A hoist hydraulic station monitoring system and method with hierarchical early warning function

Abstract

The application discloses a hoist hydraulic station monitoring system with hierarchical early warning function and a method thereof, and relates to the technical field of mine hoist equipment safety monitoring. The system comprises a continuous liquid level sensing unit, which collects liquid level displacement to obtain a digital liquid level value; a logic control unit, which generates an action characteristic label corresponding to the current car operation stage through monitoring; a compensation liquid level value is obtained based on the action characteristic label; a residual safety time is obtained based on the compensation liquid level value; an analog signal is identified to determine whether there is a constant signal fault and / or a slow drift fault, and if there is, the logic control unit cuts the safety relay coil excitation, causing the normally closed contact to physically disconnect, and forcibly cutting off the power supply of the hoist safety chain to lock it, thereby solving the problems of high false alarm rate of fixed liquid level threshold, inability to predict the residual safety time of the hydraulic station, inability to identify the gradual failure of the liquid level sensor and insufficient safety protection reliability of the existing hydraulic station liquid level monitoring system.

Description

Technical Field

[0001] This invention relates to the field of safety monitoring technology for mine hoisting equipment, specifically to a monitoring system and method for a hoisting hydraulic station with graded early warning function. Background Technology

[0002] Mine hoists are crucial equipment for transporting personnel and materials underground in both coal and non-coal mines, and their operational safety is directly related to mine production safety. The hoist's operating system typically uses a hydraulic station as the power actuator, providing hydraulic power to devices such as the car stopper, rocker platform, and brake. The hydraulic oil level in the hydraulic station directly affects the normal operation of the hydraulic system. When the oil tank leaks or the level is too low, the hydraulic actuators can easily malfunction, and in severe cases, it can even lead to the hoist going out of control, posing a significant safety hazard.

[0003] Existing hydraulic station level monitoring systems typically use float switches or single-point level switches for monitoring, triggering an alarm or interlock when the level drops to a fixed threshold. This type of technology has at least the following drawbacks: 1) Using a fixed liquid level threshold for alarm does not take into account the periodic transfer of oil between the actuator and the oil tank during the operation, which is prone to false alarms and false lockouts. 2) Judging solely based on the current liquid level value cannot analyze the trend of hydraulic oil leakage, nor can it predict the remaining safe operating time of the system in advance; 3) Using a fixed-window slope calculation method to estimate the leakage rate makes it difficult to balance response speed and noise resistance, resulting in low leakage prediction accuracy. 4) After long-term operation, liquid level sensors are prone to progressive failures such as jamming, drifting, scaling and temperature drift. Existing technologies can usually only identify sudden faults such as wire breakage, but cannot detect slow degradation faults. 5) Existing liquid level monitoring systems usually use the liquid level signal as the sole criterion for judgment, lacking a cross-verification mechanism between multiple physical quantities. Once the liquid level sensor itself fails, it is very easy to cause misjudgment or even lose the safety protection function. Summary of the Invention

[0004] To address the problems of existing hydraulic station level monitoring systems, such as high false alarm rates with fixed level thresholds, inability to predict remaining safe time, inability to identify progressive failure of level sensors, and insufficient reliability of safety protection, this invention provides a hoist hydraulic station monitoring system with tiered early warning capabilities. The system includes: A continuous liquid level sensing unit is installed in the hydraulic station oil tank of the hoist operator's car. It is used to collect the liquid level displacement to obtain an analog signal, perform analog-to-digital conversion on the analog signal to obtain a digital liquid level signal, and obtain a digital liquid level value based on the digital liquid level signal. The logic control unit has its analog input terminal connected to the continuous liquid level sensing unit, and internally includes a working condition sensing module, a dynamic threshold module, a leakage rate estimation and remaining safe time prediction module, and a signal morphology diagnosis module. The interlocking execution unit, controlled by the logic control unit, includes a safety relay with a forced-guide contact. The normally closed contact of the safety relay is connected in series to the safety chain circuit of the mine hoist electrical control system by physical hard wiring. The working condition perception module is used to identify the current operating action stage by monitoring the control signal status output by the logic control unit to the operating hydraulic valve group, and generate an action feature label corresponding to the current operating action stage. The dynamic threshold module is used to call a dynamic volume compensation coefficient that matches the action feature label within a time window based on the action feature label, and to perform dynamic volume change compensation on the digital liquid level value to obtain a compensated liquid level value. The logic control unit is configured with a first preset threshold L1 and a second preset threshold L2, and the first preset threshold L1 is greater than the second preset threshold L2. The leakage rate estimation and remaining safe time prediction module is used to calculate the current leakage rate based on the compensation liquid level value of multiple consecutive sampling periods using sliding window linear regression, and to predict the remaining safe time for the liquid level to drop from the digital liquid level value to the second preset threshold L2 based on the current leakage rate. The signal morphology diagnostic unit is used to extract morphological features and recognize time series patterns of the analog signal to determine whether there is a constant signal fault and / or a slow drift fault. When any fault is determined to exist, the logic control unit cuts off the excitation of the safety relay coil, causing the normally closed contact to physically open and forcing the hoist safety chain to lose power and lock out.

[0005] This system constructs a multi-layered safety monitoring system: a continuous level sensing unit collects the oil tank level displacement in real time and converts it into a digital level value; the logic control unit identifies the current operating status of the hydraulic actuator through the working condition perception module and compensates for volume changes in the level signal based on the dynamic threshold module, thereby obtaining the true level value and reducing the level fluctuation error caused by the hydraulic actuator's operation; the leakage rate estimation and remaining safe time prediction module analyzes the level change trend and predicts the remaining safe operating time of the hydraulic system, enabling early prediction of the remaining safe operating time of the hydraulic station; simultaneously, the signal morphology diagnostic module monitors the health status of the level sensor itself in real time, identifying abnormalities such as sensor jamming and drift, and can identify progressive failure of the level sensor; finally, based on the level status and diagnostic results, the system drives the safety relay to achieve graded early warning and hardware-level safety interlocking. The hardware power failure interlocking method improves the fault safety level of the system, realizes the integration of level monitoring, leakage prediction and safety interlocking, and significantly improves the intrinsic safety level of the hoist hydraulic station.

[0006] Furthermore, the input port of the working condition sensing module is connected in parallel to the intermediate relay auxiliary contact of the digital output module of the logic control unit, and the working condition sensing module has a built-in truth table; The operating condition perception module identifies the current vehicle operation stage based on the control signal state and the truth table, and generates an action feature label corresponding to the current vehicle operation stage.

[0007] The input of the working condition sensing module is connected in parallel to the auxiliary contact of the PLC digital output module. By reading the status of the solenoid valve control signal in real time and combining it with the pre-established truth table, the current action stage of the hydraulic actuator is deduced in reverse. The current working condition can be accurately identified without the need for additional position sensors, reducing the system implementation cost; and providing an accurate working condition basis for subsequent dynamic compensation.

