Mining liquid level sensor based on multi-data fusion and monitoring method thereof

Abstract

The invention belongs to the technical field of liquid level monitoring under complex working conditions of a coal mine, and aims to solve the problems that an existing liquid level sensor cannot adapt to complex working conditions and is low in detection precision under the complex working conditions. According to the mining liquid level sensor based on multi-data fusion, a magnet is fixedly connected to one side of the upper end of a liquid level rod, the liquid level rod is of a bent rod structure arranged in an obtuse angle shape, the liquid level rod is rotationally arranged at the position of a fulcrum in a connecting cylinder, and a hollow floating ball is fixedly connected to the lower end of the liquid level rod; a plurality of reed switches are arranged in the junction box side by side along the motion trail of the magnet, the temperature sensor and the Hall element are integrated in the junction box, the reed switches serve as discrete switching value alarm devices for liquid level indication, the Hall element serves as a continuous switching value alarm device for liquid level indication, and the reed switches and the Hall element are both matched with the magnet; the temperature sensor is used for acquiring the temperature of the junction box in real time. The controller dynamically compensates the output of the Hall element based on the temperature of the junction box. A multi-sensor data fusion technology is adopted, and the monitoring precision is high.

Description

Technical Field

[0001] This invention belongs to the field of liquid level monitoring technology under complex working conditions in coal mines, specifically relating to a mine liquid level sensor based on multi-data fusion and its monitoring method. Background Technology

[0002] Trackless rubber-tired mining vehicles are commonly used for transporting materials such as ore and coal within mines. During this process, the liquid levels (e.g., oil, water, liquids) inside the vehicle need to be accurately monitored and controlled to ensure the vehicle's safety and normal operation. Traditional liquid level detection methods use devices such as floats and pressure measuring equipment. However, in the mining environment, due to factors such as high temperature, high pressure, corrosive gases, and particulate matter, sensors are easily damaged or malfunction, leading to inaccurate measurements and safety hazards. To solve these problems, a liquid level sensor for trackless rubber-tired mining vehicles has been developed.

[0003] These types of liquid level sensors typically employ advanced materials science and sensor technology, using high-temperature and corrosion-resistant materials and undergoing special treatments to ensure stable operation over long periods in harsh environments. Simultaneously, the sensors need to possess high precision, high sensitivity, and rapid response characteristics to accurately detect changes in liquid level in real time.

[0004] However, the existing liquid level sensors for trackless rubber-tired mining vehicles have the following problems: 1. The liquid level sensor's level rod and junction box are integrated, allowing heat from inside the housing to be transferred to the junction box via the level rod. This easily leads to circuit aging, burnout of internal wiring, or melting of solder. 2. The liquid level sensor uses a vertical float to determine the liquid level in the waste disposal tank. Oil easily accumulates in the waste disposal tank and adheres to the level rod and slider over time. Furthermore, the temperature in the waste disposal tank can reach 130℃ during use. At high temperatures, the oil adheres to the sensor surface in a liquid state. As the temperature gradually cools, the oil solidifies on the sensor surface, interfering with the slider's movement and failing to accurately indicate the liquid level. 3. The float in the liquid level sensor is small, and the buoyancy force is fixed. The built-in magnet in the float generates a magnetic signal, increasing the float's mass and resulting in lower accuracy in liquid level monitoring. 4. The liquid level sensor cannot adapt to complex operating conditions, exhibiting low detection accuracy under such conditions. Summary of the Invention

[0005] In order to solve at least one of the above-mentioned technical problems in the prior art, the present invention provides a mine liquid level sensor based on multi-data fusion and a monitoring method thereof.

[0006] This invention is achieved using the following technical solution: a mining liquid level sensor based on multi-data fusion, comprising a liquid level rod, a magnet, a hollow float, a connecting cylinder located inside the oil tank, and a junction box, reed switches, a temperature sensor, a Hall element, and a controller located outside the oil tank; the magnet is fixedly connected to one side of the upper end of the liquid level rod, and the liquid level rod is a bent rod structure with an obtuse angle. The liquid level rod is rotatably positioned at the fulcrum inside the connecting cylinder, and the hollow float is fixedly connected to the lower end of the liquid level rod; multiple reed switches are arranged in parallel along the movement trajectory of the bar magnet inside the junction box; the temperature sensor and the Hall element are integrated inside the junction box; the reed switches serve as discrete switch alarm devices for liquid level indication, and the Hall element serves as a liquid level indicator, with the output voltage proportional to the magnetic field strength, continuously reflecting the liquid level height; both are in conjunction with the magnet; the temperature sensor is used to acquire the junction box temperature in real time, and the controller dynamically compensates the output of the Hall element based on the junction box temperature; the reed switches, temperature sensor, and Hall element are all electrically connected to the controller.

