On-line monitoring system for trace moisture in oil of airport refueling vehicle

Abstract

The invention discloses an on-line monitoring system for trace moisture in oil of an airport refueling truck, and belongs to the technical field of airport refueling, the system does not take whether the total moisture concentration in aviation fuel exceeds a fixed threshold value as a safety control basis, but evaluates the risk of free water precipitation in real time on the basis of a thermodynamic phase change principle; the hierarchical interlocking control is triggered accordingly; the system comprises a multi-mode moisture sensing probe which is installed in an oil conveying pipeline of a refueling vehicle and used for synchronously obtaining a dielectric constant, an infrared absorption spectrum, a temperature signal and a pressure signal of fuel oil; and the oil product type self-adaptive module is in communication connection with the multi-mode moisture sensing probe and is used for automatically identifying the current oil product type according to the combined characteristics of the dielectric constant and the sound velocity and loading a corresponding saturated dissolved water concentration-temperature relation curve. According to the method, the industrial pain point that water detection is qualified but a fuel oil system is still frozen at high altitude is effectively solved, and the safety degree of aviation fuel oil ground guarantee is greatly improved.

Description

Technical Field

[0001] This invention belongs to the field of airport refueling technology, and in particular, it is an online monitoring system for trace moisture in fuel from airport refueling trucks. Background Technology

[0002] Airport refueling trucks are core specialized vehicles ensuring the refueling of civil aircraft on the ground. Their fuel delivery systems must ensure the cleanliness and safety of the aviation fuel being refueled, especially the control of moisture content. Although trace amounts of moisture can dissolve in fuel at room temperature, they easily precipitate and freeze in the low-temperature environment of high altitudes, clogging fuel filters or control system pipelines. In severe cases, this can lead to engine fuel supply interruption, posing a significant flight safety hazard. Therefore, real-time and reliable monitoring of the moisture status of fuel during refueling is a crucial link in the civil aviation ground support system.

[0003] Currently, the industry commonly uses offline laboratory testing (such as the Karl Fischer method) or online capacitive / optical moisture sensors for monitoring. However, these methods have significant drawbacks. For example, existing online monitoring equipment generally uses "whether the total moisture concentration exceeds a fixed threshold (such as 30 ppm)" as the sole safety criterion, completely ignoring the thermodynamic dynamic process of moisture phase change. This approach stems from oversimplifying complex physical phenomena into static index control. In reality, the water-dissolving capacity of aviation fuel is highly temperature-dependent—for example, the saturated dissolved water concentration of a Jet A-1 is approximately 60 ppm at 20°C, but drops sharply to about 15 ppm at -20°C. This means that even if the ground-based moisture detection shows only 25 ppm (below the 30 ppm limit), once the aircraft climbs to the low-temperature airspace, a large amount of dissolved water will precipitate as free water and freeze at even lower temperatures. More dangerously, rapid changes in ambient temperature during refueling (such as a refueling truck moving from direct sunlight into the shaded area of ​​a hangar) or drastic fluctuations in pipeline pressure (such as pump start-up / stopping or valve switching) will accelerate the phase change process. Existing equipment is completely incapable of detecting such dynamic risks. Failure to address this issue will result in a fatal disconnect between "passing ground inspections and icing in the air," rendering moisture monitoring ineffective and unable to truly intercept the risk of high-altitude icing. Summary of the Invention

[0004] The purpose of this invention is to provide an online monitoring system for trace moisture in fuel from airport refueling trucks, in order to solve the problems mentioned in the background art.

