously receives high-frequency synchronous acquisition data from Embodiment 1, including real-time speed difference, active drum 1 torque, and tail tension; as well as the upper limit of the dynamic speed difference threshold, which decays exponentially over time, and the static threshold as a reference, provided by the adaptive dynamic threshold generation module of Embodiment 3; these parameters are organized by the main control PLC in a logic matrix, which is a dynamically updated feature set that can reflect the temporal correlation between the parameters. By maintaining a fixed-length sliding time window, it is used to store the historical sequence of parameters within a recent period (e.g., corresponding to the most recent 500 milliseconds), enabling the identification logic to analyze the changing trend of the parameters.

[0054] It should be noted that the real-time speed difference is the difference between the linear speed of the drive roller 1 measured by the high-resolution absolute encoder and the linear speed of the tail redirecting roller 2.

[0055] When identifying genuine slippage, the system first calculates the actual speed, which is the difference between the linear speed of the drive roller 1 and the linear speed of the tail redirecting roller 2.

[0056] If the actual speed difference does not exceed the static threshold, the system determines that the operating state is normal.

[0057] If the actual speed difference is between the upper limit of the static threshold and the upper limit of the dynamic speed difference threshold, the system will not immediately trigger the highest level of slippage alarm. Instead, it will start the collaborative comparison program of the feature matrix. At this time, the system will analyze the status characteristics of other key equipment parameters to analyze whether a "false slippage" phenomenon has occurred. Specifically, the system will analyze whether the output torque of the active roller 1 is in a normal rising or stable range, that is, whether the deviation between the current torque value and the previous moment is within a reasonable percentage range (e.g., <±10%). At the same time, it will detect whether the tension force signal fed back by the hydraulic tensioning device 3 at the tail of the machine has experienced a sudden and significant drop (e.g., the tension force drops by no more than 5% within 1 second). If both conditions are met, the system will determine this situation as a temporary elastic elongation phenomenon of the conveyor belt during heavy-load start-up or acceleration, that is, a "false slippage" phenomenon. At this time, the main control PLC will block the slippage stop signal, but will record this phenomenon in the warning log for post-event analysis.

[0058] If the actual speed difference exceeds the upper limit of the dynamic speed difference threshold, or if abnormal speed fluctuations are accompanied by a sharp drop in motor output torque within a very short time (e.g., a drop of more than 30%), and the tail tension also drops instantaneously (e.g., a drop of 20% within 1 second), then it is determined to be a real slippage or belt breakage accident, and emergency braking is immediately triggered.

[0059] To further improve the accuracy of identification, the system calculates the statistical correlation (such as Pearson correlation coefficient) between the change in actual speed difference and the change in torque of drive roller 1 within the slip time window. Specifically, in real slippage accidents, due to the loss of traction, the increase in speed loss is often accompanied by a sharp drop in torque, and the two will show a significant negative correlation. However, in the case of pseudo-slippage, the increase in speed difference may be due to belt elongation, while the torque may remain stable or steadily increase due to the adjustment of the control system. The two may not have a significant correlation or show different relationship patterns.

[0060] In addition, the system can be configured with a confidence management mechanism. If the system judges the slippage as false for several consecutive cycles, but the actual speed difference does not decrease but continues to widen, the system will automatically reduce the confidence of this judgment and gradually increase the response level, providing a safety redundancy for possible misjudgments. Example 5

[0061] Current control strategies are usually quite passive after identifying operational anomalies, mostly dealing with potential risks by directly shutting down the system. They lack an active disturbance rejection mechanism that can actively suppress elastic oscillations and maintain continuous system operation through coordinated operation of the drive end and tension end after identifying false disturbances.

[0062] Therefore, in this embodiment, when "false slippage" is determined in Embodiment 4, a cooperative disturbance rejection program is simultaneously initiated. The cooperative disturbance rejection program includes two parallel execution paths: drive adjustment and tension compensation. Specifically: Reference Figures 1 to 4 As shown, on the drive side, the main control PLC issues a flexible adjustment command to the main drive inverter. The command is based on the PID feedforward algorithm to make the main drive inverter instantly reduce the current acceleration set value (specifically, it can reduce the current acceleration set value by 20%-30%), thereby appropriately extending the acceleration time to reduce the disturbance caused by the surge in traction force on the conveyor belt. At the same time, the system will also calculate a feedforward compensation amount based on the real-time speed difference to fine-tune the torque output of the inverter, ensuring that the change in driving force is smoother.

