Cloud water-ice content detection dynamic calibration method and system of airborne terahertz radiometer

Abstract

The present application relates to the technical field of data calibration, and discloses a cloud water icing detection dynamic calibration method and system for an airborne terahertz radiometer, comprising: collecting external environment temperature and humidity data and calibration matching parameters; constructing an influence function of antenna icing on a flight platform on the radiation efficiency of a preset terahertz frequency band; performing icing analysis on the external environment temperature and humidity data to obtain an equivalent icing attenuation factor and correct a target source radiation reference value stored in the airborne terahertz radiometer; analyzing an environment compensation term and a system response term of the external environment temperature and humidity data; synthesizing the icing correction term, the environment compensation term and the system response term with the target source radiation reference value to generate a calibration coefficient; and using the calibration coefficient to compensate and correct real-time received original radiation signals of the terahertz frequency band to obtain a terahertz effective detection signal. The present application can improve the accuracy of cloud water icing detection dynamic calibration.

Description

Technical Field

[0001] This invention relates to the field of data calibration technology, and in particular to a dynamic calibration method and system for cloud water ice detection of an airborne terahertz radiometer. Background Technology

[0002] Airborne terahertz radiometers receive radiation signals from atmospheric targets such as clouds, water droplets, and ice crystals in the terahertz frequency band, enabling real-time detection of icing conditions along flight paths, which is of great significance for aviation safety and meteorological observation. However, in the airborne environment, the radiometer antenna and receiving link are affected by factors such as temperature changes, mechanical vibration, radome contamination, or icing, causing drift in system gain and noise characteristics. If only ground-based static calibration results are used, significant detection errors will be introduced.

[0003] Traditional calibration schemes often rely on ground calibration data or attempt to compensate by monitoring antenna VSWR and establishing complex physical models. However, antenna VSWR is difficult to measure accurately in real time during flight, and the model is affected by a variety of environmental factors, making it difficult to achieve reliable calibration in actual airborne dynamic environments. Summary of the Invention

[0004] This invention provides a dynamic calibration method and system for cloud water icing detection of an airborne terahertz radiometer, the main purpose of which is to solve the problem of low accuracy when performing dynamic calibration for cloud water icing detection.

[0005] To achieve the above objectives, the present invention provides a dynamic calibration method for cloud water icing detection using an airborne terahertz radiometer, comprising:

[0006] The reference load switching device of the airborne terahertz radiometer is controlled to align with the hot load and warm load in sequence according to a preset cycle, and the calibration matching parameters of the hot load and warm load, as well as the temperature and humidity data of the external environment outside the flight platform are collected.

[0007] Based on the simulation data and historical experimental data of the antenna feeder unit in the airborne terahertz radiometer, a function is constructed to show the influence of antenna icing on the radiation efficiency of the preset terahertz frequency band on the flight platform.

[0008] Using the aforementioned influence function, icing analysis is performed on the external environmental temperature and humidity data to obtain the equivalent icing attenuation factor of the antenna in the current terahertz detection path;

[0009] The target source radiation reference value stored in the airborne terahertz radiometer is corrected based on the equivalent icing attenuation factor to obtain the icing correction term. The environmental compensation term of the external ambient temperature and humidity data is analyzed, and the system response term of the airborne terahertz radiometer is analyzed based on the calibration matching parameters.

[0010] The icing correction term, the environmental compensation term, and the system response term are combined with the target source radiation reference value to generate the calibration coefficient of the original radiation signal on the flight platform.

[0011] In the microwave receiver subsystem of the radiometer, the calibration coefficient is used to compensate and correct the original radiation signal of the terahertz band received in real time, so as to obtain an effective terahertz detection signal after calibration for cloud water ice accumulation on the flight platform.

[0012] To address the aforementioned problems, the present invention also provides a dynamic calibration system for cloud ice detection using an airborne terahertz radiometer, the system comprising:

[0013] The data acquisition module is used to control the reference load switching device of the airborne terahertz radiometer, aligning it with the hot load and warm load sequentially according to a preset cycle, and collecting the calibration matching parameters of the hot load and warm load, as well as the temperature and humidity data of the external environment outside the flight platform.

[0014] The influence function construction module is used to construct the influence function of antenna icing on the preset terahertz frequency band radiation efficiency based on the simulation data and historical experimental data of the antenna feeder sub-unit in the airborne terahertz radiometer.

[0015] The equivalent icing attenuation factor analysis module is used to perform icing analysis on the external environmental temperature and humidity data using the influence function, and to obtain the equivalent icing attenuation factor of the antenna in the current terahertz detection path.

[0016] The reference value correction module is used to correct the target source radiation reference value stored in the airborne terahertz radiometer based on the equivalent icing attenuation factor to obtain the icing correction term, and to analyze the environmental compensation term of the external environment temperature and humidity data, and to analyze the system response term of the airborne terahertz radiometer according to the calibration matching parameters.

[0017] The calibration coefficient generation module is used to synthesize the icing correction term, the environmental compensation term, and the system response term with the target source radiation reference value to generate the calibration coefficient of the original radiation signal on the flight platform.

[0018] The terahertz effective detection signal analysis module is used in the microwave receiver subsystem of the radiometer to compensate and correct the original radiation signal of the terahertz band received in real time using the calibration coefficient, so as to obtain the terahertz effective detection signal after calibration for cloud water and ice accumulation on the flight platform.

[0019] This invention, through synchronous acquisition of ambient temperature and humidity, as well as calibration matching parameters for thermal and warm loads, constructs an influence function of antenna icing on terahertz band radiation efficiency by combining simulation and historical experimental data from antenna feeder sub-units. It accurately quantifies the equivalent icing attenuation factor and then specifically generates three types of compensation terms—icing, environment, and system response—to synthesize dynamic calibration coefficients with the target source radiation reference value. Finally, it compensates and corrects the original radiation signal, comprehensively solving the problem of traditional static calibration neglecting dynamic interferences such as icing, temperature and humidity drift, and system response during flight. This effectively improves the accuracy and stability of airborne terahertz radiometers in detecting cloud water icing, ensuring that the detected signal truly reflects the cloud water icing state on the flight platform, providing reliable data support for aviation safety early warning and accurate meteorological observation. Therefore, the dynamic calibration method and system for cloud water icing detection of airborne terahertz radiometers proposed in this invention can solve the problem of low accuracy in dynamic calibration of cloud water icing detection. Attached Figure Description

[0020] Figure 1 A flowchart illustrating a dynamic calibration method for cloud water icing detection using an airborne terahertz radiometer, provided in an embodiment of the present invention.