[0005] To achieve the above objectives, the present invention provides the following technical solution: an online monitoring system for trace moisture in fuel in airport refueling trucks. The system does not use whether the total moisture concentration in aviation fuel exceeds a fixed threshold as the basis for safety control. Instead, it is based on the principle of thermodynamic phase change to assess the risk of free water precipitation in real time and trigger graded interlocking control accordingly. The system includes: A multimodal moisture sensor probe is installed in the fuel delivery pipeline of a refueling truck to simultaneously acquire the dielectric constant, infrared absorption spectrum, temperature, and pressure signals of the fuel. An oil type adaptive module is communicatively connected to the multimodal moisture sensing probe, used to automatically identify the current oil type based on the combined characteristics of dielectric constant and sound velocity, and load the corresponding saturated dissolved water concentration-temperature relationship curve; The free water precipitation risk prediction engine calculates the dynamic risk score of free water precipitation under the current operating conditions based on the oil type, real-time water concentration, oil temperature T, oil temperature change rate dT / dt, and pressure. The integrated self-maintaining sensing unit includes a main measuring cavity, a sealed reference cavity, an ultrasonic transducer, and a differential signal conditioning circuit, which is used to perform in-situ contaminant removal and zero-point drift compensation without interrupting the refueling operation. An edge intelligent fusion processor is used to perform confidence-weighted fusion of multi-source sensor data and output moisture status information with uncertainty assessment. The intelligent interlocking execution interface communicates with the hydraulic / electrical control system of the refueling truck. When the risk score exceeds the preset level threshold, it automatically executes flow regulation or oil supply cut-off actions that match the risk level.

[0006] In a preferred embodiment, this solution also includes a vehicle coupling feedforward compensation module for the environment, the module comprising: An ambient temperature sensor and a solar radiation intensity sensor are installed on the outside of the refueling truck; The vehicle-mounted GNSS positioning unit is used to obtain the current geographical location and driving status; The fuel thermal inertia prediction sub-model obtains the ambient temperature change rate from the historical meteorological database based on the geographical location, and combines this with the vehicle's stationary / driving status, fuel tank surface area-to-volume ratio, and fuel thermal conductivity parameters to predict the change trend of the main fuel temperature over the next 5-10 minutes. ; The free water precipitation risk prediction engine further By substituting the saturated dissolved water concentration into the calculation, a feedforward enhanced risk score is generated, enabling early warning of potential icing risks caused by vehicle movement or sudden environmental changes.

[0007] In this preferred embodiment, the risk scoring model used by the free water precipitation risk prediction engine is: ; in, This is an estimated value for the water concentration after fusion. This represents the saturated dissolved water concentration of the oil at temperature T. This refers to the amplitude of pipeline pressure fluctuation. The weight coefficients are obtained by training with historical fault data. The model explicitly characterizes the physical mechanism by which a sudden drop in temperature or pressure disturbance induces the phase transformation of dissolved water into free water, overcoming the misjudgment defect of relying solely on static moisture thresholds.

[0008] In a preferred embodiment of this scheme, the self-calibration process of the integrated self-maintenance sensing unit includes: During refueling intervals or low flow rate phases, close the main measuring chamber inlet valve and open the reference chamber passage. The reference cavity contains a pre-stored dry standard oil sample, the moisture content of which is known and stable. The differential signal conditioning circuit synchronously acquires the capacitance values ​​of the main cavity and the reference cavity, and calculates the offset ΔC; ΔC is fed back to the edge intelligent fusion processor to correct subsequent main cavity measurement results; This mechanism enables the system to achieve a moisture measurement drift of less than ±2 ppm over 30 days of continuous operation, which is significantly better than existing equipment that requires monthly manual calibration.

[0009] In this preferred embodiment, the infrared absorption sensor in the multimodal moisture sensing probe operates in the 1.94 μm water molecule characteristic absorption band and is equipped with a dual-optical-path structure with a reference wavelength of 1.85 μm. The edge intelligent fusion processor uses Dempster-Shafer evidence theory to perform confidence fusion on the moisture estimation results of the capacitance signal and the infrared signal. When the deviation between the two exceeds 15%, it is determined that there is oil contamination or sensor failure, and the output confidence is reduced.

[0010] In this preferred embodiment, when the oil type adaptive module identifies sustainable aviation fuel (SAF), it automatically activates a nonlinear moisture-dielectric mapping table. This mapping table is obtained through laboratory calibration and compensates for the nonlinear dielectric response of SAF caused by oxygen-containing functional groups, reducing the moisture measurement error from ±10 ppm in traditional methods to within ±3 ppm.

[0011] In this preferred embodiment, the intelligent interlock execution interface is connected to the refueling truck PLC controller via a CAN bus and follows the ARINC 825 aviation ground service communication protocol. When the risk score is "high", the system not only cuts off the solenoid valve, but also sends a "fuel quality abnormality" fault code to the PLC to prevent the refueling arm from automatically resetting until manual confirmation.