[0008] Furthermore, the logic control unit has a built-in vehicle operation action feature signal library, which includes at least brake release action tags, brake holding action tags, brake extend action tags, brake retract action tags, rocker lift action tags, rocker lower action tags, and static standby tags. Each action tag corresponds to a tag data, and the tag data includes an enable logic matrix, a standard time-domain waveform, a nonlinear compensation coefficient matrix, a follow-up offset, and an adaptive correction factor. The dynamic threshold module is specifically used for: The action feature tags are matched with the vehicle operation action feature signal library to obtain the target action tag and target tag data, and a synchronization trigger signal is generated and an internal timer is started. Based on the current timing value of the internal timer and the standard time-domain waveform of the target tag data, the per-unit invasion depth corresponding to the current timing value is obtained, and the predicted liquid level fluctuation amount is obtained based on the per-unit invasion depth and the preset maximum liquid level fluctuation amplitude. Based on the digital liquid level value and the action speed parameter of the hydraulic actuator, interpolation is performed on the nonlinear compensation coefficient matrix of the target label data to obtain the dynamic volume compensation coefficient. The predicted liquid level fluctuation, the dynamic volume compensation coefficient, and the adaptive correction factor of the target tag data are coupled and calculated to obtain the dynamic compensation amount. The digital liquid level value is corrected based on the dynamic compensation amount to obtain the compensated liquid level value of the current timing value. Based on the compensation liquid level value and the follow-up offset of the target tag data, the first preset threshold L1 and the second preset threshold L2 are dynamically offset to obtain the third dynamic threshold L3 and the fourth dynamic threshold L4 of the current timing value, respectively; the first preset threshold L1 and the second preset threshold L2 are updated to the third dynamic threshold L3 and the fourth dynamic threshold L4, respectively. When the target action tag disappears, the dynamic compensation amount is gradually attenuated according to the falling edge curve of the standard time domain waveform of the target tag data, and the dynamic compensation amount is smoothly transitioned to zero within a preset ebb tide time to maintain the continuity of the compensation liquid level value.

[0009] When operating conditions change, the system matches action tags, predicts the current level disturbance based on standard time-domain waveforms, obtains compensation coefficients from tables based on tank level and action speed, and calculates dynamic compensation amounts. This effectively reduces false level fluctuations caused by vehicle operation actions, ensuring that the level value accurately reflects the remaining oil in the tank. The system also dynamically corrects alarm thresholds to improve their rationality. After the action ends, the system smoothly retracts compensation according to the falling edge curve, achieving feedforward prediction + nonlinear compensation + adaptive correction, thus improving the system's adaptability to different operating conditions.

[0010] Furthermore, the first calculation formula for obtaining the compensated liquid level value is: ; in, Indicates the compensation liquid level value. Indicates the digital liquid level value. This indicates the predicted amount of liquid level fluctuation. This represents the dynamic volume compensation coefficient. This represents the adaptive correction factor.

[0011] Furthermore, the leakage rate estimation and remaining safe time prediction module is specifically used for: The compensation change rate is obtained based on the compensation liquid level value; If the target action tag is a static standby tag, or if the compensation change rate is within a preset range for a preset time, then the sampling point corresponding to the compensation liquid level value is determined to be a valid sampling point, and the valid sampling point is stored in the regression data pool. Based on the regression data pool, when the range of the compensation liquid level value within multiple consecutive sampling periods is less than the preset dead zone threshold, the first effective leakage sampling point is determined. Establish an effective leakage sampling window, and initialize the basic window width and basic step size of the effective leakage sampling window; Obtain the residual standard deviation of the compensated liquid level value within the effective leakage sampling window, and update the base window width based on the residual standard deviation; Each effective leakage sampling point within the effective leakage sampling window is assigned an exponential decay weight according to its time sequence, and a weighted least squares method is used for linear fitting to obtain the current leakage rate. The base step size is then updated based on the current leakage rate. The leakage acceleration is calculated based on the current leakage rate and the historical leakage rate obtained from the previous effective leakage sampling window. If the leakage acceleration is greater than 0, the leakage acceleration is smoothed to obtain the emergency expansion factor. Based on the geometric model of the oil tank, the cross-sectional area function corresponding to the digital liquid level value is obtained. Based on the compensation liquid level value of the current timing value, the second preset threshold L2, the current leakage rate and the emergency expansion factor, the remaining safe time integral model is established. The remaining safety time integral model is solved iteratively using a numerical integration method to calculate the remaining safety time corresponding to the liquid level dropping from the digital liquid level value to the second preset threshold L2; The predicted standard deviation of the remaining safety time is obtained based on the regression fitting residual, and the predicted value of the remaining safety time and the predicted confidence interval are obtained based on the predicted standard deviation.

[0012] By establishing an effective leakage sampling window and eliminating non-leakage disturbance data generated by the hydraulic actuator's actions, only valid sampling points from the free leakage phase are retained. An adaptive variable-window sliding regression and a weighted least squares regression algorithm with a forgetting factor are used to dynamically fit the liquid level decline trend to obtain the current leakage rate. The regression window length and step size are dynamically adjusted based on the residual standard deviation to balance response speed and noise resistance. Simultaneously, leakage acceleration is introduced to describe the leakage deterioration trend, and a remaining safety time integral model is established in conjunction with the tank's geometric model. The actual time required for the liquid level to drop to the safety threshold is inverted using numerical integration methods, achieving physical constraint prediction of the hydraulic system's remaining safety time. This enables real-time prediction of hydraulic oil leakage trends and remaining safe operating time, as well as multi-domain joint analysis integrating physical and statistical laws, improving the accuracy of leakage rate estimation and remaining safety time prediction, and enhancing sensitivity to sudden leaks and accelerated leakage deterioration.

[0013] Furthermore, the second formula for calculating the exponentially decaying weight is as follows: ; The third calculation formula for the remaining safety time integral model is: ; in, Indicates the first The exponential decay weight of each effective leakage sampling point This indicates a pre-defined forgetting factor. Indicates the current moment. Indicates the first Sampling time of each effective leak sampling point Indicates the number of the valid leak sampling point. Indicates the remaining safe time. This indicates the compensation liquid level value for the current timing. This indicates the second preset threshold. Indicates digital liquid level value The corresponding cross-sectional area function, Indicates the current leakage rate. Indicates the emergency expansion factor. Indicates time, This indicates the digital liquid level value.

[0014] By utilizing exponential forgetting weights to reduce the contribution of historical data to prediction results, a nonlinear residual time model is established to quickly respond to changes in leakage and improve prediction accuracy under complex tank structures.

[0015] Furthermore, the signal morphology diagnostic unit is specifically used for: Within a preset diagnostic time window, the variance of the sampled value of the analog signal is calculated. If the variance of the sampled value is less than a preset variance threshold and exceeds a preset constant time threshold within a first preset duration, then a signal constant fault is determined to exist. Calculate the rate of change of the sliding average value of the analog signal within a preset drift detection window. If the absolute value of the rate of change of the sliding average value is greater than the preset drift rate threshold within a second preset duration, and the absolute value is less than the disconnection judgment threshold, then a slow drift fault is determined to exist. When any fault is detected, the logic control unit cuts off the energization of the safety relay coil, causing the normally closed contact to physically open, thus forcing the hoist safety chain to lose power and lock out.