[0021] Figure 2 This is a functional block diagram of a dynamic calibration system for cloud water ice detection of an airborne terahertz radiometer provided in an embodiment of the present invention.

[0022] The realization of the objective, functional features and advantages of the present invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0023] It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.

[0024] This application provides a dynamic calibration method for cloud water ice detection using an airborne terahertz radiometer. The execution entity of this method includes, but is not limited to, at least one of the following electronic devices that can be configured to execute the method provided in this application: a server, a terminal, etc. In other words, the dynamic calibration method for cloud water ice detection using an airborne terahertz radiometer can be executed by software or hardware installed on a terminal device or a server device. The software can be a blockchain platform. The server includes, but is not limited to, a single server, a server cluster, a cloud server, or a cloud server cluster. The server can be an independent server or a cloud server providing basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communication, middleware services, domain name services, security services, content delivery networks (CDNs), and big data and artificial intelligence platforms.

[0025] Reference Figure 1 The diagram shown is a flowchart illustrating a dynamic calibration method for cloud water icing detection using an airborne terahertz radiometer according to an embodiment of the present invention. In this embodiment, the dynamic calibration method for cloud water icing detection using an airborne terahertz radiometer includes:

[0026] S1. Control the reference load switching device of the airborne terahertz radiometer to sequentially align with the hot load and warm load according to a preset cycle, and collect the calibration matching parameters of the hot load and warm load, as well as the temperature and humidity data of the external environment outside the flight platform.

[0027] In this embodiment of the invention, the external environmental temperature and humidity data are parameter data characterizing the atmospheric temperature and humidity state in the flight path; the calibration matching parameters include the voltage values ​​corresponding to the radiation signals of the thermal load and the warm load, as well as the known physical temperatures of the thermal load and the warm load.

[0028] In detail, the reference load switching device is located in the mechanical structure behind or to the side of the antenna and includes: a hot load, made of ccosorb or equivalent microwave absorbing material, with a built-in thin-film heating element and a high-precision temperature sensor, whose temperature is maintained at 340±0.5K by a PID temperature control circuit; a warm load, also made of microwave absorbing material, whose temperature is maintained at 300±1K by thermal coupling to the cabin environment or by independent temperature control; and a switching mechanism, which uses a dual-position electromagnetic push rod or a micro stepper motor to drive the load tray to move, so that the hot load or warm load can be accurately placed into the focal plane area of ​​the antenna.

[0029] In this embodiment of the invention, the step of sequentially aligning the thermal load and the warm load according to a preset cycle, and collecting the calibration matching parameters of the thermal load and the warm load, as well as the external environmental temperature and humidity data outside the flight platform, includes:

[0030] During the first period of the preset cycle, the antenna of the airborne terahertz radiometer is pointed at the sky and a preset number of first raw voltage signals of sky radiation are continuously collected. At the same time, the temperature and humidity data of the external environment outside the flight platform are collected based on the temperature and humidity sensor.

[0031] During the second time period of the preset cycle, the solenoid valve controlling the airborne terahertz radiometer drives the thermal load to move to the center of the antenna field of view and continuously collects a preset number of thermal load radiation signals.

[0032] During the third time period of the preset cycle, the antenna of the airborne terahertz radiometer is pointed at the sky again, and a preset number of second raw voltage signals of sky radiation are continuously collected.

[0033] During the fourth time period of the preset cycle, the solenoid valve controlling the airborne terahertz radiometer drives the warm load to move to the center of the antenna field of view and continuously collects a preset number of warm load radiation signals.

[0034] The first original voltage signal, the thermal load radiation signal, the second original voltage signal, and the warm load radiation signal are combined into calibration matching parameters.

[0035] In detail, the flight platform is an airborne carrier equipped with an airborne terahertz radiometer for cloud water icing detection. The airborne terahertz radiometer is a device used to receive terahertz frequency radiation signals to detect cloud water icing. Temperature and humidity sensors measure environmental parameters near the radiometer head. These sensors are specifically designed to sense atmospheric temperature and humidity, with a measurement range consistent with the temperature and humidity variations of the flight environment; for example, the temperature measurement range is -40℃ to 60℃, and the humidity measurement range is 0% to 100%RH. When monitoring external environmental temperature and humidity data, the power supply module of the airborne terahertz radiometer is first activated to provide a stable 5V DC voltage to the temperature and humidity sensors. The sensors sense the heat conduction and water vapor content of the surrounding atmosphere, converting physical quantities into electrical signals. These signals are then converted from analog to digital signals by an analog-to-digital converter, ultimately outputting external environmental temperature and humidity data characterizing the atmospheric state along the flight path. For example, if the temperature is -15℃ and the relative humidity is 85% at a certain moment, real-time atmospheric environmental parameters are collected, providing fundamental data for subsequent icing analysis and ensuring the accuracy of icing assessment.

[0036] Specifically, the thermal load is made of microwave absorbing material, and its physical temperature is maintained at approximately 340K by a temperature control device; the warm load is also made of microwave absorbing material, and its physical temperature is maintained at approximately 300K. The thermal and warm loads are alternately aligned with the antenna's field of view via an electromagnetic or motor-driven device. The preset cycle is a cyclical sequence containing multiple sub-segments, namely: the first sky observation period, the thermal load observation period, the second sky observation period, and the warm load observation period, starting a calibration cycle (e.g., lasting 4 seconds). Each cycle is divided into: First period A (0-1 seconds): The antenna is aligned with the sky, continuously acquiring 20 sets of the first raw voltage signal V_sky1 of sky radiation, and simultaneously recording the ambient temperature and humidity (Temp, RH). Second period B (1-2 seconds): The solenoid valve is controlled to drive the thermal load to the center of the antenna's field of view, continuously acquiring 20 sets of thermal load radiation signals V_hot. Third period C (2-3 seconds): The antenna is aligned with the sky again, continuously acquiring 20 sets of the second raw voltage signal V_sky2 of sky radiation. Fourth time period D (3-4 seconds): Control the solenoid valve to drive the warm load to the center of the antenna field of view, and continuously collect 20 sets of warm load radiation signals V_warm.