[0007] Preferably, the junction box is connected to the flange on the oil tank wall by bolts, and a heat dissipation gap is left between the junction box and the flange. The junction box is connected to the connecting cylinder inside the oil tank by a snap ring.

[0008] Preferably, a shaped washer is provided at the hinge joint between the level rod and the fulcrum inside the connecting cylinder. The shaped washer is used to continuously crack and peel off the oil stains that begin to solidify around the fulcrum, preventing them from forming a firm adhesion layer.

[0009] Preferably, a miniature piezoelectric ceramic vibrator is provided at the hinge of the connection between the liquid level rod and the fulcrum inside the connecting cylinder. The miniature piezoelectric ceramic vibrator is periodically excited by the circuit to generate high-frequency micro-vibrations, which are then transmitted to the connecting hinge.

[0010] Preferably, an angle encoder is installed at the hinge joint connecting the level rod and the fulcrum inside the connecting cylinder. By monitoring the slight increase in rotational resistance, the encoder can predict the decrease in rotational performance at the hinge joint due to oil accumulation or mechanical wear. The angle encoder is electrically connected to the controller.

[0011] Preferably, the hollow float has a cylindrical cavity structure inside, and the surfaces of the hollow float and the level rod have a superoleophobic nano-coating; it also includes an excitation coil and an induction coil, which are miniaturized and integrated into the lower half of the cavity structure inside the hollow float, and the sensing surface is in direct contact with the oil to integrate a miniature three-coil electromagnetic abrasive sensor.

[0012] In a second aspect, the present invention provides a monitoring method for a mine liquid level sensor based on multi-data fusion, comprising the following steps: setting a dynamic alarm threshold related to the oil tank level, the dynamic alarm threshold being set based on a standard alarm threshold, equipment operating status, and oil temperature; acquiring the status of a reed switch and the output of a Hall element; comparing the liquid level value converted from the status of the reed switch and the output of the Hall element with the dynamic alarm threshold; and outputting liquid level alarm information, sensor fault information, or normal liquid level information based on the comparison result.

[0013] Preferably, when the equipment is running at the point where the vehicle has just started or is running unloaded and the oil temperature is below the preset high oil temperature threshold, the standard alarm threshold is selected for the dynamic alarm threshold; when the equipment is running at full load or experiencing severe vibrations, the high liquid level alarm threshold in the dynamic alarm threshold is increased and the low liquid level alarm threshold in the dynamic alarm threshold is decreased to avoid false judgments caused by liquid level fluctuations; at the same time, delayed judgment or average value judgment can be temporarily enabled to avoid triggering alarms by a single fluctuation peak; when the oil temperature is above the preset high oil temperature threshold, the low liquid level alarm threshold in the dynamic alarm threshold is increased to provide early warning. It also includes dynamic compensation for the output of the Hall element, the compensation method being: A temperature-voltage compensation model was established; tests were conducted in a high and low temperature test chamber, and the following data were obtained: Record the original output voltage V of the Hall element at different temperature points T; the different temperature points T are the internal temperatures of the junction box obtained by the temperature sensor. Through data analysis, a compensation function or lookup table is obtained; the compensation function of the temperature-voltage compensation model is: V_compensated=V [a+b (T-T_ref)] Where V_compensated is the output voltage of the Hall element after temperature compensation; a, b are the fitting coefficients; and T_ref is the reference temperature. The current raw output voltage V_current of the Hall element and the current internal temperature T_current of the junction box are obtained in real time. Based on the trained temperature-voltage compensation model, the output voltage V_compensated of the Hall element after temperature compensation is obtained; Based on the pre-calibrated voltage-level percentage curve, the real-time oil tank level value L_actual is obtained.

[0014] Preferably, the system also includes a hinge jamming early warning algorithm, which includes the following steps: establishing a baseline database for querying the expected rotation rate range of the level rod when different liquid levels change; obtaining the actual rotation rate of the level rod when the liquid level changes in real time, the actual rotation rate being obtained by dividing the current rotation angle of the level rod by the current rotation time; if the actual rotation rate of the level rod is less than the lower limit of the expected rotation rate range, the jamming confidence counter is incremented by 1; if it is greater than the expected rotation rate range, the jamming confidence counter is set to MAX(0, jamming confidence counter-1); if the jamming confidence counter is greater than a preset threshold, to avoid false alarms from a single fluctuation, an alarm is only triggered after multiple consecutive anomalies; the triggered warning is an increase in the rotation resistance of the level rod, indicating possible oil accumulation in the hinge, suggesting an inspection, and the warning level is a maintenance reminder, not a fault shutdown; In addition, a float failure diagnosis algorithm is included. This algorithm is used to determine whether the static angle of the level rod is normal at a specific liquid level. The algorithm includes: establishing an "angle-liquid level" standard model; obtaining the real-time angle of the level rod based on an angle encoder; comparing the real-time angle of the level rod with the expected angle of the level rod at the current liquid level obtained based on the "angle-liquid level" standard model; calculating the angle deviation between the two; if the angle deviation is greater than a preset tolerance, it is determined to be an angle abnormality; if the real-time angle of the level rod is less than the expected angle, an alarm is triggered: float failure; if the real-time angle of the level rod is greater than the expected angle, an alarm is triggered: mechanical structure abnormality.