[0016] By using analysis of variance to identify frozen liquid level signals and by using the moving average rate of change to identify slow liquid level drift, we can identify both stuck and slow drift faults, improve the reliability of liquid level data, and reduce scenarios where sensor failure leads to false locking.

[0017] Furthermore, the signal morphology diagnostic unit is also used for: Construct a theoretical mapping relationship between the direction of volume change and the direction of pressure change of the hydraulic actuator, and generate the theoretical direction of liquid level change and the theoretical direction of pressure change; Acquire the pressure signal from the pressure sensor; Within a preset sliding time window, the fluctuation characteristic parameters of the digital liquid level signal and the pressure signal are calculated, and the normalized activity index is calculated based on the full scale of the fluctuation characteristic parameters to obtain the first index and the second index respectively. If, within a preset duration, the first index remains less than the first activity threshold and the second index remains greater than the second activity threshold, then it is determined that the continuous liquid level sensing unit has a signal freezing fault. Based on the rate of change of the digital liquid level signal and the pressure signal, a liquid level-pressure phase plane trajectory is constructed, the instantaneous motion direction angle of the liquid level-pressure phase plane trajectory is obtained, the instantaneous motion direction angle is compared with the theoretical liquid level change direction, and when the directions of the two are inconsistent, it is determined that the continuous liquid level sensing unit has a slow drift fault. The virtual liquid level value is calculated based on the standard liquid level-pressure correspondence table and the pressure signal. The residual between the virtual liquid level value and the digital liquid level value is obtained. The residual is subjected to a cumulative sum test. When the cumulative residual exceeds the preset residual threshold, it is determined that the continuous liquid level sensing unit has a nondeterministic failure. Based on the signal freezing fault determination results, the slow drift fault determination results, and the nondeterministic failure results, the nondeterministic failure confidence level is calculated. When the nondeterministic failure confidence level is greater than a preset global threshold, the logic control unit is controlled to execute a graded response.

[0018] By introducing pressure signals as a heterogeneous reference frame for liquid level signals, a physical causal relationship model between liquid level and pressure is constructed. Utilizing wave-activity cross-validation, liquid level-pressure phase plane orientation consistency verification, and virtual liquid level redundancy calculation based on pressure-volume mapping, a comprehensive diagnosis of liquid level sensor jamming, slow drift, and nondeterministic failures is performed. The sensor health status is output through an accumulation-sum algorithm and a confidence fusion mechanism. This achieves online identification of progressive degradation and metastable failures of liquid level sensors, as well as collaborative decision-making for hydraulic system state prediction and sensor health diagnosis. This improves the credibility and reliability of liquid level monitoring data, the system's fault tolerance to sensor faults and false alarms, and enhances the intrinsic safety level and safety integrity grade of the hydraulic station monitoring system.

[0019] Furthermore, when it is determined that the compensation liquid level value is less than the first preset threshold L1 and greater than the second preset threshold L2, the logic control unit drives an audible and visual alarm device independent of the safety circuit. When the compensation liquid level value is determined to be less than the second preset threshold L2, the logic control unit cuts off the excitation of the safety relay coil, causing the normally closed contact to physically disconnect, and forcing the hoist safety chain to lose power and lock.

[0020] Establish a two-level safety response mechanism, with alarm circuits and safety circuits designed independently to avoid direct shutdowns affecting production, give maintenance personnel time to handle the situation, and prevent further drops in liquid level from causing accidents; thus balancing safety and production continuity.

[0021] The present invention also provides a monitoring method for a hoist hydraulic station with a graded early warning function, the method comprising: A continuous level sensor unit installed in the hydraulic station oil tank of the hoisting machine acquires an analog signal by collecting the level displacement, performs analog-to-digital conversion on the analog signal to obtain a digital level signal, and obtains a digital level value based on the digital level signal. Monitor the status of the control signal output by the logic control unit to the hydraulic valve group, identify the current stage of the vehicle operation, and generate an action feature label corresponding to the current stage of the vehicle operation. Based on the action feature label, within the time window, the dynamic volume compensation coefficient matching the action feature label is called to perform dynamic volume change compensation on the digital liquid level value to obtain the compensated liquid level value. A first preset threshold L1 and a second preset threshold L2 are preset, and the first preset threshold L1 is greater than the second preset threshold L2; Based on the compensated liquid level value of multiple consecutive sampling periods, the current leakage rate is calculated using sliding window linear regression, and the remaining safe time for the liquid level to drop from the digital liquid level value to the second preset threshold L2 is predicted based on the current leakage rate. The analog signal is subjected to morphological feature extraction and time series pattern recognition to determine whether there is a constant signal fault and / or slow drift fault. When either fault is determined to exist, the logic control unit cuts off the excitation of the safety relay coil, causing the normally closed contact to physically open and forcing the hoist safety chain to lose power and lock out.

[0022] The principle and effect of this method are similar to those of this system, and will not be elaborated further.

[0023] One or more technical solutions provided by this invention have at least the following technical effects or advantages: 1. The current operating status of the vehicle is identified by the working condition perception module, and the liquid level signal is corrected by the dynamic volume compensation algorithm, which effectively reduces the impact of liquid level fluctuation caused by the action of the hydraulic actuator and significantly reduces the false alarm rate. 2. By establishing a database of engine operation action characteristic signals, feedforward prediction and dynamic threshold adjustment of liquid level changes were achieved, making the liquid level monitoring results more consistent with the actual hydraulic state; 3. By using the adaptive variable window length sliding regression algorithm and the weighted least squares algorithm with forgetting factor, the leakage rate can be obtained in real time, taking into account both prediction sensitivity and noise resistance. 4. By introducing leakage acceleration and a nonlinear geometric model of the oil tank, a physical-statistical coupled remaining safety time prediction model is established, which significantly improves the accuracy of remaining safety time prediction; 5. By verifying the consistency of heterogeneous sensors for liquid level and pressure signals, a redundant diagnostic mechanism based on physical causality was constructed, which can identify freezing faults, slow drift faults, and progressive degradation faults of liquid level sensors. 6. By setting a first preset threshold and a second preset threshold, graded responses to audible and visual alarms and safety interlocks are achieved, thereby improving the safety protection capability of the hydraulic station; 7. By hard-wiring the normally closed contacts of the safety relay in series with the safety chain circuit of the hoist, hardware-level power failure interlocking under fault conditions is realized, which improves the system's fault safety and intrinsic safety level. Attached Figure Description

[0024] The accompanying drawings, which are provided to further illustrate embodiments of the invention and constitute a part of this invention, are not intended to limit the scope of the invention. Figure 1 This is a flowchart illustrating a monitoring system for a hoist hydraulic station with tiered early warning capabilities. Detailed Implementation

[0025] To better understand the above-mentioned objectives, features, and advantages of the present invention, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments. It should be noted that, where there is no conflict, the embodiments of the present invention and the features thereof can be combined with each other.