[0037] Furthermore, through periodic built-in reference load measurements (with a period as short as a few seconds), real-time calibration of the entire system during flight is achieved, avoiding steps that are difficult to perform in real time, such as measuring the antenna VSWR. All measurements are completed in a closed loop within the radiometer, resulting in strong anti-interference capabilities.

[0038] S2. Construct a function to determine the effect of antenna icing on the radiation efficiency of the preset terahertz frequency band on the flight platform based on the simulation data and historical experimental data of the antenna feeder unit in the airborne terahertz radiometer.

[0039] In this embodiment of the invention, the influence function is a mathematical function that characterizes the relationship between the antenna icing state and the terahertz band radiation efficiency.

[0040] In this embodiment of the invention, the step of constructing the influence function of antenna icing on the radiation efficiency of a preset terahertz frequency band on the flight platform based on simulation data and historical experimental data of the antenna feeder unit in the airborne terahertz radiometer includes:

[0041] Extract the simulated radiation pattern dataset of multiple discrete frequency points of the target band of the antenna feeder unit under the preset target state;

[0042] Extract the measured radiation efficiency attenuation dataset from the historical experimental data recorded by the antenna feeder unit under different controllable ice thickness conditions;

[0043] The normalized ratio of the main lobe gain in the measured radiation efficiency attenuation dataset and the simulated radiation pattern dataset is fitted to a bivariate nonlinear attenuation function with ice thickness and ambient temperature as independent variables.

[0044] The binary nonlinear attenuation function is physically constrained and verified based on the preset terahertz wave penetration depth.

[0045] Once the verification is successful, the binary nonlinear attenuation function is output as the function that reflects the effect of antenna icing on the radiation efficiency of the preset terahertz frequency band.

[0046] In detail, the antenna feeder unit is a functional unit in an airborne terahertz radiometer, consisting of an antenna and feeder responsible for signal transmission and reception. The simulation data is obtained by simulating and analyzing the antenna feeder unit using electromagnetic simulation software. HFSS software was selected for the electromagnetic simulation. The preset target conditions were set to a standard ambient temperature of 25℃, relative humidity of 50%, and no icing. The target frequency band was set to 0.3THz to 0.5THz, and five equally spaced discrete frequency points were selected within this band: 0.3THz, 0.35THz, 0.4THz, 0.45THz, and 0.5THz. A three-dimensional model of the antenna feeder unit was established using the simulation software. Simulation parameters such as mesh generation accuracy and solution frequency range were set. After running the simulation, the radiation pattern data corresponding to each discrete frequency point was obtained, including radiation intensity values ​​at different angles. These data were integrated to form a simulated radiation pattern dataset, which is a set of data representing the distribution of antenna radiation intensity in different directions. Historical experimental data refers to data recorded under different conditions for the antenna feeder unit in laboratory or real-world scenarios. Experiments were conducted in a low-temperature environment simulation laboratory. The antenna feeder unit was fixed on the experimental platform. The ambient temperature and ice thickness were controlled using cooling equipment and a spray device. Different controllable ice thicknesses were set: 0mm (no ice), 1mm, 2mm, 3mm, 4mm, and 5mm. At each ice thickness, the ambient temperature was maintained at -20℃. A fixed-power terahertz signal of 10mW was input to the antenna feeder unit via a signal source. The radiated power of the antenna was measured using a power meter, and the radiation efficiency (the ratio of radiated power to input power) was calculated and compared with the radiation efficiency without ice to obtain the radiation efficiency attenuation value. The attenuation value corresponding to each ice thickness was recorded, forming a measured radiation efficiency attenuation dataset. This dataset represents the data set showing the degree of radiation efficiency reduction under different ice thicknesses, as measured during the experiment. For example, the radiation efficiency attenuation was 10% with 1mm of ice, 25% with 3mm, and 40% with 5mm.

[0047] Specifically, the main lobe gain value of each discrete frequency point is extracted from the simulated radiation pattern dataset. For example, the main lobe gain is 15dB at 0.3THz and 16dB at 0.35THz. The main lobe gain of each frequency point is obtained sequentially. The main lobe gain is the maximum gain value of the main lobe in the antenna radiation pattern. The normalized ratio of the measured radiation efficiency attenuation value to the corresponding main lobe gain at each ice thickness is calculated. The normalized ratio is the ratio obtained after standardizing the measured radiation efficiency attenuation data and the simulated main lobe gain data. For example, at an ice thickness of 1 mm and a frequency of 0.3 THz, the normalized ratio of 10% attenuation to 15 dB main lobe gain is 0.0067. Using the normalized ratios at all frequencies and for all ice thicknesses as sample data, with ice thickness (mm) and ambient temperature (°C) as independent variables and the normalized ratio as the dependent variable, a bivariate nonlinear attenuation function is constructed using the least squares method. This function is a nonlinear mathematical function with ice thickness and ambient temperature as input variables and the degree of radiative efficiency attenuation as the output variable. The function form is as follows: ,in The normalized ratio, This refers to the thickness of the ice accumulation. The ambient temperature.