[0015] Preferably, the output of liquid level alarm information, sensor fault information, or normal liquid level information based on the comparison result includes: In the case of Hall effect drift, when the reed switch is triggered at normal liquid level, the liquid level value L_actual converted by the Hall element is 15% (extremely low), and the angle deviation is within the preset tolerance, it is determined that the Hall sensor is faulty, and the angle encoder and reed switch are accepted. In the case of float damage, when the reed switch is triggered at high liquid level, the liquid level value L_actual converted by the Hall element is 65%, and the real-time angle is less than the expected angle, it is determined that the float is faulty. In the early stage of hinge jamming, when the reed switch is triggered at normal liquid level, the liquid level value L_actual converted by the Hall element changes slowly, and the angle encoder reading is consistently low, triggering a predictive maintenance warning. In the case of severe jamming, the reed switch is stuck in a certain state, the liquid level value L_actual converted by the Hall element remains unchanged or jumps, the actual rotation rate of the angle sensor reading is ≈0 or the real-time angle of the liquid level rod remains unchanged, it is determined that the liquid level rod mechanism is jammed. If the low-level reed switch is closed and the level value L_actual converted by the Hall element is less than 25%, a low-level alarm is triggered. If the low-level reed switch is closed and the level value L_actual converted by the Hall element is greater than 60%, the states of the reed switch and the Hall element conflict, which is determined to be a "sensor system fault," triggering the highest level fault alarm and prompting a check of the Hall element. At the same time, the reed switch signal is given priority. If the normal-level reed switch is triggered and the level value L_actual converted by the Hall element is 45%, and the angle deviation is within the preset tolerance range, the level is displayed normally, and there is no alarm.

[0016] Compared with the prior art, the beneficial effects of the present invention are: Employing multi-sensor data fusion technology, it combines the reliability of discrete alarms with the high precision of continuous monitoring. Through the redundant design of reed switches and linear Hall elements, the system achieves self-verification. Adaptive temperature compensation and dynamic alarm threshold algorithms make the sensor output more accurate and intelligent, and can adapt to complex and ever-changing working conditions.

[0017] A composite self-cleaning structure combining active and passive methods ensures long-term reliability and maintenance-free operation of mechanical components in extreme environments. A liquid level rod structure replaces the traditional float-based design, avoiding the limitations of traditional float movement and improving oil resistance. The buoyancy of the hollow float drives the liquid level rod to rotate around a fulcrum, changing the relative position of the magnet and reed switch to open or close the circuit. Furthermore, it incorporates a hinge jamming warning algorithm and a float failure algorithm, making it more suitable for complex operating conditions. Attached Figure Description

[0018] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0019] Figure 1 This is a schematic diagram of the overall structure of the liquid level sensor of the present invention; Figure 2 This is a flowchart of the method of the present invention. Detailed Implementation

[0020] The technical solutions of the embodiments of the present invention will be clearly and completely described 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 implementation methods obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0021] It should be noted that the structures, proportions, sizes, etc., shown in the accompanying drawings of this specification are only for the purpose of assisting those skilled in the art in understanding and reading the content disclosed in the specification, and are not intended to limit the conditions under which the present invention can be implemented. Therefore, they have no substantial technical significance. Any modifications to the structure, changes in the proportional relationships, or adjustments to the size, without affecting the effects and objectives that the present invention can produce, should fall within the scope of the technical content disclosed in the present invention. It should be noted that in this specification, relational terms such as "first" and "second" are only used to distinguish one entity from several other entities, and do not necessarily require or imply any actual relationship or order between these entities.