[0012] In this preferred embodiment, the edge intelligent fusion processor is deployed with a lightweight convolutional neural network (CNN), which takes multimodal sensor time-series segments as input and outputs moisture concentration, confidence level, and anomaly labels. The network has an inference latency of <50 ms on the embedded GPU and supports 20 risk reassessments per second, meeting the real-time requirements of high-speed refueling (>3000 L / min) scenarios.

[0013] In a preferred embodiment of this solution, the system further includes a digital twin interface module, which is used to upload real-time risk scores, moisture concentration, and oil temperature change rate to the airport's digital fuel platform to construct a "digital profile of moisture safety" for a single refueling task, supporting post-event traceability and fleet-level fuel quality trend analysis.

[0014] In this preferred embodiment, the system automatically performs a "cold start self-check" before each refueling task begins: Start the ultrasonic transducer to clean the probe surface; Switch to the reference cavity to verify the zero point; Injecting simulated oil samples to verify the full-range response; Refueling mode can only be entered after all three self-checks pass. This mechanism ensures that the system is always in a reliable state and avoids missed detections due to sensor failure.

[0015] Compared with the prior art, the technical effects and advantages of the present invention are as follows: This airport refueling truck's online monitoring system for trace moisture in fuel abandons the traditional design that uses a fixed moisture concentration threshold (such as 30 ppm) as the basis for safety control. Instead, it assesses the risk of free water precipitation in real time based on the principle of thermodynamic phase change, triggering tiered interlocking control accordingly. This allows the system to identify the real threat to flight safety from a physical perspective—that is, free water that may precipitate and freeze in the low-temperature environment at high altitudes—rather than simply focusing on the total moisture content. This design represents a paradigm shift from "passive compliance detection" to "proactive risk prevention." Compared with existing technologies, it effectively solves the industry pain point of "fuel system icing occurring at high altitudes even when moisture detection is qualified," significantly improving the safety of ground-based aviation fuel supply.

[0016] By simultaneously acquiring the dielectric constant, infrared absorption spectrum, temperature, and pressure signals of fuel through a multimodal moisture sensing probe, and combining this with an edge intelligent fusion processor to perform confidence-weighted fusion of multi-source data, the moisture state information not only achieves high accuracy but also includes uncertainty assessment. When the results from sensors based on different principles deviate significantly (e.g., the deviation between the capacitance method and the infrared method exceeds 15%), the system can automatically identify potential contamination or malfunctions and reduce the output confidence level to avoid false alarms. This mechanism significantly improves the system's robustness under complex fuel types and harsh operating conditions, overcoming the shortcomings of existing single sensors that are susceptible to interference from additives, bubbles, or impurities, leading to false alarms or missed alarms.

[0017] The system automatically identifies the current fuel type based on the combined characteristics of dielectric constant and sound velocity using an adaptive fuel type module, and dynamically loads the corresponding saturated dissolved water concentration-temperature relationship curve. Specifically, it employs a nonlinear moisture-dielectric mapping table for Sustainable Aviation Fuel (SAF), enabling the system to accurately adapt to various fuels such as Jet A-1, JP-8, and SAF without manual configuration. This design compensates for the nonlinear dielectric response of SAF caused by oxygen-containing functional groups, reducing the moisture measurement error from ±10 ppm in traditional methods to within ±3 ppm. This meets the high-precision monitoring requirements in the context of the future widespread adoption of green aviation fuels, a capability generally lacking in existing equipment. Attached Figure Description

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

[0019] Figure 1 This is a flowchart of an online monitoring system for trace moisture in fuel from an airport refueling truck, according to the present invention. Detailed Implementation

[0020] In the following description, numerous specific details are set forth in order to provide a more thorough understanding of the invention. However, it will be apparent to those skilled in the art that the invention can be practiced without one or more of these details. In other instances, certain technical features well-known in the art have not been described in order to avoid obscuring the invention.

[0021] Unless otherwise defined, the directions mentioned herein, such as up, down, left, right, front, back, inside, and outside, are based on the directions shown in the figures of this invention, and are explained here together.

[0022] This embodiment provides, for example Figure 1 The present invention provides an online monitoring system for trace moisture in fuel in airport refueling trucks. Its core lies in abandoning the traditional approach of judging safety solely based on whether the total water concentration in aviation fuel exceeds a fixed threshold (e.g., 30 ppm). Instead, it uses the principle of thermodynamic phase change to assess the dynamic risk of free water precipitation in real time and trigger graded interlocking control accordingly, thereby effectively intercepting the risk of high-altitude icing at the ground stage.