[0048] Furthermore, the terahertz wave penetration depth refers to the propagation distance of a terahertz wave in matter when its energy decays to 1 / e of its initial value. The penetration depth of terahertz waves in ice is a known physical parameter. For example, in the 0.3 THz band, the penetration depth of terahertz waves in ice is approximately 5 mm. This means that when the ice thickness exceeds 5 mm, the attenuation trend of radiation efficiency tends to level off and will not increase indefinitely. Substituting the constructed binary nonlinear attenuation function into ice thickness values ​​exceeding the penetration depth (e.g., 6 mm, 7 mm), the change in the normalized ratio of the function output is observed. If the growth rate of the normalized ratio decreases significantly after the ice thickness exceeds 5 mm, conforming to the physical characteristics of the penetration depth, the verification passes. If the function output continues to grow rapidly, the fitting coefficients are adjusted and refitted until the physical constraints are met. The physical constraint verification verifies the rationality of the constructed function based on the physical characteristics of terahertz waves, thus ensuring the physical rationality of the influence function, resolving the problem of discrepancies between pure mathematical fitting and actual physical phenomena, and improving the reliability of the influence function. Once the physical constraint verification is passed, the final determined binary nonlinear attenuation function is used as the function of the influence of antenna icing on the radiation efficiency of the preset terahertz frequency band and stored in the memory of the airborne terahertz radiometer for subsequent steps.

[0049] Furthermore, to quantify the impact of icing on radiation efficiency, it is necessary to establish a clear correlation between icing status and radiation efficiency. Therefore, an influence function is constructed by integrating simulation data and historical experimental data. This function serves as a bridge connecting real-time environmental data and radiation efficiency attenuation. Only through this function can the collected environmental data be transformed into an assessment of its impact on radiation efficiency, thereby enabling the analysis of the equivalent icing attenuation factor using the influence function.

[0050] S3. Using the influence function, perform icing analysis on the external environmental temperature and humidity data to obtain the equivalent icing attenuation factor of the antenna in the current terahertz detection path.

[0051] In this embodiment of the invention, the equivalent icing attenuation factor is a comprehensive parameter that characterizes the loss of radiation efficiency caused by antenna icing in the terahertz detection path, taking into account icing thickness, ambient temperature, and antenna matching status.

[0052] In this embodiment of the invention, the step of using the influence function to perform icing analysis on the external environmental temperature and humidity data to obtain the equivalent icing attenuation factor of the antenna in the current terahertz detection path includes:

[0053] The current ambient temperature and relative humidity data of the flight platform are analyzed, and the current ambient temperature and relative humidity data are input into a preset dew point temperature calculation model to obtain the potential amount of liquid water adhering to the antenna reflector under the current environment.

[0054] The amount of liquid water adhering is converted into the equivalent uniform ice thickness in the current terahertz detection path. The influence function is used to analyze the equivalent uniform ice thickness to obtain the theoretical radiation efficiency loss coefficient of each terahertz channel for antenna icing under the current operating conditions.

[0055] Electrical matching state correction is applied to the theoretical radiation efficiency loss coefficient to obtain the equivalent icing attenuation factor of the antenna in the current terahertz detection path.

[0056] In detail, the current ambient temperature and relative humidity data are extracted; for example, the current ambient temperature is analyzed to be -10℃ and the relative humidity is 90%. The dew point temperature calculation model is a mathematical model based on ambient temperature and relative humidity, employing the classic Magnus-Tetens equation, the expression of which is: ,in Dew point temperature, Relative humidity, For ambient temperature, and These are all empirical constants. For example, when T ≥ 0℃, a = 17.5, b = 237.3; when T < 0℃, , The dew point temperature calculation model inputs extracted temperature and humidity data into the model to calculate the dew point temperature, which refers to the temperature at which air reaches saturation when cooled under constant water vapor content and air pressure. For example, the dew point temperature is approximately -11.2℃. Since the current ambient temperature (-10℃) is higher than the dew point temperature (-11.2℃), liquid water will adhere to the antenna reflector. The calculation is based on the difference between the dew point temperature and the ambient temperature, relative humidity, and the area of ​​the antenna reflector. (Preset to 0.1m²), the amount of liquid water adhering is calculated using an empirical formula, which is: ,in This refers to the amount of liquid water adhering to the surface. The proportionality coefficient is 0.001 g / (℃·%·m²). The amount of liquid water adhering refers to the mass or volume of liquid water that may adhere to the antenna reflector surface. It can accurately determine whether the antenna will accumulate ice and the potential for ice accumulation.

[0057] Specifically, the equivalent uniform ice thickness converts the amount of liquid water adhering to the surface of the antenna into the thickness of ice uniformly distributed on the antenna surface. For example, if the density of ice is 0.9 g / cm³ and the area of ​​the antenna reflector is 0.1 m² (i.e., 100 cm²), the volume of ice corresponding to the liquid water is calculated according to the formula: volume = mass / density; the equivalent uniform ice thickness = volume / area. Substituting the equivalent uniform ice thickness and the current ambient temperature into a binary nonlinear attenuation function, the theoretical radiation efficiency loss coefficient is calculated. This theoretical radiation efficiency loss coefficient is obtained through the influence function, considering only the ice thickness and ambient temperature as radiation efficiency loss parameters.

[0058] Furthermore, the electrical matching state correction is an adjustment to the theoretical radiation efficiency loss coefficient to reflect the impact of the antenna matching state on radiation efficiency. The antenna's voltage standing wave ratio (VSWR) is also considered. ) and reflection coefficient ( The relationship is Thus, the reflection coefficient and the influence coefficient of impedance mismatch on radiation efficiency are obtained. By multiplying the theoretical radiation efficiency loss coefficient by this influence coefficient, the equivalent icing attenuation factor is obtained. This incorporates the influence of antenna matching state on radiation efficiency loss, solves the analysis error problem caused by neglecting impedance mismatch when only considering icing, and improves the accuracy of the icing attenuation factor.

[0059] Furthermore, the real-time equivalent icing attenuation factor was obtained through analysis. This factor quantifies the actual impact of icing on radiation efficiency. However, the target source radiation reference value of the airborne terahertz radiometer is calibrated under standard conditions without considering the effects of icing, ambient temperature and humidity, and system response. Directly using the reference value for calibration will lead to signal deviation. Therefore, it is necessary to correct the reference value based on this attenuation factor. At the same time, it is also necessary to quantify the effects of ambient temperature and humidity and system response to ensure the relevance and accuracy of the correction.