[0063] At the tensioning end, the system issues a pressure superposition command to the hydraulic tensioning device 3 to rapidly increase the local tensile stiffness of the system. The pressure increment value of this pressure superposition command is dynamically determined by an action decision algorithm that includes proportional and differential actions. The proportional term is proportional to the magnitude of the real-time speed difference to ensure response speed, while the differential term is proportional to the rate of change of the real-time speed difference to predictively suppress fluctuation trends and provide a damping effect. After receiving this command, the servo valve of the hydraulic tensioning device 3 will precisely control the flow and pressure of the oil, causing the hydraulic tensioning device 3 to instantly and actively increase the dynamic tension by 10-30kN, thereby improving the local tensile stiffness.

[0064] It should be noted that when implementing this control strategy that combines proportional and derivative actions, the parameters usually need to be determined through on-site debugging to achieve the best balance between response speed and stability. Applications similar to fuzzy PID control can enhance its adaptive capabilities.

[0065] Throughout the entire collaborative anti-interference program, the system continuously monitors whether the real-time speed difference returns to within the static threshold. Simultaneously, a high-frequency tension sensor mounted on the hydraulic tensioning device 3 monitors whether the tail tension has stabilized. If both conditions are met, the collaborative control program uses a ramp function to smoothly cancel the additional control inputs, gradually and smoothly restoring the additional tension compensation and drive acceleration correction to the normal operating mode. If the conditions are not met or the effect is not as expected, the system gradually increases the gain coefficient in the action decision algorithm to enhance the control effect.

[0066] In addition, during the coordinated control process, if the speed difference not only does not decrease but increases and even exceeds the upper limit of the dynamic speed difference threshold, the system will determine that the control has failed or the judgment is incorrect. At this time, the system will interrupt the anti-interference control, determine it as a real fault, and trigger emergency braking.

[0067] The above description is only a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope of the technology disclosed in the present invention, based on the technical solution and inventive concept of the present invention, should be covered within the scope of protection of the present invention.

Claims

1. A method for identifying abnormal operation of mining electromechanical equipment, characterized in that, include: A multi-source information synchronous acquisition architecture is constructed to obtain underlying synchronous data with a unified timestamp. The underlying synchronous data includes operating parameters, tail tension force, and original weighing signal. The original weighing signal is decoupled and reconstructed using digital signal processing algorithms to extract the instantaneous linear load equivalent that reflects the actual material distribution. A viscoelastic model of the conveyor belt is introduced, and combined with adaptively updated equipment parameters, a dynamic speed difference threshold is generated. The feature identification logic is executed in a multivariate coupling manner. It integrates the actual speed difference, dynamic speed difference threshold, static threshold and multiple parameter features in the underlying synchronization data to distinguish between false slippage and real abnormal operating state.

2. The method for identifying abnormal operation of mining electromechanical equipment according to claim 1, characterized in that, Obtain underlying synchronized data with a unified timestamp, specifically including: The angular displacement fed back by the absolute encoder installed on the shaft end of the drive drum and the tail redirecting drum is acquired through high-speed industrial real-time Ethernet and converted into linear velocity. The operating parameters of the drive drum are obtained through the high-speed communication interface of the frequency converter, the tail tension force is obtained through the high-frequency tension sensor at the tail hydraulic tensioning device, and the original weighing signal is obtained through the dynamic weighing sensor array downstream of the receiving point. By using a timestamp alignment algorithm, the data sequence with standard intervals is reconstructed through interpolation compensation, thereby achieving alignment of multivariate time series.

3. The method for identifying abnormal operation of mining electromechanical equipment according to claim 1, characterized in that, Extracting the instantaneous linear load equivalent that reflects the actual material distribution, specifically including: A multi-level wavelet packet decomposition signal processing procedure is used to divide the original weighing signal into sub-signals of different frequency bands. High-frequency oscillation components identified as physical shocks and local jitter noise are filtered out by threshold denoising, and low-frequency components representing stable load changes are extracted. The extracted low-frequency components are subjected to inverse wavelet transform, and the reconstructed signal is synthesized and cleaned by zero-phase filtering. The amplitude integral of the reconstructed signal is divided by the belt-related parameters to obtain the true instantaneous line load equivalent.

4. The method for identifying abnormal operation of mining electromechanical equipment according to claim 1, characterized in that, Generating a dynamic speed difference threshold specifically includes: Establish a parameter adaptive matrix, and dynamically update the elastic modulus and viscous damping coefficient of the conveyor belt by reading ambient temperature sensor data and the cumulative running time of the equipment; Substitute the motor output torque, tail tension force, and instantaneous linear load equivalent into the viscoelastic model to solve for the maximum transient stress and the corresponding theoretical maximum allowable strain rate. Based on the theoretical maximum allowable strain rate and effective transmission length, the maximum theoretical speed difference allowed by the system is calculated, thereby generating the envelope of the dynamic speed difference threshold that converges smoothly with the system's operating state.