[0023] In this embodiment, the system is deployed as a whole in the oil pipeline and vehicle control system of the airport-dedicated refueling vehicle. It mainly includes a multimodal moisture sensor probe, an oil type adaptive module, a free water precipitation risk prediction engine, an integrated self-maintaining sensor unit, an edge intelligent fusion processor, and an intelligent interlock execution interface. It further includes an environment-vehicle coupling feedforward compensation module and a digital twin interface module.

[0024] In this embodiment, a multimodal moisture sensor probe is installed in the fuel delivery pipeline of the refueling truck to simultaneously acquire multiple physical parameters of the aviation fuel during the refueling process. This probe integrates a capacitive dielectric sensor, a near-infrared absorption sensor, a temperature sensor, and a pressure sensor.

[0025] In this embodiment, a capacitive dielectric sensor is used to measure the dielectric constant of fuel; the near-infrared absorption sensor adopts a dual-optical-path structure, with a main detection wavelength of 1.94 μm (corresponding to the characteristic absorption peak of the O–H bond in water molecules) and a reference wavelength of 1.85 μm (without significant water absorption). The dual-optical-path differential method eliminates interference from light source fluctuations and fuel color; the temperature sensor uses a high-precision PT100 element, and the pressure sensor has a piezoresistive structure with a sampling frequency of not less than 10 Hz.

[0026] In this embodiment, all signals are transmitted to the edge intelligent fusion processor via shielded cables to achieve millisecond-level synchronous acquisition. The oil type adaptive module is communicatively connected to a multimodal moisture sensor probe to automatically identify the type of oil being injected based on the combined characteristics of the measured dielectric constant and sound velocity. The sound velocity is obtained through a secondary ultrasonic transducer.

[0027] In this embodiment, the system pre-stores a characteristic database of typical aviation fuels, including JetA-1, JP-8, and Sustainable Aviation Fuel (SAF). When the identification result is sustainable aviation fuel, the system automatically activates a nonlinear moisture-dielectric mapping table specifically constructed for SAF. This mapping table was obtained through laboratory calibration of SAF samples with different moisture contents (0–50 ppm), effectively compensating for the nonlinear dielectric response of SAF caused by oxygen-containing functional groups. Experimental results show that using this mapping table, the moisture measurement error can be reduced from ±10 ppm using traditional linear methods to within ±3 ppm.

[0028] At the same time, the system automatically loads the corresponding saturated dissolved water concentration-temperature relationship curve based on the identified oil type. For example, for JetA-1, the empirical formula is used: ; Where T is the fuel temperature (unit: °C). The unit is ppm.

[0029] In this embodiment, the free water precipitation risk prediction engine is the core logic unit of the system, and its inputs include oil type and real-time water concentration after fusion. Measured oil temperature T, oil temperature change rate and pipeline pressure fluctuation amplitude The output is a dynamic risk score R.

[0030] The engine uses the following risk scoring model: ; The meanings of the symbols in the formula are as follows: : Estimated fused moisture concentration (unit: ppm) output by the edge intelligent fusion processor; : The saturated dissolved water concentration (ppm) of the oil at the current oil temperature T; Oil temperature change rate (unit: °C / min), calculated by differential calculation using a sliding time window; The pipeline pressure fluctuation amplitude is defined as the standard deviation of the pressure signal over the most recent 10 seconds (unit: kPa). The weighting coefficients are obtained through training and optimization using historical icing events and false alarm datasets; a typical value is... .

[0031] In this embodiment, the model explicitly characterizes the physical mechanism that "even if the total moisture content does not exceed the standard, a sudden drop in temperature or severe pressure disturbance may still induce a phase transition of dissolved water to free water," fundamentally overcoming the defects of underreporting or misjudgment caused by relying solely on static moisture thresholds.

[0032] In this embodiment, to address the issue of sudden changes in ambient temperature caused by frequent movement of airport refueling trucks during operations (such as driving from the fuel depot to the apron), the system adds an environment-vehicle coupling feedforward compensation module. This module includes an ambient temperature sensor (accuracy ±0.5°C) and a solar radiation intensity sensor (range 0–1200W / m²) installed externally on the refueling truck, as well as an onboard GNSS positioning unit (supporting RTK, positioning accuracy better than 1 meter). The GNSS unit not only provides the geographical location but also outputs the vehicle's driving status (stationary / moving).