[0022] This invention provides an embodiment: like Figure 1 As shown, a mining liquid level sensor based on multi-data fusion includes a liquid level rod, a magnet, a hollow float, and a connecting cylinder located inside the oil tank, and a junction box, reed switches, a temperature sensor, a Hall element, and a controller located outside the oil tank. The magnet is fixedly connected to one side of the upper end of the liquid level rod, which is a bent rod structure with an obtuse angle. The liquid level rod is rotatably mounted at a fulcrum inside the connecting cylinder, and the hollow float is fixedly connected to the lower end of the liquid level rod. Multiple reed switches are arranged in parallel along the movement trajectory of the bar magnet inside the junction box. The temperature sensor and the Hall element are integrated inside the junction box. The reed switches serve as discrete switching alarm devices for liquid level indication, while the Hall element serves as an alarm device for continuously reflecting the liquid level height. Both work in conjunction with the magnet. The temperature sensor is used to acquire the temperature of the junction box in real time, and the controller dynamically compensates the output of the Hall element based on the junction box temperature. The reed switches, temperature sensor, and Hall element are all electrically connected to the controller.

[0023] In this embodiment, reed switches are used to provide reliable, discrete switching alarm points, with each reed switch representing a different liquid level (e.g., high liquid level alarm, normal liquid level, low liquid level alarm, and extremely low liquid level shutdown). The magnet at the top of the liquid level rod triggers these switches sequentially during rotation, thereby achieving segmented and precise monitoring of the liquid level.

[0024] The junction box is bolted to a flange attached to the tank wall, with a heat dissipation gap between the junction box and the flange. The junction box is connected to the connecting cylinder inside the tank via a snap ring. After the high temperature of the enclosure is dissipated through the flange and bolts, the remaining heat is conducted to the junction box or its interior. At this point, the circuit temperature is below 60℃, and the circuit components experience relatively low temperatures, ensuring good heat dissipation performance, improving the sensor's high-temperature resistance, preventing circuit damage and aging, and avoiding direct contact between the junction box and the high-temperature enclosure, further enhancing heat dissipation and high-temperature resistance. The original sensor circuit was located inside the level gauge, where the internal temperature of the enclosure reached over 130℃, and the circuit operated directly at 135℃, making it prone to damage. Furthermore, titanium alloy replaces traditional 304 stainless steel as the sensor material.

[0025] In addition, to address the issue that oil condensation may affect the flexibility of the hinge point between the fulcrum and the liquid level rod, a passive self-cleaning irregularly shaped washer, a miniature piezoelectric ceramic oscillator, and an angle encoder were designed.

[0026] Specifically, a shaped washer is installed at the hinge joint connecting the level rod and the fulcrum inside the connecting cylinder. This washer continuously breaks and peels away the oil contamination that begins to solidify around the fulcrum, preventing it from forming a firmly adhering layer. A miniature piezoelectric ceramic vibrator is also installed at the hinge joint connecting the level rod and the fulcrum inside the connecting cylinder. This vibrator is periodically excited by a circuit, generating high-frequency micro-vibrations that are transmitted to the connecting hinge. An angle encoder is also installed at the hinge joint connecting the level rod and the fulcrum inside the connecting cylinder. This encoder not only diagnoses "jamming" but also monitors torque changes or angular deviations during rotation in real time. By monitoring subtle increases in rotational resistance, the system can predictively determine if the hinge is experiencing reduced flexibility due to oil accumulation or mechanical wear, issuing a warning before complete jamming. If the float itself is damaged by corrosion or impact, its buoyancy will decrease, resulting in a discrepancy between the level rod's rotation angle and the expected level. The system can identify this anomaly through an algorithm, triggering an alarm for "float failure," rather than a simple "abnormal level."

[0027] The hollow float has a cylindrical cavity structure inside. The surfaces of the hollow float and the level rod have a super oleophobic nano-coating. The super oleophobic nano-coating uses biomimetic techniques (such as the lotus leaf effect) to fundamentally reduce the impact of oil contamination at a physical level. It also includes an excitation coil and an induction coil. The excitation coil and induction coil are miniaturized and integrated into the lower half of the cavity structure inside the hollow float. The sensing surface is in direct contact with the oil, integrating a miniature three-coil electromagnetic abrasive sensor.

[0028] Employing a "one-excitation, two-induction" coil distribution (30 turns for the excitation coil and 200 turns for the induction coil, wound in 5 layers), it can detect ferromagnetic abrasive particles ≥82μm and non-ferromagnetic abrasive particles ≥294μm in oil. The excitation coil, energized with high-frequency alternating current, establishes a stable alternating magnetic field around the float. The differential induction coil pair consists of two induction coils with identical parameters arranged in a specific geometrical symmetry. When no abrasive particles pass through, the output potential difference between the two is zero. When ferromagnetic or non-ferromagnetic abrasive particles pass through the magnetic field, they disturb the magnetic field distribution, causing the two induction coils to generate differential voltage signals.