[0026] Many specific details are set forth in the following description in order to provide a full understanding of the invention. However, the invention may also be practiced in other ways different from those described herein, and therefore the scope of protection of the invention is not limited to the specific embodiments disclosed below.

[0027] Example 1 refer to Figure 1 This embodiment provides a monitoring system for a hoist hydraulic station with a graded early warning function. The system includes: A continuous liquid level sensing unit is installed in the hydraulic station oil tank of the hoist operator's car. It is used to collect the liquid level displacement to obtain an analog signal, perform analog-to-digital conversion on the analog signal to obtain a digital liquid level signal, and obtain a digital liquid level value based on the digital liquid level signal. In this embodiment, the continuous liquid level sensing unit can be an existing intrinsically safe magnetostrictive displacement sensor, installed on the top of the oil tank, with the probe vertically inserted into the oil. The sensor integrates an analog-to-digital conversion circuit, which converts the analog signal into a digital liquid level signal and then uploads it to the logic control unit via a standard industrial bus protocol (such as Modbus RTU or Profibus PA).

[0028] The logic control unit has its analog input terminal connected to the continuous liquid level sensing unit, and internally includes a working condition sensing module, a dynamic threshold module, a leakage rate estimation and remaining safe time prediction module, and a signal morphology diagnosis module. In this embodiment, the logic control unit can be a safety-type programmable logic controller (PLC, such as Siemens S7-1500F or Rockwell CompactLogix L43S), which integrates a condition sensing module, a dynamic threshold module, a leakage rate estimation and remaining safe time prediction module, and a signal morphology diagnostic module. The above modules are stored in the program storage area of ​​the PLC in the form of function blocks or structured text programs.

[0029] The interlocking execution unit, controlled by the logic control unit, includes a safety relay with a forced-guide contact (such as Pilz PNOZ s4 or Siemens 3SK1 series). The normally closed contact of the safety relay is connected in series in the safety chain circuit of the mine hoist electrical control system by physical hard wiring. The working condition perception module is used to identify the current operating action stage by monitoring the control signal status output by the logic control unit to the operating hydraulic valve group, and generate an action feature label corresponding to the current operating action stage. The input port of the working condition sensing module is connected in parallel to the intermediate relay auxiliary contact of the digital output module of the logic control unit, and the working condition sensing module has a built-in truth table. The operating condition perception module identifies the current vehicle operation stage based on the control signal state and the truth table, and generates an action feature label corresponding to the current vehicle operation stage.

[0030] In this embodiment, the truth table is a mapping matrix between each action tag and the combined state of the solenoid valve control signal.

[0031] When the operating condition sensing module detects a change in the combination of control signals, it executes the following steps: Step A: Compare the current control signal combination state with the previous state using differential analysis, and detect rising edge and falling edge events; Step B: Match the current stable combined state code with the truth table. If the match is successful, output the corresponding action feature label; if the match fails, generate an unknown action label, the system maintains the previous valid label unchanged and records the abnormal event; Step C: To prevent accidental tag switching caused by solenoid valve contact jitter, the module has a built-in anti-jitter timer. When a new combination state is matched, the anti-jitter timer is started (preset 50ms). The current action feature tag is only officially updated if the combination state remains stable within the anti-jitter timer; otherwise, it reverts to the previous tag. Step D: After the action feature label is generated, it is output to the dynamic threshold module to trigger volume compensation calculation, and recorded in the internal event log for subsequent fault tracing.

[0032] Taking the brake release action as an example: When the operator issues a brake release command by operating the hoist handle, the PLC's digital output module Q0.1 outputs a 24V high level to energize the brake release solenoid valve. The operating condition sensing module scans and detects that the state of Q0.1 changes from 0 to 1 and remains stable for more than 50ms. That is, it generates a brake release action tag according to the truth table [Q0.1=1, Q0.2=0, Q0.3=0], and records the system timestamp of the tag generation time as a timing reference point.

[0033] The dynamic threshold module is used to call a dynamic volume compensation coefficient that matches the action feature label within a time window based on the action feature label, and to perform dynamic volume change compensation on the digital liquid level value to obtain a compensated liquid level value. The logic control unit is configured with a first preset threshold L1 and a second preset threshold L2, and the first preset threshold L1 is greater than the second preset threshold L2. The logic control unit has a built-in vehicle operation action feature signal library. The vehicle operation action feature signal library includes at least the following tags: brake release action tag, brake holding action tag, brake extend action tag, brake retract action tag, rocker lift action tag, rocker lower action tag, and static standby tag. Each action tag corresponds to a tag data. The tag data includes an enable logic matrix, a standard time domain waveform, a nonlinear compensation coefficient matrix, a follow-up offset, and an adaptive correction factor. Enable logic matrix: Records stable combination state codes of each solenoid valve control signal when a vehicle operation action (such as brake release, brake engagement, brake extension, brake retraction, rocker lift, rocker lower) is activated. For example, brake release corresponds to [valve A=1, valve B=0, valve C=0], and brake extension corresponds to [valve A=0, valve B=1, valve C=0]. This matrix also includes an anti-jitter confirmation time (e.g., 50ms), meaning that the combined state must continue for more than this time threshold before the action is confirmed as activated.

[0034] Standard time-domain waveform: The extracted motion waveform is normalized in time (mapping the motion cycle to the 0-1 interval) and normalized in amplitude (based on the maximum fluctuation amplitude) to generate a waveform curve represented by the per-unit damage depth, which is stored as a discrete point sequence, and linear interpolation is used between adjacent points.

[0035] Nonlinear compensation coefficient matrix: Under combined operating conditions of different initial liquid level heights (e.g., 200mm, 400mm, 600mm, 800mm) and different action speeds (achieved through fast / medium / slow settings by adjusting valve port current), each action is executed separately, and the actual liquid level fluctuation amplitude is recorded. This is compared with the nominal fluctuation amplitude under standard conditions to form a two-dimensional compensation coefficient lookup table. Each entry in the table represents the ratio of the measured amplitude to the nominal amplitude.

[0036] Follow-up offset: The threshold dynamic rise value corresponding to each action is set according to 50% of the maximum liquid level fluctuation amplitude caused by each action. For example, if the maximum fluctuation amplitude of the brake release action is 65mm, then its follow-up offset is 32.5mm.

[0037] Adaptive correction factor: A weighting coefficient that can be erased and rewritten online, used to correct for actual stroke deviations caused by oil temperature and pump wear.