[0033] In this embodiment, the system queries the local meteorological API or the airport's historical meteorological database based on the current location to obtain the ambient temperature change rate. Furthermore, a fuel thermal inertia prediction sub-model is constructed to predict the change trend of the main fuel temperature over the next 5-10 minutes. The model considers the fuel tank's geometric parameters (surface area to volume ratio), fuel thermal properties (density ρ, specific heat capacity c, thermal conductivity), and the effects of solar radiation. Its simplified form is as follows: ; in: This is the current measured oil temperature; The ambient temperature; h is the overall heat transfer coefficient; A / V is the ratio of the surface area to the volume of the fuel tank; η is the solar energy absorption coefficient; This represents the intensity of solar radiation.

[0034] In this embodiment, the free water precipitation risk prediction engine further predicts temperature. Substitution The calculation generates a feedforward enhanced risk score, thereby enabling early warning of potential icing risks caused by sudden environmental changes such as vehicle movement or changes in sunlight shading. The measured warning time can be more than 2 minutes in advance.

[0035] In this embodiment, the integrated self-maintaining sensing unit includes a main measuring chamber, a sealed reference chamber, a miniature ultrasonic transducer, and a differential signal conditioning circuit. The main measuring chamber is connected to the fuel pipeline and is used for real-time contact with the fuel to be tested; the sealed reference chamber pre-seals a dry standard oil sample (moisture content <1 ppm, long-term stable).

[0036] In this embodiment, during refueling intervals or at low flow rates below 50 L / min, the system automatically closes the main measurement chamber inlet valve and opens the reference chamber passage. The differential signal conditioning circuit synchronously acquires the capacitance values ​​of the main chamber and the reference chamber, and calculates the zero-point offset ΔC. This offset is fed back to the edge intelligent fusion processor to correct subsequent main chamber measurement results, achieving in-situ zero-point drift compensation.

[0037] In this embodiment, a miniature ultrasonic transducer (operating frequency 40 kHz, power 2 W) is periodically activated to vibrate and clean the surface of the main measuring electrode for 5–10 seconds, removing contaminants such as colloids and microbial films. According to actual measurements, this self-calibration and self-cleaning mechanism enables the system to have a moisture measurement drift of less than ±2 ppm within 30 days of continuous operation, which is significantly better than existing equipment that requires monthly manual disassembly and calibration.

[0038] In this embodiment, the edge intelligent fusion processor is deployed on an embedded GPU platform (such as NVIDIA Jetson AGXOrin) and is responsible for performing confidence-weighted fusion of multi-source sensor data and outputting moisture status information with uncertainty assessment.

[0039] On one hand, the processor uses Dempster-Shafer evidence theory to fuse the confidence levels of the moisture estimation results from the capacitance method and the infrared method. When the relative deviation between the two exceeds 15%, the system determines that there may be oil contamination or sensor failure, and actively reduces the output confidence level to avoid false control.

[0040] On the other hand, a lightweight one-dimensional convolutional neural network (CNN) is deployed inside the processor. The input is a multi-sensor time sequence within a 1-second window (including dielectric constant, infrared absorbance, temperature, and pressure), and the output is moisture concentration, confidence score, and anomaly label (normal / abnormal). The network has an inference latency of less than 50 milliseconds on the embedded GPU and supports 20 risk reassessments per second, fully meeting the real-time requirements of high-speed refueling (flow rate > 3000 L / min) scenarios.

[0041] In this embodiment, the intelligent interlock execution interface communicates with the refueling truck PLC controller via a CAN bus and follows the ARINC 825 aviation ground support equipment communication protocol to ensure compatibility with mainstream airport ground support systems.

[0042] The system sets three risk thresholds: Low risk (R < 5.0): Warnings are displayed only on the human-machine interface; Medium risk (5.0 ≤ R < 8.0): Automatically limits refueling flow to 30% of the rated value; High risk (R ≥ 8.0): Immediately shut off the fuel delivery solenoid valve and send a "fuel quality abnormality" fault code to the PLC.

[0043] This fault code will trigger the PLC logic, preventing the refueling arm from automatically resetting or retracting until ground staff confirm and manually deactivate it on-site, effectively preventing substandard fuel from being injected into the aircraft's fuel tank.