[0029] By integrating liquid level signals and abrasive particle concentration data, a "liquid level-oil health" correlation algorithm is established. When the abrasive particle concentration exceeds the standard and the liquid level is abnormal, a combined alarm of "oil replacement + liquid level replenishment" is triggered first. This upgrades the system from single liquid level monitoring to dual monitoring of "liquid level + oil lubrication status," providing early warning of equipment wear failures and preventing mechanical damage caused by oil deterioration. The abrasive particle detection sensitivity is 13.75 times higher than traditional solutions, enabling online real-time monitoring of oil status and reducing the cost of manual sampling and testing. This differential design greatly suppresses common-mode interference (such as temperature drift and circuit noise) and is specifically optimized for electromagnetic interference in harsh mining environments, resulting in a significantly improved signal-to-noise ratio.

[0030] The steps of the "liquid level-oil health" correlation algorithm include: 1. The microprocessor synchronously acquires liquid level signals (from a linear Hall effect sensor) and abrasive particle signals (peak value, quantity, waveform).

[0031] 2. Feature Extraction: Abrasive Particle Concentration Trend (PCL): The moving average of abrasive pulse counts per unit time. Catastrophic Particle Event: Monitors the presence of abrasive particles of a specific size (e.g., >150 μm), a clear sign of severe wear. Liquid Level Change Rate (LCR): Calculates the rate of change of liquid level over time.

[0032] 3: Correlation Diagnostic Model (“Liquid Level-Oil Health” Correlation Algorithm): Scenario 1: Normal liquid level + slow increase in abrasive particle concentration; Diagnosis: Normal wear period or early abnormal wear.

[0033] Action: Record data and trigger a maintenance reminder to "recommend performing oil laboratory analysis", providing a basis for planned maintenance.

[0034] Scenario 2: Normal liquid level + sudden increase in large-sized abrasive particles; Diagnosis: Severe wear failure is occurring (such as bearing surface peeling, gear tooth breakage).

[0035] Action: Immediately trigger the highest level "Severe Wear Shutdown Inspection" alarm to avoid catastrophic failure.

[0036] Scenario 3: Rapid drop in liquid level + normal abrasive particle concentration; Diagnosis: High probability of external leakage.

[0037] Action: Triggers the "Liquid level drops rapidly, check for leaks" alarm.

[0038] Scenario 4: Slow liquid level drop + sharp increase in abrasive particle concentration; Diagnosis: Internal leakage or seal failure leads to contaminant entry, or insufficient lubricating oil results in poor lubrication and aggravated wear.

[0039] Action: Prioritize triggering the combined alarm of "oil replacement + fluid level replenishment + system check".

[0040] like Figure 2 As shown, in a second aspect of the present invention, a monitoring method for a mine liquid level sensor based on multi-data fusion includes the following steps: S1: Set a dynamic alarm threshold related to the oil tank level, which is based on the standard alarm threshold, equipment operating status, and oil temperature setting.

[0041] Specifically, when the equipment is in the state of vehicle startup or no-load operation and the oil temperature is lower than the preset high oil temperature threshold, the dynamic alarm threshold is selected from the standard alarm threshold, such as the low liquid level alarm line is set to 20%.

[0042] When the equipment is operating under full load or experiencing severe vibrations, the fluid level will fluctuate violently, and the L_actual value will fluctuate rapidly. In this situation, the system can temporarily activate "delayed judgment" or "average value judgment" to prevent a single fluctuation peak from triggering an alarm. Simultaneously, the threshold can be temporarily relaxed, for example, dynamically lowering the low fluid level alarm threshold to 15%, and then restoring it after the driving stabilizes. The oil temperature inside the tank can be estimated using the junction box temperature or other methods. Under high-temperature, heavy-load conditions, the oil may expand and its lubricating performance may decrease, making the system more sensitive to "low fluid level." Alarms can be triggered earlier; for example, when the oil temperature exceeds 100℃, the low fluid level alarm threshold can be increased from 20% to 25%, providing operators with earlier warning and response time.

[0043] S2: Obtain the state of the reed switch and the output of the Hall element.