5. The method for identifying abnormal operation of mining electromechanical equipment according to claim 1, characterized in that, Distinguishing between false slippage and genuine abnormal operating conditions specifically includes: The underlying synchronous data is used to build a dynamically updated feature matrix, and the historical sequence of parameters is stored through a sliding time window. When the actual speed difference is between the upper limit of the static threshold and the dynamic speed difference threshold, the collaborative comparison program is triggered to analyze the state characteristics of the motor output torque and tail tension. If the motor output torque is within the preset rising or stable range, and the tail tension does not suddenly drop, it is determined to be a false slip caused by transient elastic elongation, and the stop signal is blocked. If the actual speed difference exceeds the upper limit of the dynamic speed difference threshold, or if there is a sudden and sharp drop in the motor output torque and tail tension, it is determined to be a real slippage or belt breakage accident, triggering emergency braking.

6. The method for identifying abnormal operation of mining electromechanical equipment according to claim 5, characterized in that, include: When a false slip is detected, a collaborative disturbance rejection procedure is initiated, which includes parallel execution of drive adjustment and tension compensation. A flexible adjustment command is issued at the drive end, and the acceleration set value is instantly reduced based on the feedforward algorithm, and the torque output of the frequency converter is finely adjusted and compensated. A pressure superposition command is issued to the hydraulic tensioning device at the tensioning end, and a motion decision algorithm including proportional and differential actions is used to actively increase the dynamic tension to improve the local tensile stiffness.

7. A system for identifying abnormal operation of mining electromechanical equipment, used to implement the method for identifying abnormal operation of mining electromechanical equipment as described in any one of claims 1 to 6, characterized in that, The identification system includes: The multi-source synchronous acquisition module is used to build a high-frequency synchronous acquisition architecture to acquire low-level synchronous data with a unified timestamp, including absolute displacement, tension force, operating parameters and weighing signals. The signal cleaning and reconstruction module is used to remove high-frequency interference from the weighing signal through wavelet packet decomposition and digital filtering algorithms, and to calculate the true instantaneous line load equivalent. The dynamic threshold generation module is used to introduce a parameter-adaptive viscoelastic model, combine it with the real-time stress state to derive the theoretical maximum allowable strain rate, and generate a dynamic velocity difference threshold that evolves with the working conditions. The multivariate feature identification module is used to identify normal elastic deformation and real operational faults by comparing the relationship between the actual speed difference and the dynamic speed difference threshold and static threshold based on the feature matrix and the multidimensional state trend within the sliding time window.

8. The identification system for abnormal operation of mining electromechanical equipment according to claim 7, characterized in that, In the multi-source synchronous acquisition module: The mechanical interface of the absolute encoder that acquires absolute displacement adopts a heavy-duty bearing structure, and the electrical output is a differential signal; The high-frequency tension sensor for collecting tension has an overload-resistant design to resist damage from instantaneous impacts; The weighing signal is acquired by a dynamic weighing sensor array covering the full width of the belt and uploaded after preliminary filtering by a local transmitter.

9. The identification system for abnormal operation of mining electromechanical equipment according to claim 7, characterized in that, The dynamic threshold generation module is equipped with a model verification and correction mechanism: Offline calibration is achieved by executing a calibration program under no-load conditions, collecting strain response data, and then back-calculating and updating the actual elastic modulus and viscous damping coefficient. The relationship between the actual speed difference and the dynamic speed difference threshold is continuously monitored, and the online model self-check program is triggered when the actual value continuously approaches the upper limit of the dynamic speed difference threshold.

10. The identification system for abnormal operation of mining electromechanical equipment according to claim 7, characterized in that, The system also includes an active cooperative anti-interference module, which is used for: When the multivariate feature identification module determines that the slippage is false, an electromechanical-hydraulic active cooperative anti-disturbance command is sent to the drive end and the tensioning end. Continuously monitor the actual speed difference and tail tension status. Once the operation returns to stability, the additional control quantity is smoothly canceled through the ramp function, allowing the equipment to return to the normal closed-loop operation mode. When the speed difference exceeds the upper limit of the dynamic speed difference threshold, the disturbance rejection control is interrupted and emergency braking is triggered.

Images