[0044] In this embodiment, to ensure the system remains reliable at all times, the system automatically performs a "cold start self-check" process before each refueling task begins, including the following steps: (1) starting the ultrasonic transducer to clean the probe surface; (2) switching to the reference cavity to verify whether the zero point is within the allowable range; (3) injecting a simulated oil sample (containing 30 ppm standard water) through the built-in micro pump to verify whether the full-range response is accurate. Only after all three self-checks pass can the system be allowed to enter the formal refueling mode. This mechanism can effectively avoid false alarms caused by sensor contamination, drift, or failure, and significantly improve the robustness of the system.

[0045] In this embodiment, the system also includes a digital twin interface module, which uploads key data such as real-time risk scores, moisture concentration, oil temperature change rate, and control actions during each refueling process to the airport digital fuel management platform via 4G / 5G or airport private network. Based on this data, the platform constructs a "digital profile of moisture safety" for each refueling task, supports post-event quality traceability and abnormal event review, and can aggregate data from multiple vehicles and multiple days to perform fleet-level fuel quality trend analysis and predictive maintenance, providing data support for airport fuel management.

[0046] In this embodiment, a typical refueling task is taken as an example: (1) The refueling truck departs from the oil depot. The environment-vehicle coupling feedforward compensation module predicts that the oil temperature will drop by 7°C in the next 8 minutes. (2) After arriving at the apron, the system completes the cold start self-test. (3) Refueling begins, and the multimodal probe collects data in real time. (4) The edge processor fuses the data and calculates the risk score R=6.8 (medium risk). (5) The system automatically limits the flow to 30% and prompts "Pay attention to the risk of low temperature water separation". (6) If the shadow near the aircraft causes the oil temperature to drop sharply and R rises to 8.5, the system immediately cuts off the fuel supply and locks the refueling arm. (7) The data of the whole process is synchronously uploaded to the digital fuel platform to form an auditable record.

[0047] It should be noted that, in this document, relational terms such as "one" and "two" are used merely to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, the phrase "comprising an element defined as..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0048] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. An online monitoring system for trace moisture content in fuel from airport refueling trucks, characterized in that, The system assesses the risk of free water precipitation in real time based on thermodynamic phase change and triggers hierarchical interlocking control accordingly. The system includes: A multimodal moisture sensor probe is installed in the fuel delivery pipeline of a refueling truck to simultaneously acquire the dielectric constant, infrared absorption spectrum, temperature, and pressure signals of the fuel. An oil type adaptive module is communicatively connected to the multimodal moisture sensing probe, used to automatically identify the current oil type based on the combined characteristics of dielectric constant and sound velocity, and load the corresponding saturated dissolved water concentration-temperature relationship curve; The free water precipitation risk prediction engine calculates the dynamic risk score of free water precipitation under the current operating conditions based on the oil type, real-time water concentration, oil temperature T, oil temperature change rate dT / dt, and pressure. The integrated self-maintaining sensing unit includes a main measuring cavity, a sealed reference cavity, an ultrasonic transducer, and a differential signal conditioning circuit, which is used to perform in-situ contaminant removal and zero-point drift compensation without interrupting the refueling operation. An edge intelligent fusion processor is used to perform confidence-weighted fusion of multi-source sensor data and output moisture status information with uncertainty assessment. The intelligent interlocking execution interface communicates with the hydraulic / electrical control system of the refueling truck. When the risk score exceeds the preset level threshold, it automatically executes flow regulation or oil supply cut-off actions that match the risk level.

2. The online monitoring system for trace moisture in fuel from airport refueling trucks according to claim 1, characterized in that: It also includes a vehicle coupling feedforward compensation module for the environment, the module comprising: An ambient temperature sensor and a solar radiation intensity sensor are installed on the outside of the refueling truck; The vehicle-mounted GNSS positioning unit is used to obtain the current geographical location and driving status; The fuel thermal inertia prediction sub-model obtains the ambient temperature change rate from the historical meteorological database based on the geographical location, and combines this with the vehicle's stationary / driving status, fuel tank surface area-to-volume ratio, and fuel thermal conductivity parameters to predict the change trend of the main fuel temperature over the next 5-10 minutes. ; The free water precipitation risk prediction engine further By substituting the saturated dissolved water concentration into the calculation, a feedforward enhanced risk score is generated, enabling early warning of potential icing risks caused by vehicle movement or sudden environmental changes.