[0044] This also includes dynamic compensation for the Hall element's output. The magnetic field strength of a magnet decreases as temperature increases, and the Hall element's sensitivity to magnetic fields also changes with temperature. The compensation method is as follows: A temperature-voltage compensation model was established; tests were conducted in a high and low temperature test chamber to obtain the following data: the original output voltage V of the Hall element was recorded at different temperature points T; the different temperature points T were the internal temperatures of the junction box obtained based on temperature sensors; through data analysis, a compensation function or lookup table was obtained; the compensation function of the temperature-voltage compensation model was: V_compensated = V [a+b [(T-T_ref)]; where V_compensated is the output voltage of the Hall element after temperature compensation; a, b are fitting coefficients, and T_ref is the reference temperature; The system acquires the current raw output voltage V_current of the Hall element and the current internal temperature T_current of the junction box in real time; based on the trained temperature-voltage compensation model, it acquires the output voltage V_compensated of the Hall element after temperature compensation; and based on the pre-calibrated voltage-level percentage curve, it acquires the real-time oil tank level L_actual.

[0045] After compensation, regardless of whether the junction box temperature is 20℃ or 60℃, when the liquid level is actually 50%, the liquid level value displayed by the system and used for judgment is an accurate 50%.

[0046] S3: Compare the state of the reed switch and the liquid level value converted by the Hall element output with the dynamic alarm threshold.

[0047] S4: Output liquid level alarm information, sensor fault information, or normal liquid level information based on the comparison results.

[0048] This invention uses a simple and reliable mechanism to verify complex and precise mechanisms. Mechanical structure (reed switch): Provides discrete, absolute, and reliable switching signals. Its advantages are simple principle, strong anti-interference ability, and resistance to temperature and circuit drift. Its disadvantage is that it can only provide a limited number of points and cannot perform continuous measurements. Algorithm / electronics (Hall element): Provides continuous, high-precision, and flexible analog signals. Its disadvantage is that it may drift due to temperature, electromagnetic interference, and device aging.

[0049] The system continuously compares the state of the reed switch with the liquid level value interpreted by the Hall element. There is a strict logical correspondence between them. This invention specifies the logic for choosing which value to accept when the two conflict: If the low-level reed switch is closed and the level value L_actual converted by the Hall element is less than 25%, a low-level alarm is triggered. If the low-level reed switch is closed and the level value L_actual converted by the Hall element is greater than 60%, the states of the reed switch and the Hall element conflict, which is determined to be a "sensor system fault," triggering the highest level fault alarm and prompting a check of the Hall element. At the same time, the reed switch signal is used first. If the normal-level reed switch is triggered and the level value L_actual converted by the Hall element is 45%, the level is displayed normally, and there is no alarm.

[0050] Meanwhile, this invention also designs a hinge jamming early warning algorithm, which includes the following steps: establishing a baseline database, which is used to query the expected rotation rate range of the liquid level rod when different liquid levels change; obtaining the actual rotation rate of the liquid level rod when the liquid level changes in real time, the actual rotation rate is obtained by dividing the current rotation angle of the liquid level rod by the current rotation time; if the actual rotation rate of the liquid level rod is less than the lower limit of the expected rotation rate range, the jamming confidence counter is incremented by 1; if it is greater than the expected rotation rate range, the jamming confidence counter is set to MAX(0, jamming confidence counter-1); if the jamming confidence counter is greater than a preset threshold, to avoid false alarms from a single fluctuation, an alarm is only triggered after multiple consecutive abnormalities; the triggered warning is that the rotation resistance of the liquid level rod has increased, and the hinge may have accumulated oil, suggesting an inspection. The warning level is a maintenance reminder, not a fault shutdown.

[0051] In addition, a float failure diagnosis algorithm was designed to determine whether the static angle of the level rod is normal at a specific liquid level. This algorithm includes: establishing an "angle-liquid level" standard model: during initial sensor installation or after maintenance, the model learns and records the normal angle-liquid level correspondence; due to the lever principle, this is a fixed and calculable physical relationship (obtained through a lookup table via a geometric model or on-site calibration); obtaining the real-time angle of the level rod based on an angle encoder; comparing the real-time angle of the level rod with the expected angle of the level rod at the current liquid level obtained based on the "angle-liquid level" standard model; calculating the angle deviation between the two; if the angle deviation is greater than a preset tolerance, it is determined to be an angle abnormality; if the real-time angle of the level rod is less than the expected angle, an alarm is triggered: float failure; if the real-time angle of the level rod is greater than the expected angle, an alarm is triggered: mechanical structure abnormality.

[0052] The final diagnostic network formed by reed switches, Hall elements, and angle encoders: In the case of Hall effect drift, when the reed switch is triggered at normal liquid level, the liquid level value L_actual converted by the Hall element is 15% (extremely low), and the angle deviation is within the preset tolerance, it is determined that the Hall sensor is faulty, and the angle encoder and reed switch are accepted. In the case of float damage, when the reed switch is triggered at high liquid level, the liquid level value L_actual converted by the Hall element is 65%, and the real-time angle is less than the expected angle, it is determined that the float is faulty. In the early stage of hinge jamming, when the reed switch is triggered at normal liquid level, the liquid level value L_actual converted by the Hall element changes slowly, and the angle encoder reading is continuously low, triggering a predictive maintenance warning. In the case of severe jamming, the reed switch is stuck in a certain state, the liquid level value L_actual converted by the Hall element remains unchanged or jumps, the actual rotation rate of the angle sensor reading is ≈0 or the real-time angle of the liquid level rod remains unchanged, it is determined that the liquid level rod mechanism is jammed.