[0038] The dynamic threshold module is specifically used for: The motion feature labels are matched with the vehicle operation motion feature signal library to obtain target motion labels and target label data (including standard time domain waveform, nonlinear compensation coefficient matrix, follow-up offset and adaptive correction factor) and a synchronization trigger signal is generated and an internal timer is started. This timer is used to determine the phase position currently in the motion cycle. Based on the current timing value of the internal timer and the standard time-domain waveform of the target tag data, the per-unit invasion depth corresponding to the current timing value is obtained. Based on the per-unit invasion depth and the preset maximum liquid level fluctuation amplitude, the predicted liquid level fluctuation amount is obtained, that is, the predicted liquid level fluctuation amount = per-unit invasion depth * preset maximum liquid level fluctuation amplitude. In this embodiment, the per-unit damage depth refers to the ratio of the degree to which the fuel level deviates from the nominal value at a certain moment during the operation to the maximum fuel level fluctuation amplitude of the operation, with a value ranging from 0 to 1. This parameter describes the instantaneous degree of disturbance to the fuel tank level caused by the operation. For example, the value is 0 at the beginning of the operation, 1 at the moment of maximum fuel level fluctuation, and gradually returns to zero during the retraction process after the operation ends. This proportionalization process allows fuel level fluctuations of different amplitudes (full fuel level / low fuel level) under the same operation label to be described using a unified waveform template.

[0039] Based on the digital liquid level value and the action speed parameters of the hydraulic actuator (which can be obtained by monitoring the PWM duty cycle of the solenoid valve drive current or the valve core position feedback signal, quantifying the speed of the current action), interpolation is performed in the nonlinear compensation coefficient matrix of the target label data to obtain the dynamic volume compensation coefficient; if (digital liquid level value, action speed parameter) is used as the query coordinate, two-dimensional interpolation is performed in the nonlinear compensation coefficient matrix. First, two adjacent calibrated liquid level heights are found in the dimension of digital liquid level value, and two adjacent calibrated action speeds are found in the dimension of action speed parameter. Then, the existing bilinear interpolation formula is used to calculate the dynamic volume compensation coefficient corresponding to the current working condition.

[0040] The predicted liquid level fluctuation, the dynamic volume compensation coefficient, and the adaptive correction factor of the target tag data are coupled and calculated to obtain the dynamic compensation amount. The digital liquid level value is corrected based on the dynamic compensation amount to obtain the compensated liquid level value of the current timing value. The first calculation formula for obtaining the compensation liquid level value is: ; in, Indicates the compensation liquid level value. Indicates the digital liquid level value. This indicates the predicted amount of liquid level fluctuation. This represents the dynamic volume compensation coefficient. Indicates the adaptive correction factor. This refers to the dynamic compensation amount.

[0041] When the action causes a drop in liquid level (such as brake release or actuator extension), A positive value indicates the compensation level. Higher than digital liquid level value This replenishes the false drop in oil level caused by the engine operation drawing away fluid; when the operation causes the fluid level to rise (such as when the brake is engaged or the actuator retracts, causing fluid to flow back into the tank), A negative value indicates a compensation level. Below digital liquid level This effectively reduces the false rise in fluid level caused by oil backflow. Through this compensation, the fluid level value is adjusted. It remains relatively stable throughout the entire cycle of engine operation, accurately reflecting the actual amount of fuel in the tank.

[0042] Based on the compensated liquid level value and the follow-up offset of the target tag data, the first preset threshold L1 and the second preset threshold L2 are dynamically offset to obtain the third dynamic threshold L3 and the fourth dynamic threshold L4 of the current timing value, respectively; the first preset threshold L1 and the second preset threshold L2 are updated to the third dynamic threshold L3 and the fourth dynamic threshold L4, respectively; that is, the third dynamic threshold L3 = the first preset threshold L1 + the follow-up offset, and the fourth dynamic threshold L4 is obtained in the same way.

[0043] When the target action tag disappears, the dynamic compensation amount is gradually attenuated according to the falling edge curve of the standard time domain waveform of the target tag data, and the dynamic compensation amount is smoothly transitioned to zero within a preset ebb tide time to maintain the continuity of the compensation liquid level value.

[0044] The leakage rate estimation and remaining safe time prediction module is used to calculate the current leakage rate based on the compensation liquid level value of multiple consecutive sampling periods using sliding window linear regression, and to predict the remaining safe time for the liquid level to drop from the digital liquid level value to the second preset threshold L2 based on the current leakage rate; this soft landing method avoids the jump in compensation liquid level value caused by sudden changes in compensation amount, and maintains data continuity.

[0045] Specifically, the leakage rate estimation and remaining safe time prediction module is used for: The compensation change rate (i.e., dP / dt, which can be obtained by calculating the difference between adjacent sampling points) is obtained based on the compensation liquid level value. If the target action tag is a static standby tag, that is, no vehicle operation action is being performed, or the compensation change rate is continuously within the preset range within a preset time, such as within 5 seconds, the absolute value of the compensation change rate is continuously less than 0.5 mm / s, indicating that the liquid level is in a stable change state, then the sampling point corresponding to the compensation liquid level value is determined to be a valid sampling point, and the valid sampling point is stored in the regression data pool. Based on the regression data pool, when the range of the compensation liquid level value within multiple consecutive sampling periods (such as 5 consecutive sampling points) is less than the preset dead zone threshold, it is considered that the system has entered the steady-state free leakage stage, and the current sampling point is determined as the first effective leakage sampling point. Establish an effective leakage sampling window, and initialize the basic window width and basic step size of the effective leakage sampling window; Obtain the residual standard deviation of the compensated liquid level values ​​within the effective leakage sampling window (i.e., the standard deviation of the vertical deviation of each sampling point relative to the fitted straight line), and update the base window width based on the residual standard deviation; if the residual standard deviation is <0.1mm (signal smoothing), the module shortens the window width to 10 sampling points to quickly respond to changes in the leakage rate; if the residual standard deviation is >0.5mm (signal jitter is large), the module extends the window width to 50 sampling points to suppress random noise using the large number averaging effect; if the residual standard deviation is between the two, the base window width is maintained.

[0046] Each valid leakage sampling point within the effective leakage sampling window is assigned an exponentially decaying weight according to its time sequence, and the existing weighted least squares method is used for linear fitting to obtain the current leakage rate. The base step size is updated based on the current leakage rate. If the leakage rate changes rapidly, the step size is shortened to 1 sampling point (updated point by point). If the leakage rate is stable, the step size is widened to 10 sampling points (to reduce the computational load). The leakage acceleration is calculated based on the current leakage rate and the historical leakage rate obtained from the previous effective leakage sampling window. That is, leakage acceleration = (current leakage rate - historical leakage rate) / Δt, where Δt is the time interval between the two windows. If the leakage acceleration is greater than 0, the leakage acceleration is smoothed to obtain the emergency expansion factor. Based on the existing fuel tank geometric model, obtain the cross-sectional area function (i.e., the horizontal cross-sectional area of ​​the fuel tank) corresponding to the digital liquid level value. Based on the compensation liquid level value of the current timing value, the second preset threshold L2, the current leakage rate and the emergency expansion factor, establish the remaining safe time integral model. The remaining safety time integral model is iteratively solved using numerical integration methods (existing adaptive step size Simpson integral method or fourth-order Runge-Kutta method) to calculate the remaining safety time corresponding to the liquid level dropping from the digital liquid level value to the second preset threshold L2; The predicted standard deviation of the remaining safety time is obtained based on the regression fitting residuals. For example, the root mean square error of the regression fitting residuals is calculated and passed to the remaining safety time integral model. The predicted standard deviation is calculated by Monte Carlo simulation or by the error propagation method based on the first-order Taylor expansion. Based on the predicted standard deviation, the predicted remaining safety time and the predicted confidence interval are obtained (e.g., [predicted remaining safety time, ±1.96 * predicted standard deviation]).