[0060] S4. Based on the equivalent icing attenuation factor, the target source radiation reference value stored in the airborne terahertz radiometer is corrected to obtain the icing correction term. The environmental compensation term of the external ambient temperature and humidity data is analyzed, and the system response term of the airborne terahertz radiometer is analyzed according to the calibration matching parameters.

[0061] In this embodiment of the invention, the target source radiation reference value is the radiation reference data obtained by calibrating an airborne terahertz radiometer under standard laboratory conditions using cold and hot targets. The icing correction term is a correction parameter used to compensate for the deviation in radiation signal caused by antenna icing.

[0062] In this embodiment of the invention, the step of correcting the target source radiation reference value stored in the airborne terahertz radiometer based on the equivalent icing attenuation factor to obtain an icing correction term includes:

[0063] Retrieve the cold target reference brightness temperature and hot target reference brightness temperature values ​​calibrated in a standard laboratory environment from the memory of the airborne terahertz radiometer;

[0064] On the calibration line segment between the cold target reference brightness temperature value and the hot target reference brightness temperature value, a brightness temperature offset proportional to the equivalent ice accumulation attenuation factor is determined;

[0065] The brightness temperature offset is converted into an equivalent noise temperature increment in the receive link of the airborne terahertz radiometer for antenna icing.

[0066] The normalization factor characterizing the contribution weight of icing noise is determined based on the ratio between the equivalent noise temperature increment and the preset theoretical noise temperature of the antenna, and the icing correction term is calculated based on the normalization factor and the voltage coefficient of the total circuit in the receiving link.

[0067] In detail, the standard laboratory environment was set at a temperature of 25°C, relative humidity of 50%, no icing, and good antenna matching. The cold target was a liquid nitrogen-cooled blackbody target with a preset reference brightness temperature of 77K, and the hot target was a temperature-controlled blackbody target with a preset reference brightness temperature of 300K. These two reference values ​​were pre-stored in the memory of the airborne terahertz radiometer. Data read commands retrieved these values ​​from memory, thus providing a standard reference for calculating the icing correction term and solving the problem of inaccurate correction due to the lack of a unified correction reference. The cold target reference brightness temperature is the radiation brightness temperature reference value corresponding to the cold target under standard laboratory conditions; the hot target reference brightness temperature is the radiation brightness temperature reference value corresponding to the hot target under standard laboratory conditions. The cold target reference brightness temperature (77K) and the hot target reference brightness temperature (300K) constitute a calibration line segment, ranging from 77K to 300K. The brightness temperature offset is proportional to the equivalent icing attenuation factor, and the proportionality coefficient is determined to be 0.1K / 100% based on experimental data (i.e., for every 1% increase in the attenuation factor, the brightness temperature offset increases by 0.1K). The brightness temperature offset is the value of the radiation brightness temperature deviating from the reference value due to icing. By converting the icing attenuation factor into the brightness temperature offset, a direct correlation between the icing effect and the radiation brightness temperature is established, solving the problem that the icing effect is difficult to directly use for reference value correction.

[0068] Specifically, the equivalent noise temperature increment is the additional noise temperature introduced by antenna icing. There is a linear relationship between the brightness temperature shift and the equivalent noise temperature increment, expressed by the following formula: ,in, This is the equivalent noise temperature increment. The conversion factor, determined experimentally to be 1.2, is the conversion factor. The brightness temperature offset is converted into a quantifiable noise temperature increment in the receiving link, allowing it to be directly used for circuit signal correction. The normalization factor is a standardized parameter characterizing the contribution weight of icing noise. The preset theoretical antenna noise temperature is 200K, and the normalization factor = equivalent noise temperature increment / preset theoretical antenna noise temperature. The voltage coefficient of the total circuit in the receiving link is preset to 0.01V / K (i.e., a voltage change of 0.01V for every 1K of noise temperature). The receiving link is the signal transmission path from the antenna receiving signal to the signal processing unit in the airborne terahertz radiometer. The voltage coefficient is the proportionality coefficient in the receiving link that converts noise temperature into back-end circuit voltage. The icing correction term is the normalization factor multiplied by the voltage coefficient, thus accurately correcting the icing signal and providing a key correction parameter for the subsequent generation of calibration coefficients.

[0069] In this embodiment of the invention, the environmental compensation term is a correction parameter used to compensate for the deviation of the radiation signal caused by changes in the external environment temperature and humidity; a pre-stored temperature and humidity drift error comparison table under different temperature and humidity noise figure conditions is obtained; the environmental temperature and humidity data are matched and queried with the temperature and humidity drift error comparison table to obtain the environmental compensation term caused by the performance drift of the airborne terahertz radiometer receiving link itself.

[0070] In detail, the temperature and humidity drift error reference table records the error data caused by the performance drift of the receiver link under different temperature and humidity conditions. This table was obtained in a laboratory setting by measuring the performance drift error of the receiver link under varying ambient temperature and humidity. The table records the error values ​​under different combinations of temperature (e.g., -40℃ to 60℃, in 5℃ intervals) and humidity (e.g., 0% to 100%RH, in 10%RH intervals). The error values ​​are expressed as voltage deviations; for example, at -10℃ and 85% humidity, the error value is 20μV. This table is pre-stored in the storage module of the airborne terahertz radiometer and is retrieved via a data retrieval command. Extract the current ambient temperature and humidity data (temperature -10℃, relative humidity 85%), match this data with the temperature and humidity drift error comparison table, and find the error value of 20μV for the corresponding temperature and humidity combination in the table. This error value is the environmental compensation item. The performance drift of the receiving link itself refers to the change in the parameters of electronic components in the receiving link caused by the change in temperature and humidity, which in turn causes the signal processing performance to change. Therefore, the problem of inaccurate calibration caused by temperature and humidity drift is taken into account, and environmental correction parameters are provided for the calibration coefficient.