[0053] The above description is merely a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A mine liquid level sensor based on multi-data fusion, characterized in that: It includes a level gauge, magnet, hollow float, and connecting cylinder located inside the oil tank, as well as a junction box, reed switch, temperature sensor, Hall element, and controller located outside the oil tank; A magnet is fixedly connected to the upper side of the liquid level rod, and the liquid level rod is a bent rod structure with an obtuse angle. The liquid level rod is rotatably set at the fulcrum inside the connecting cylinder, and a hollow float is fixedly connected to the lower end of the liquid level rod. Multiple reed switches are arranged in parallel along the movement trajectory of the magnet inside the junction box. The temperature sensor and Hall element are integrated inside the junction box. The reed switches serve as discrete switching alarm devices for liquid level indication, and the Hall element serves as a liquid level indicator. The output voltage is proportional to the magnetic field strength and can continuously reflect the liquid level height. Both work in conjunction with the magnet. The temperature sensor is used to obtain the junction box temperature in real time. The controller dynamically compensates the output of the Hall element based on the junction box temperature. The reed switches, temperature sensor, and Hall element are all electrically connected to the controller.

2. The mine liquid level sensor based on multi-data fusion according to claim 1, characterized in that: The junction box is connected to the flange on the oil tank wall by bolts, with a heat dissipation gap between the junction box and the flange. The junction box is connected to the connecting cylinder inside the oil tank by a snap ring.

3. A mine liquid level sensor based on multi-data fusion according to claim 1, characterized in that: A shaped washer is installed at the hinge joint between the level rod and the fulcrum inside the connecting cylinder. The shaped washer is used to continuously crack and peel off the oil stains that begin to solidify around the fulcrum, preventing them from forming a firm adhesion layer.

4. A mine liquid level sensor based on multi-data fusion according to claim 1, characterized in that: A miniature piezoelectric ceramic vibrator is installed at the hinge of the fulcrum connecting the liquid level rod and the connecting cylinder. The miniature piezoelectric ceramic vibrator is periodically excited by the circuit to generate high-frequency micro-vibration, which is transmitted to the connecting hinge.

5. A mine liquid level sensor based on multi-data fusion according to claim 1, characterized in that: An angle encoder is installed at the hinge joint connecting the level rod and the fulcrum inside the connecting cylinder. By monitoring the slight increase in rotational resistance, it can predict the decrease in rotational performance at the connecting hinge due to oil accumulation or mechanical wear. The angle encoder is electrically connected to the controller.

6. A mine liquid level sensor based on multi-data fusion according to claim 1, characterized in that: The hollow float has a cylindrical cavity structure inside, and the surfaces of the hollow float and the level rod have a super oleophobic nano-coating. It also includes an excitation coil and an induction coil, which are miniaturized and integrated into the lower half of the cavity structure inside the hollow float. The sensing surface is in direct contact with the oil, integrating a miniature three-coil electromagnetic abrasive sensor.

7. A monitoring method for a mine liquid level sensor based on multi-data fusion, wherein the mine liquid level sensor is based on multi-data fusion as described in any one of claims 1 to 6, characterized in that, Includes the following steps: Set a dynamic alarm threshold related to the oil tank level, the dynamic alarm threshold being based on a standard alarm threshold, equipment operating status, and oil temperature; Obtain the state of the reed switch and the output of the Hall element; The state of the reed switch and the liquid level value converted by the Hall element output are compared with the dynamic alarm threshold. Based on the comparison results, output liquid level alarm information, sensor fault information, or normal liquid level information.