[0047] The second formula for obtaining the exponential decay weight is as follows: ; The third calculation formula for the remaining safety time integral model is: ; in, Indicates the first The exponential decay weight of each effective leakage sampling point This indicates a pre-defined forgetting factor. Indicates the current moment. Indicates the first Sampling time of each effective leak sampling point Indicates the number of the valid leak sampling point. Indicates the remaining safe time. This indicates the compensation liquid level value for the current timing. This indicates the second preset threshold. Indicates digital liquid level value The corresponding cross-sectional area function, Indicates the current leakage rate. Indicates the emergency expansion factor. Indicates time, This indicates the digital liquid level value.

[0048] The signal morphology diagnostic unit is used to extract morphological features and recognize time series patterns of the analog signal to determine whether there is a constant signal fault and / or a slow drift fault. When any fault is determined to exist, the logic control unit cuts off the excitation of the safety relay coil, causing the normally closed contact to physically open and forcing the hoist safety chain to lose power and lock out.

[0049] Specifically, the signal morphology diagnostic unit is used for: Within a preset diagnostic time window, the variance of the sampled value of the analog signal is calculated. If the variance of the sampled value is less than a preset variance threshold and exceeds a preset constant time threshold within a first preset duration, then a signal constant fault is determined to exist. Calculate the rate of change of the sliding average value of the analog signal within a preset drift detection window. If the absolute value of the rate of change of the sliding average value is greater than the preset drift rate threshold within a second preset duration, and the absolute value is less than the disconnection judgment threshold, then a slow drift fault is determined to exist. When any fault is detected, the logic control unit cuts off the energization of the safety relay coil, causing the normally closed contact to physically open, thus forcing the hoist safety chain to lose power and lock out.

[0050] In this embodiment, when it is determined that the compensation liquid level value is less than the first preset threshold L1 and greater than the second preset threshold L2, the logic control unit drives the audible and visual alarm device independent of the safety circuit. When the compensation liquid level value is determined to be less than the second preset threshold L2, the logic control unit cuts off the excitation of the safety relay coil, causing the normally closed contact to physically disconnect, and forcing the hoist safety chain to lose power and lock.

[0051] Example 2 Based on Embodiment 1, in this embodiment, the signal morphology diagnostic unit is further used for: Construct a theoretical mapping relationship between the direction of volume change and the direction of pressure change of the hydraulic actuator, and generate the theoretical direction of liquid level change and the theoretical direction of pressure change; In this embodiment, the theoretical mapping relationship can be expressed as follows: under the premise that the system does not perform oil replenishment operation and there is no external leakage, the actuator extends (volume increases) - the system pressure decreases, and the oil tank level decreases (oil is drawn into the actuator); the actuator retracts (volume decreases) - the system pressure increases, and the oil tank level increases (oil flows back into the oil tank).

[0052] Based on the above mapping, the module generates the theoretical liquid level change direction (+ indicates rising, - indicates falling) and the theoretical pressure change direction (+ indicates rising, - indicates falling) for each vehicle operation action label.

[0053] The pressure signal from the pressure sensor is acquired. In this embodiment, the pressure sensor can be an existing stainless steel diaphragm pressure transmitter, which is installed at the outlet of the accumulator in the main oil supply pipeline of the hydraulic station. Its analog signal is connected to the analog input module of the PLC for cross-consistency verification.

[0054] Within a preset sliding time window, calculate the fluctuation characteristic parameters of the digital liquid level signal and the pressure signal, such as calculating the standard deviation of the digital liquid level signal and the standard deviation of the pressure signal, and calculate the normalized activity index based on the full scale of the fluctuation characteristic parameters, that is, divide the two by their corresponding full scale values ​​to obtain the first index and the second index respectively. In this embodiment, the normalized activity index refers to the dimensionless ratio obtained by dividing the standard deviation (fluctuation amplitude) of the liquid level signal and the pressure signal by their full-scale value within the same sliding time window.

[0055] If, within the preset duration, the first index remains less than the first activity threshold and the second index remains greater than the second activity threshold, then it is determined that the continuous liquid level sensing unit has a signal freezing fault. This is because normal fluctuations in the pressure signal indicate that the hydraulic system is performing an action, and the liquid level should fluctuate synchronously. However, the constant liquid level signal violates the physical linkage law. The liquid level-pressure phase plane trajectory is constructed based on the rate of change of the digital liquid level signal and the pressure signal, that is, the rate of change of the liquid level signal is used as the X-axis and the rate of change of the pressure signal is used as the Y-axis to construct the liquid level-pressure phase plane trajectory.

[0056] Obtain the instantaneous motion direction angle of the liquid level-pressure phase plane trajectory, such as instantaneous motion direction angle = arctan(liquid level change rate / pressure change rate); The instantaneous motion direction angle is compared with the theoretical liquid level change direction. When the signs of the two directions are inconsistent, it means that the system is performing an action that causes the volume to increase (theoretically the liquid level should drop), but the liquid level signal shows an increase. This indicates that the continuous liquid level sensing unit has a slow drift fault. Based on the standard liquid level-pressure correspondence table (the standard liquid level values ​​corresponding to different working pressures are recorded during the system debugging phase to establish the standard liquid level-pressure correspondence table) and the pressure signal, the virtual liquid level value is calculated, and the residual between the virtual liquid level value and the digital liquid level value is obtained. The cumulative sum (CUSUM) algorithm is used to perform a cumulative sum test on the residual. When the cumulative residual exceeds the preset residual threshold, it is determined that the continuous liquid level sensing unit has a nondeterministic failure, that is, although the liquid level signal is not disconnected or completely frozen, it has been continuously deviating from the expected value calculated from the pressure signal, indicating that the sensor may output a distorted liquid level value due to scaling, temperature drift or local contamination. Based on the signal freeze fault determination results, slow drift fault determination results, and nondeterministic failure results, the nondeterministic failure confidence level is calculated. When the nondeterministic failure confidence level is greater than a preset global threshold, the logic control unit is controlled to execute a graded response. For example, the output of each sub-criteria is converted into independent uncertainty probabilities P1 (constant signal), P2 (slow drift), P3 (freeze-fluctuation activity), P4 (inconsistent direction), and P5 (CUSUM exceeding limit), with each probability value ranging from 0 to 1. The nondeterministic failure confidence level is calculated using a weighted summation method. When the nondeterministic failure confidence level > the preset global threshold, the level sensor is determined to have a nondeterministic failure: if 0.5 ≤ nondeterministic failure confidence level < 0.6, a sensor maintenance prompt state is triggered, and a yellow warning is displayed on the warning interface, suggesting maintenance but not locking out temporarily; if the nondeterministic failure confidence level ≥ 0.6, a sensor failure lockout state is triggered, and the PLC main control program immediately cuts off the excitation of the safety relay coil, forcing the hoist safety chain to lose power and lock out.