[0071] In this embodiment of the invention, the system response term is a set of core parameters that describe the quantized correspondence between the input physical quantity and the output electrical signal of the airborne terahertz radiometer, including system gain and system offset.

[0072] In this embodiment of the invention, the step of analyzing the system response term of the airborne terahertz radiometer based on the calibration matching parameters includes:

[0073] Obtain the first voltage value of the thermal load radiation signal corresponding to the thermal load and the second voltage value of the warm load radiation signal corresponding to the warm load from the calibration matching parameters;

[0074] Obtain the first physical temperature for the thermal load and the second physical temperature for the warm load;

[0075] The ratio of the difference between the first voltage value and the second voltage value to the difference between the first physical temperature and the second physical temperature is used as the system gain in the system response term;

[0076] Calculate the product between the system gain and the second physical temperature, and use the difference between the first voltage value and the product result as the system offset in the system response term.

[0077] Specifically, the first voltage value (V_hot) is the average measured voltage corresponding to the thermal load radiation signal. The second voltage value (V_warm) is the average measured voltage corresponding to the warm load radiation signal. The first physical temperature (T_hot) is the known physical temperature of the thermal load (e.g., 340K). The second physical temperature (T_warm) is the known physical temperature of the warm load (e.g., 300K). These data are derived from periodic measurements of the internal reference load of the radiometer, ensuring data synchronization and representativeness of the current system state.

[0078] Specifically, calculate the system gain (G): The system gain is the ratio of the voltage difference between the hot load and the warm load to their temperature difference. This calculation is based on the linear response assumption, determining the slope of the voltage-temperature line using two known temperature points. The steeper the slope, the more sensitive the system is to temperature changes. The system offset is calculated as follows: Using the obtained system gain and warm load data, the system offset is calculated. This value represents the theoretical output voltage of the system when the input temperature is 0 K. In practice, it corrects the reference voltage offset of the system, ensuring the accuracy of the zero point of brightness temperature inversion. The system response term is the system gain and system offset calculated in real time using the above two-point calibration method. They serve as the basic parameters for radiometer signal processing, converting the raw voltage into brightness temperature. In the dynamic calibration method, this system response term (as a reference) is further synthesized with an environmental compensation term (compensating for temperature and humidity drift) and an icing correction term (compensating for the effects of radome icing) to generate more accurate final calibration coefficients, thereby achieving high-precision detection of cloud water icing in complex flight environments.

[0079] Furthermore, the icing correction term, environmental compensation term, and system response term correspond to the main dynamic interference factors affecting the radiation signal during flight. The target source radiation reference value is a fixed value under standard conditions. Relying on a single reference value alone cannot compensate for the signal deviation caused by dynamic interference. Only by combining these compensation terms with the reference value can a calibration coefficient applicable to real-time operating conditions be obtained, thereby achieving accurate correction of the original radiation signal and ensuring the comprehensiveness and specificity of the calibration coefficient.

[0080] S5. The icing correction term, the environmental compensation term, and the system response term are combined with the target source radiation reference value to generate the calibration coefficient of the original radiation signal on the flight platform.

[0081] In this embodiment of the invention, the calibration coefficient is a comprehensive coefficient used to compensate and correct the original radiation signal so that it can reflect the true atmospheric radiation state.

[0082] In this embodiment of the invention, the step of synthesizing the icing correction term, the environmental compensation term, and the system response term with the target source radiation reference value to generate calibration coefficients for the original radiation signal on the flight platform includes:

[0083] The system response term is algebraically added to the target source radiation reference value to obtain the preliminary gain parameter and the preliminary offset parameter;

[0084] Identify the gain compensation component and offset compensation component in the environmental compensation item according to the preset temperature and humidity-gain offset correction table, and superimpose the gain compensation component and the offset compensation component onto the initial gain parameter and the initial offset parameter respectively to obtain the intermediate gain quantity and the intermediate offset quantity;

[0085] The icing correction term is assigned a dynamic weight that is adaptive to the equivalent icing attenuation factor, wherein the dynamic weight is used to adjust the degree of influence of the icing correction term on the intermediate gain and the intermediate offset.

[0086] Multiply the first correction amount related to gain in the ice accumulation correction term by the dynamic weight to obtain the weighted gain correction amount, and combine the weighted gain correction amount with the intermediate gain amount to generate the system gain coefficient;

[0087] The second correction amount related to the offset in the ice accumulation correction term is multiplied by the dynamic weight to obtain the weighted offset correction amount, and the weighted offset correction amount is combined with the intermediate offset amount to generate the system offset amount;

[0088] The system gain coefficient and the system offset together constitute a calibration coefficient for real-time correction of the original radiation signal.

[0089] In detail, the target source radiation reference values ​​include the cold target reference brightness temperature (77K) and the hot target reference brightness temperature (300K), which are the "factory" or "historical" calibration coefficients of the system under ideal ground conditions or the previous stable cycle. It serves as a reference starting point. A pre-calibrated "temperature and humidity-gain / offset correction table" is used in the laboratory. This table quantifies the systematic drift patterns of the radiometer circuitry and components under different temperature and humidity environments. Based on real-time temperature and humidity data, the pre-defined correction table is consulted to identify the gain compensation component and offset compensation component in the environmental compensation item. These are then superimposed onto the preliminary parameters from the previous step to obtain intermediate gain and offset values. In other words, after establishing the current system response reference, a general compensation for environmental temperature and humidity is immediately introduced. This step corrects the preliminary parameters to a level matching the current cabin environment, yielding intermediate values. At this point, the calibration coefficients already include corrections for the system's instantaneous response and known environmental drift.

[0090] Specifically, the dynamic weights are adaptively adjusted based on the magnitude of the equivalent icing attenuation factor. These weight coefficients, used to characterize the importance of the icing correction term, are not fixed but dynamically change according to the severity of the icing (attenuation factor). For example, when icing is severe, the weights automatically increase, making the correction effect stronger; when there is no icing, the weights decrease or even become zero to avoid unnecessary intervention.