8. A monitoring method for a mine liquid level sensor based on multi-data fusion according to claim 7, characterized in that: When the equipment is running at the point where the vehicle has just started or is running unloaded and the oil temperature is below the preset high oil temperature threshold, the standard alarm threshold is selected for the dynamic alarm threshold. When the equipment is running at full load or experiencing severe vibrations, the high liquid level alarm threshold in the dynamic alarm threshold is increased, and the low liquid level alarm threshold in the dynamic alarm threshold is decreased to avoid false alarms caused by liquid level fluctuations. At the same time, delayed judgment or average value judgment can be temporarily enabled to avoid triggering alarms by a single fluctuation peak. When the oil temperature is above the preset high oil temperature threshold, the low liquid level alarm threshold in the dynamic alarm threshold is increased to provide early warning. It also includes dynamic compensation for the output of the Hall element, the compensation method being: A temperature-voltage compensation model was established; tests were conducted in a high and low temperature test chamber, and the following data were obtained: Record the original output voltage V of the Hall element at different temperature points T; the different temperature points T are the internal temperatures of the junction box obtained based on the temperature sensor. Through data analysis, a compensation function or lookup table is obtained; the compensation function of the temperature-voltage compensation model is: V_compensated=V [a+b (T-T_ref)] Where V_compensated is the output voltage of the Hall element after temperature compensation; a, b are the fitting coefficients; and T_ref is the reference temperature. The current raw output voltage V_current of the Hall element and the current internal temperature T_current of the junction box are obtained in real time. Based on the trained temperature-voltage compensation model, the output voltage V_compensated of the Hall element after temperature compensation is obtained; Based on the pre-calibrated voltage-level percentage curve, the real-time oil tank level value L_actual is obtained.

9. A monitoring method for a mine liquid level sensor based on multi-data fusion according to claim 8, characterized in that: It also includes a hinge jamming early warning algorithm, which includes the following steps: establishing a baseline database, which is used to query the expected rotation rate range of the liquid level rod when different liquid levels change; obtaining the actual rotation rate of the liquid level rod when the liquid level changes in real time, the actual rotation rate is obtained by dividing the current rotation angle of the liquid level rod by the current rotation time; if the actual rotation rate of the liquid level rod is less than the lower limit of the expected rotation rate range, the jamming confidence counter is incremented by 1; if it is greater than the expected rotation rate range, the jamming confidence counter is set to MAX(0, jamming confidence counter-1); if the jamming confidence counter is greater than a preset threshold, to avoid false alarms from a single fluctuation, an alarm is only triggered after multiple consecutive abnormalities; the triggered warning is an increase in the rotation resistance of the liquid level rod, which may indicate oil accumulation in the hinge, and an inspection is recommended. The warning level is a maintenance reminder, not a fault shutdown. In addition, a float failure diagnosis algorithm is included. This algorithm is used to determine whether the static angle of the level rod is normal at a specific liquid level. The algorithm includes: establishing an "angle-liquid level" standard model; obtaining the real-time angle of the level rod based on an angle encoder; comparing the real-time angle of the level rod with the expected angle of the level rod at the current liquid level obtained based on the "angle-liquid level" standard model; calculating the angle deviation between the two; if the angle deviation is greater than a preset tolerance, it is determined to be an angle abnormality; if the real-time angle of the level rod is less than the expected angle, an alarm is triggered: float failure; if the real-time angle of the level rod is greater than the expected angle, an alarm is triggered: mechanical structure abnormality.

10. A monitoring method for a mine liquid level sensor based on multi-data fusion according to claim 9, characterized in that: Based on the comparison results, the output may include liquid level alarm information, sensor fault information, or normal liquid level information, including: In the case of Hall effect drift, when the reed switch is triggered at normal liquid level, the liquid level value L_actual converted by the Hall element is 15%, and the angle deviation is within the preset tolerance, it is determined that the Hall sensor is faulty, and the angle encoder and reed switch are accepted. In the case of float damage, when the reed switch is triggered at high liquid level, the liquid level value L_actual converted by the Hall element is 65%, and the real-time angle is less than the expected angle, it is determined that the float is faulty. In the early stage of hinge jamming, when the reed switch is triggered at normal liquid level, the liquid level value L_actual converted by the Hall element changes slowly, and the angle encoder reading is consistently low, triggering a predictive maintenance warning. In the case of severe jamming, the reed switch is stuck in a certain state, the liquid level value L_actual converted by the Hall element remains unchanged or jumps, the actual rotation rate of the angle sensor reading is ≈0 or the real-time angle of the liquid level rod remains unchanged, it is determined that the liquid level rod mechanism is jammed. If the low-level reed switch is closed and the level value L_actual converted by the Hall element is less than 25%, a low-level alarm is triggered. If the low-level reed switch is closed and the level value L_actual converted by the Hall element is greater than 60%, the states of the reed switch and the Hall element conflict, which is determined to be a "sensor system fault," triggering the highest level fault alarm and prompting a check of the Hall element. At the same time, the reed switch signal is used first. If the normal-level reed switch is triggered and the level value L_actual converted by the Hall element is 45%, and the angle deviation is within the preset tolerance range, the level is displayed normally, and there is no alarm.

Images