3. The online monitoring system for trace moisture in fuel from airport refueling trucks according to claim 2, characterized in that: The risk scoring model used by the free water precipitation risk prediction engine is as follows: ; in, This is an estimated value for the water concentration after fusion. This represents the saturated dissolved water concentration of the oil at temperature T. This refers to the amplitude of pipeline pressure fluctuation. The weight coefficients are obtained by training with historical fault data. The model explicitly characterizes the physical mechanism by which a sudden drop in temperature or pressure disturbance induces the phase transformation of dissolved water into free water, overcoming the misjudgment defect of relying solely on static moisture thresholds.

4. The online monitoring system for trace moisture in fuel from airport refueling trucks according to claim 3, characterized in that: The self-calibration process of the integrated self-maintenance sensing unit includes: During refueling intervals or low flow rate phases, close the main measuring chamber inlet valve and open the reference chamber passage. The reference cavity contains a pre-stored dry standard oil sample, the moisture content of which is known and stable. The differential signal conditioning circuit synchronously acquires the capacitance values ​​of the main cavity and the reference cavity, and calculates the offset ΔC; ΔC is fed back to the edge intelligent fusion processor to correct subsequent main cavity measurement results; This mechanism enables the system to achieve a moisture measurement drift of less than ±2 ppm over 30 days of continuous operation, which is significantly better than existing equipment that requires monthly manual calibration.

5. The online monitoring system for trace moisture in fuel from airport refueling trucks according to claim 4, characterized in that: The infrared absorption sensor in the multimodal moisture sensing probe operates in the 1.94 μm water molecule characteristic absorption band and is equipped with a dual optical path structure with a reference wavelength of 1.85 μm. The edge intelligent fusion processor uses Dempster-Shafer evidence theory to perform confidence fusion of the moisture estimation results of the capacitance signal and the infrared signal. When the deviation between the two exceeds 15%, it is determined that there is oil contamination or sensor failure, and the output confidence is reduced.

6. The online monitoring system for trace moisture in fuel from airport refueling trucks according to claim 5, characterized in that: When the oil type adaptive module identifies sustainable aviation fuel (SAF), it automatically activates a nonlinear moisture-dielectric mapping table. This mapping table, obtained through laboratory calibration, compensates for the nonlinear dielectric response of SAF caused by oxygen-containing functional groups, reducing the moisture measurement error from ±10 ppm in traditional methods to within ±3 ppm.

7. The online monitoring system for trace moisture in fuel from airport refueling trucks according to claim 6, characterized in that: The intelligent interlock execution interface is connected to the refueling truck PLC controller via a CAN bus and follows the ARINC 825 aviation ground service communication protocol. When the risk score is "high", the system not only cuts off the solenoid valve, but also sends a "fuel quality abnormality" fault code to the PLC to prevent the refueling arm from automatically resetting until manual confirmation.

8. The online monitoring system for trace moisture in fuel in airport refueling trucks according to claim 7, characterized in that: The edge intelligent fusion processor deploys a lightweight convolutional neural network (CNN), which takes multimodal sensor time-series segments as input and outputs moisture concentration, confidence level, and anomaly labels. The network has an inference latency of <50 ms on an embedded GPU and supports 20 risk reassessments per second, meeting the real-time requirements of high-speed refueling (>3000 L / min) scenarios.

9. The online monitoring system for trace moisture in fuel from airport refueling trucks according to claim 1, characterized in that: The system also includes a digital twin interface module, which is used to upload real-time risk scores, moisture concentration, and oil temperature change rate to the airport's digital fuel platform to build a "moisture safety digital profile" for a single refueling mission, supporting post-event traceability and fleet-level fuel quality trend analysis.

10. The online monitoring system for trace moisture in fuel from airport refueling trucks according to claim 1, characterized in that: The system automatically performs a "cold start self-check" before each refueling task begins: Start the ultrasonic transducer to clean the probe surface; Switch to the reference cavity to verify the zero point; Injecting simulated oil samples to verify the full-range response; Refueling mode can only be entered after all three self-checks pass. This mechanism ensures that the system is always in a reliable state and avoids missed detections due to sensor failure.

Images