[0091] In this embodiment of the invention, the dynamic weighting of the icing correction term, which is adaptive to the equivalent icing attenuation factor, includes:

[0092] A dynamic weight decision-maker for multi-channel changes is constructed based on a pre-defined neural network, wherein the dynamic weight decision-maker learns a non-linear mapping relationship from the equivalent ice accumulation attenuation factor to the weight coefficient.

[0093] The equivalent icing attenuation factor of the current flight cycle on the flight platform is input into the dynamic weight decision-maker to obtain the instantaneous value of the equivalent icing attenuation factor and the rate of change of the equivalent icing attenuation factor per unit time in the adjacent calibration cycles of the current flight cycle.

[0094] The instantaneous value and the rate of change per unit time are combined into a two-dimensional feature vector, and the two-dimensional feature vector is used as the input layer data of the dynamic weight decision-maker.

[0095] The pre-stored multilayer perceptron model in the dynamic weight decision-maker is invoked to transform the two-dimensional feature vector sequentially through at least one hidden layer with a nonlinear activation function;

[0096] The activation function of the output layer in the dynamic weight decision-maker is used to linearly map the transformed neuron activity value to obtain the initial weight value.

[0097] The initial weight values ​​are constrained by physical characteristics using a preset limiting function to obtain dynamic weights.

[0098] In detail, the dynamic weight decision-maker is a model built on a neural network to calculate dynamic weights based on the icing attenuation factor. This dynamic weight decision-maker learns a nonlinear mapping relationship from the icing attenuation factor to the weight coefficients, where the nonlinear mapping relationship refers to the non-linear correspondence between the icing attenuation factor and the dynamic weights. Multichannel variation refers to the signal changes of multiple terahertz detection channels in an airborne terahertz radiometer. A multilayer perceptron model is selected during construction. Based on the original multilayer perceptron (MLP), the model is able to capture the multichannel variation patterns through "input layer expansion + channel fusion layer addition + output layer adaptation". The input layer design expands the single-channel "two-dimensional feature vector" into a "multi-channel feature matrix". The number of neurons in the input layer equals the number of channels × 2 (each channel has two features: instantaneous value + rate of change). For 5 channels, the number of neurons in the input layer is 5 × 2 = 10. Between the input layer and the first hidden layer, a "channel attention fusion layer" is added. Its core function is to automatically learn the importance weights of each channel, strengthening the features of key channels (such as high-frequency channels, which are more significantly affected by ice buildup) and weakening the influence of interfering channels (such as channels with abnormal data). Global average pooling is performed on the input feature matrix to obtain the global feature value for each channel. Two fully connected layers (number of hidden neurons = number of channels / 2 = 3, activation function is sigmoid) are then used to perform a non-linear transformation on the global feature value. The attention weights for each channel are output. The input feature matrix is ​​multiplied element-wise with the attention weights to obtain the feature matrix with fused attention, thus achieving differentiated fusion of channel features. There are two hidden layers, with 15 neurons in each layer (adapting to the multi-channel feature dimension) and ReLU activation function. The output layer is designed according to calibration requirements. If "unified dynamic weights" (global weights adapted to all channels to simplify calibration calculations) are required, the output layer has 1 neuron and a linear activation function. If "channel-specific dynamic weights" (precisely adapting to the differences of each channel) are required, the output layer has 5 neurons (number of channels) and a linear activation function.The model is trained by inputting the following features for each sample: instantaneous values ​​of attenuation factors and rates of change (standardized) for each of the five channels; label values ​​(optimal weights): through experimental calibration, the dynamic weights are manually adjusted under each operating condition to minimize the calibrated signal error of all channels (the error is calculated by comparing with the radiation signal of a standard blackbody target), and this weight is the label value of that sample; the Adam optimizer is used, with an initial learning rate of 0.001, which decreases by 50% every 100 epochs; the loss function is the mean squared error (MSE), which aims to minimize the difference between the model output weights and the label values; an L2 regularization term (penalty coefficient λ=0.0001) is added to prevent the model from overfitting to multi-channel noisy data; after the validation set loss does not decrease for 10 consecutive epochs, or after 1000 training epochs, the trained model parameters (weight matrix, bias term) are saved to the processing unit of the airborne terahertz radiometer.

[0099] Specifically, the current flight cycle refers to the time period during which signal calibration is currently performed, such as 1 second. The adjacent calibration cycle refers to the previous or next calibration cycle of the current flight cycle. For example, if the instantaneous value of the equivalent icing attenuation factor in the current flight cycle is 5.06%, and the equivalent icing attenuation factor in the adjacent calibration cycle (the previous 1-second cycle) is 4.86% based on historical data, and the time interval between adjacent calibration cycles is 1 second (the calibration cycle is preset to 1 second / cycle), then the rate of change per unit time = (current instantaneous value - adjacent cycle value) / time interval. The rate of change per unit time is the ratio of the change in the equivalent icing attenuation factor within an adjacent calibration cycle to the time interval. By introducing the rate of change parameter, the dynamic weights not only respond to the current icing state but also adapt to the changing trend of icing. The instantaneous value (5.06%) and the rate of change per unit time (0.2% / s) are combined into a two-dimensional feature vector [5.06, 0.2]. The first element of the vector is the instantaneous value of the decay factor (dimensionless, directly substituted as a percentage), and the second element is the rate of change per unit time (unit % / s, retaining the original dimensional value). This vector fully represents the "current state + trend" of ice accumulation decay, providing comprehensive input for the dynamic weight decision-maker. The pre-stored multilayer perceptron model in the dynamic weight decision-maker is invoked, and the two-dimensional feature vector [5.06, 0.2] is input into the trained model. After weighting by the channel fusion layer, the two-dimensional feature vector is transformed by the ReLU activation function of two hidden layers, and finally linearly mapped by the output layer to obtain the initial weight values. The initial weight values ​​are constrained by a limiting function, with the preset limiting function being... The weight limit is set at a lower limit of 0.05 (to ensure that the icing correction item has at least a basic weight) and an upper limit of 0.8 (to prevent the icing correction item from having an excessively large weight that masks the influence of other compensation items). Channel weight difference verification is added, and the coefficient of variation (CV = standard deviation / mean) of the attenuation factor for each channel is calculated. If CV > 0.3 (significant channel difference), the lower limit of the weight limit is increased to 0.08 (to enhance the weight of the icing correction item and adapt to channel differences); if CV ≤ 0.3 (small channel difference), the lower limit of 0.05 is maintained to ensure that the weight adapts to the differences while avoiding extreme values. This makes icing correction no longer a crude "on or off" switch, but a refined and intelligent adjustment based on the actual degree of impact, which significantly improves the calibration accuracy and robustness in complex and variable icing scenarios.