[0057] This confidence fusion mechanism effectively reduces unnecessary blocking caused by false alarms from a single diagnostic submodule, thereby improving system availability.

[0058] Example 3 Based on the above embodiments, this embodiment also provides a monitoring method for a hoist hydraulic station with a graded early warning function, the method comprising: A continuous level sensor unit installed in the hydraulic station oil tank of the hoisting machine acquires an analog signal by collecting the level displacement, performs analog-to-digital conversion on the analog signal to obtain a digital level signal, and obtains a digital level value based on the digital level signal. Monitor the status of the control signal output by the logic control unit to the hydraulic valve group, identify the current stage of the vehicle operation, and generate an action feature label corresponding to the current stage of the vehicle operation. Based on the action feature label, within the time window, the dynamic volume compensation coefficient matching the action feature label is called to perform dynamic volume change compensation on the digital liquid level value to obtain the compensated liquid level value. A first preset threshold L1 and a second preset threshold L2 are preset, and the first preset threshold L1 is greater than the second preset threshold L2; Based on the compensated liquid level value of multiple consecutive sampling periods, the current leakage rate is calculated using sliding window linear regression, and the remaining safe time for the liquid level to drop from the digital liquid level value to the second preset threshold L2 is predicted based on the current leakage rate. The analog signal is subjected to morphological feature extraction and time series pattern recognition to determine whether there is a constant signal fault and / or slow drift fault. When either fault is determined to exist, the logic control unit cuts off the excitation of the safety relay coil, causing the normally closed contact to physically open and forcing the hoist safety chain to lose power and lock out.

[0059] Although preferred embodiments of the invention have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including both the preferred embodiments and all changes and modifications falling within the scope of the invention.

[0060] Obviously, those skilled in the art can make various modifications and variations to this invention without departing from its spirit and scope. Therefore, if these modifications and variations fall within the scope of the claims of this invention and their equivalents, this invention also intends to include these modifications and variations.

Claims

1. A monitoring system for a hoist hydraulic station with graded early warning function, characterized in that, The system includes: A continuous liquid level sensing unit is installed in the hydraulic station oil tank of the hoist operator's car. It is used to collect the liquid level displacement to obtain an analog signal, perform analog-to-digital conversion on the analog signal to obtain a digital liquid level signal, and obtain a digital liquid level value based on the digital liquid level signal. The logic control unit has its analog input terminal connected to the continuous liquid level sensing unit, and internally includes a working condition sensing module, a dynamic threshold module, a leakage rate estimation and remaining safe time prediction module, and a signal morphology diagnosis module. The interlocking execution unit, controlled by the logic control unit, includes a safety relay with a forced-guide contact. The normally closed contact of the safety relay is connected in series to the safety chain circuit of the mine hoist electrical control system by physical hard wiring. The working condition perception module is used to identify the current operating action stage by monitoring the control signal status output by the logic control unit to the operating hydraulic valve group, and generate an action feature label corresponding to the current operating action stage. The dynamic threshold module is used to call a dynamic volume compensation coefficient that matches the action feature label within a time window based on the action feature label, and to perform dynamic volume change compensation on the digital liquid level value to obtain a compensated liquid level value. The logic control unit is configured with a first preset threshold L1 and a second preset threshold L2, and the first preset threshold L1 is greater than the second preset threshold L2. The leakage rate estimation and remaining safe time prediction module is used to calculate the current leakage rate based on the compensation liquid level value of multiple consecutive sampling periods using sliding window linear regression, and to predict the remaining safe time for the liquid level to drop from the digital liquid level value to the second preset threshold L2 based on the current leakage rate. The signal morphology diagnostic unit is used to extract morphological features and recognize time series patterns of the analog signal to determine whether there is a constant signal fault and / or a slow drift fault. When any fault is determined to exist, the logic control unit cuts off the excitation of the safety relay coil, causing the normally closed contact to physically open and forcing the hoist safety chain to lose power and lock out.

2. The hoist hydraulic station monitoring system with graded early warning function according to claim 1, characterized in that, The input port of the working condition sensing module is connected in parallel to the intermediate relay auxiliary contact of the digital output module of the logic control unit, and the working condition sensing module has a built-in truth table. The operating condition perception module identifies the current vehicle operation stage based on the control signal state and the truth table, and generates an action feature label corresponding to the current vehicle operation stage.

3. A hoist hydraulic station monitoring system with graded early warning function according to claim 1, characterized in that, The logic control unit has a built-in vehicle operation action feature signal library. The vehicle operation action feature signal library includes at least the following tags: brake release action tag, brake holding action tag, brake extend action tag, brake retract action tag, rocker lift action tag, rocker lower action tag, and static standby tag. Each action tag corresponds to a tag data. The tag data includes an enable logic matrix, a standard time domain waveform, a nonlinear compensation coefficient matrix, a follow-up offset, and an adaptive correction factor. The dynamic threshold module is specifically used for: The action feature tags are matched with the vehicle operation action feature signal library to obtain the target action tag and target tag data, and a synchronization trigger signal is generated and an internal timer is started. Based on the current timing value of the internal timer and the standard time-domain waveform of the target tag data, the per-unit invasion depth corresponding to the current timing value is obtained, and the predicted liquid level fluctuation amount is obtained based on the per-unit invasion depth and the preset maximum liquid level fluctuation amplitude. Based on the digital liquid level value and the action speed parameter of the hydraulic actuator, interpolation is performed on the nonlinear compensation coefficient matrix of the target label data to obtain the dynamic volume compensation coefficient. The predicted liquid level fluctuation, the dynamic volume compensation coefficient, and the adaptive correction factor of the target tag data are coupled and calculated to obtain the dynamic compensation amount. The digital liquid level value is corrected based on the dynamic compensation amount to obtain the compensated liquid level value of the current timing value. Based on the compensation liquid level value and the follow-up offset of the target tag data, the first preset threshold L1 and the second preset threshold L2 are dynamically offset to obtain the third dynamic threshold L3 and the fourth dynamic threshold L4 of the current timing value, respectively; the first preset threshold L1 and the second preset threshold L2 are updated to the third dynamic threshold L3 and the fourth dynamic threshold L4, respectively. When the target action tag disappears, the dynamic compensation amount is gradually attenuated according to the falling edge curve of the standard time domain waveform of the target tag data, and the dynamic compensation amount is smoothly transitioned to zero within a preset ebb tide time to maintain the continuity of the compensation liquid level value.

4. A hoist hydraulic station monitoring system with graded early warning function according to claim 3, characterized in that, The first calculation formula for obtaining the compensation liquid level value is: ； in, Indicates the compensation liquid level value. Indicates the digital liquid level value. This indicates the predicted amount of liquid level fluctuation. This represents the dynamic volume compensation coefficient. This represents the adaptive correction factor.

5. A hoist hydraulic station monitoring system with graded early warning function according to claim 3, characterized in that, The leakage rate estimation and remaining safe time prediction module is specifically used for: The compensation change rate is obtained based on the compensation liquid level value; If the target action tag is a static standby tag, or if the compensation change rate is within a preset range for a preset time, then the sampling point corresponding to the compensation liquid level value is determined to be a valid sampling point, and the valid sampling point is stored in the regression data pool. Based on the regression data pool, when the range of the compensation liquid level value within multiple consecutive sampling periods is less than the preset dead zone threshold, the first effective leakage sampling point is determined. Establish an effective leakage sampling window, and initialize the basic window width and basic step size of the effective leakage sampling window; Obtain the residual standard deviation of the compensated liquid level value within the effective leakage sampling window, and update the base window width based on the residual standard deviation; Each effective leakage sampling point within the effective leakage sampling window is assigned an exponential decay weight according to its time sequence, and a weighted least squares method is used for linear fitting to obtain the current leakage rate. The base step size is then updated based on the current leakage rate. The leakage acceleration is calculated based on the current leakage rate and the historical leakage rate obtained from the previous effective leakage sampling window. If the leakage acceleration is greater than 0, the leakage acceleration is smoothed to obtain the emergency expansion factor. Based on the geometric model of the oil tank, the cross-sectional area function corresponding to the digital liquid level value is obtained. Based on the compensation liquid level value of the current timing value, the second preset threshold L2, the current leakage rate and the emergency expansion factor, the remaining safe time integral model is established. The remaining safety time integral model is solved iteratively using a numerical integration method to calculate the remaining safety time corresponding to the liquid level dropping from the digital liquid level value to the second preset threshold L2; The predicted standard deviation of the remaining safety time is obtained based on the regression fitting residual, and the predicted value of the remaining safety time and the predicted confidence interval are obtained based on the predicted standard deviation.

6. A hoist hydraulic station monitoring system with graded early warning function according to claim 5, characterized in that, The second formula for obtaining the exponential decay weight is: ； The third calculation formula for the remaining safety time integral model is: ； in, Indicates the first The exponential decay weight of each effective leakage sampling point This indicates a pre-defined forgetting factor. Indicates the current moment. Indicates the first Sampling time of each effective leak sampling point Indicates the number of the valid leak sampling point. Indicates the remaining safe time. This indicates the compensation liquid level value for the current timing. This indicates the second preset threshold. Indicates digital liquid level value The corresponding cross-sectional area function, Indicates the current leakage rate. Indicates the emergency expansion factor. Indicates time, This indicates the digital liquid level value.

7. A hoist hydraulic station monitoring system with graded early warning function according to claim 5, characterized in that, The signal morphology diagnostic unit is specifically used for: The sample value variance of the analog signal is calculated within a preset diagnostic time window. If the sample value variance is less than a preset variance threshold and exceeds a preset constant time threshold within a first preset duration, a signal constant fault is determined to exist. Calculate the rate of change of the sliding average value of the analog signal within a preset drift detection window. If the absolute value of the rate of change of the sliding average value is greater than the preset drift rate threshold within a second preset duration, and the absolute value is less than the disconnection judgment threshold, then a slow drift fault is determined to exist. When any fault is detected, the logic control unit cuts off the energization of the safety relay coil, causing the normally closed contact to physically open, thus forcing the hoist safety chain to lose power and lock out.

8. A hoist hydraulic station monitoring system with graded early warning function according to claim 7, characterized in that, The signal morphology diagnostic unit is also used for: Construct a theoretical mapping relationship between the direction of volume change and the direction of pressure change of the hydraulic actuator, and generate the theoretical direction of liquid level change and the theoretical direction of pressure change; Acquire the pressure signal from the pressure sensor; Within a preset sliding time window, the fluctuation characteristic parameters of the digital liquid level signal and the pressure signal are calculated, and the normalized activity index is calculated based on the full scale of the fluctuation characteristic parameters to obtain the first index and the second index respectively. If, within a preset duration, the first index remains less than the first activity threshold and the second index remains greater than the second activity threshold, then it is determined that the continuous liquid level sensing unit has a signal freezing fault. Based on the rate of change of the digital liquid level signal and the pressure signal, a liquid level-pressure phase plane trajectory is constructed, the instantaneous motion direction angle of the liquid level-pressure phase plane trajectory is obtained, the instantaneous motion direction angle is compared with the theoretical liquid level change direction, and when the directions of the two are inconsistent, it is determined that the continuous liquid level sensing unit has a slow drift fault. The virtual liquid level value is calculated based on the standard liquid level-pressure correspondence table and the pressure signal. The residual between the virtual liquid level value and the digital liquid level value is obtained. The residual is subjected to a cumulative sum test. When the cumulative residual exceeds the preset residual threshold, it is determined that the continuous liquid level sensing unit has a nondeterministic failure. Based on the signal freezing fault determination results, the slow drift fault determination results, and the nondeterministic failure results, the nondeterministic failure confidence level is calculated. When the nondeterministic failure confidence level is greater than a preset global threshold, the logic control unit is controlled to execute a graded response.

9. A hoist hydraulic station monitoring system with graded early warning function according to claim 1, characterized in that, When it is determined that the compensation liquid level value is less than the first preset threshold L1 and greater than the second preset threshold L2, the logic control unit drives the audible and visual alarm device independent of the safety circuit. When the compensation liquid level value is determined to be less than the second preset threshold L2, the logic control unit cuts off the excitation of the safety relay coil, causing the normally closed contact to physically disconnect, and forcing the hoist safety chain to lose power and lock.

10. A monitoring method for a hoist hydraulic station with graded early warning function, characterized in that, The method includes: A continuous level sensor unit installed in the hydraulic station oil tank of the hoisting machine acquires an analog signal by collecting the level displacement, performs analog-to-digital conversion on the analog signal to obtain a digital level signal, and obtains a digital level value based on the digital level signal. Monitor the status of the control signal output by the logic control unit to the hydraulic valve group, identify the current stage of the vehicle operation, and generate an action feature label corresponding to the current stage of the vehicle operation. Based on the action feature label, within the time window, the dynamic volume compensation coefficient matching the action feature label is called to perform dynamic volume change compensation on the digital liquid level value to obtain the compensated liquid level value. A first preset threshold L1 and a second preset threshold L2 are preset, and the first preset threshold L1 is greater than the second preset threshold L2; Based on the compensated liquid level value of multiple consecutive sampling periods, the current leakage rate is calculated using sliding window linear regression, and the remaining safe time for the liquid level to drop from the digital liquid level value to the second preset threshold L2 is predicted based on the current leakage rate. The analog signal is subjected to morphological feature extraction and time series pattern recognition to determine whether there is a constant signal fault and / or slow drift fault. When either fault is determined to exist, the logic control unit cuts off the excitation of the safety relay coil, causing the normally closed contact to physically open and forcing the hoist safety chain to lose power and lock out.

Images