[0100] Furthermore, the gain-related portion of the icing correction term (the first correction amount) is multiplied by a dynamic weight to obtain a weighted gain correction amount, which is then combined with the intermediate gain amount to generate the final system gain coefficient. The offset-related portion of the icing correction term (the second correction amount) is multiplied by a dynamic weight to obtain a weighted offset correction amount, which is then combined with the intermediate offset amount to generate the final system offset. These are then integrated into the calibration parameters that have already been corrected for environmental and system response in a weighted manner to complete the synthetic compensation for all known error sources. The generated system gain coefficient and system offset together constitute a unified calibration coefficient that adapts to all channels, accommodates the attenuation differences of different channels, and further improves calibration accuracy. In addition, to simplify the engineering process, a unified weight is adopted, but a dedicated weight extension interface is reserved.

[0101] Furthermore, the calibration coefficient is the core basis for correcting the original radiation signal. The original radiation signal contains multiple interferences such as ice accumulation and environment. Distortion cannot be completely eliminated by a single correction. The signal needs to be systematically compensated and corrected by the calibration coefficient in order to obtain an effective signal that reflects the real cloud water ice accumulation state. Therefore, the calibration coefficient is used to correct the original radiation signal to ensure the pertinence and accuracy of the correction.

[0102] S6. In the microwave receiver subsystem of the radiometer, the calibration coefficient is used to compensate and correct the original radiation signal of the terahertz band received in real time, so as to obtain the effective terahertz detection signal after calibration for cloud water ice accumulation on the flight platform.

[0103] In this embodiment of the invention, the voltage value corresponding to the original radiation signal is the voltage value obtained by preliminary conversion of the terahertz frequency band signal received in real time by the airborne terahertz radiometer. The effective terahertz detection signal is a terahertz signal that has undergone complete dynamic calibration and can accurately characterize the cloud water icing state on the flight platform.

[0104] In this embodiment of the invention, the step of compensating and correcting the original radiation signal of the terahertz band received in real time using the calibration coefficient to obtain an effective terahertz detection signal calibrated for cloud water icing on the flight platform includes:

[0105] Identify the operating interval between the intermediate frequency amplification link and the analog-to-digital conversion operation of the microwave receiver subsystem, and within the operating interval, use the difference between the voltage value corresponding to the original radiated signal and the system offset in the calibration coefficient as the first-level scale for the gain fluctuation of the receiver link.

[0106] The ratio of the first-level scale to the system gain coefficient in the calibration coefficient is used as the second-level correction for noise floor drift in the terahertz band;

[0107] The signal after the second-level correction is used as the calibration brightness temperature digital sequence that reflects the real atmospheric radiation;

[0108] The calibration brightness temperature digital sequence is transmitted to the information processing unit of the airborne terahertz radiometer to obtain an effective terahertz detection signal calibrated for cloud water ice accumulation on the flight platform.

[0109] In detail, the intermediate frequency (IF) amplification link is the circuit link in the microwave receiver subsystem that performs IF amplification processing on terahertz signals; the analog-to-digital conversion (ADC) operation is the process of converting analog signals into digital signals. The output of the IF amplification link in the microwave receiver subsystem has a signal detection point, and the input of the ADC operation also has a signal detection point. By identifying the signal transmission path between these two detection points, the operating interval is determined. This interval is the pre-stage of analog-to-digital signal conversion, referring to the signal processing stage between the IF amplification link output and the ADC input. At this time, the signal still maintains a high degree of integrity, suitable for calibration calculations. The voltage value corresponding to the original radiated signal is obtained through the signal acquisition circuit of the microwave receiver subunit. For example, the original signal voltage value at a certain moment is acquired in real time as 2.331V (including gain fluctuation interference). The difference between the voltage value corresponding to the original radiated signal and the system offset in the calibration coefficient is used as the first-level scale for receiver link gain fluctuations. The first-level scale is for receiver link gain fluctuations, removing the DC bias of the signal. The result is a voltage signal with "zero bias" as the reference. The change in this signal will be caused purely by the change in radiation intensity received by the front-end antenna, thus completing the "first stage" basic correction for the DC drift and fixed deviation of the receiving link itself.

[0110] Specifically, noise floor drift refers to the baseline drift in terahertz frequency signals caused by electronic component noise. Dividing the first-level scale by the system gain coefficient in the calibration coefficient yields the second-level corrected signal. This operation, (voltage - offset) / gain coefficient, achieves a transition from the voltage domain to the temperature (brightness temperature), eliminating the gain fluctuations caused by temperature and power supply fluctuations in the receiving link. This unifies the signal strength scale, and through precise gain correction, ensures the stability of the system noise floor, allowing even subtle changes in sky radiation signals to be accurately extracted. Hence, this is called correction for "noise floor drift".

[0111] Furthermore, the calibration brightness temperature digital sequence is a sequence of digital signals that, after a second-level correction, reflects the true atmospheric radiation state. For example, using the second-level corrected digital signal 20340 as the calibration brightness temperature digital sequence, this sequence contains corrected digital values ​​at multiple consecutive moments. For instance, it contains 1000 data points within a 1-second calibration period, forming a sequence [20340, 20338, 20342,...]. Each data point corresponds to the true atmospheric radiation brightness temperature information at a specific moment. This "sequence" is a high-precision brightness temperature dataset arranged chronologically and dynamically compensated for in real time across the entire system. It eliminates the influence of the instrument itself and reflects the radiation characteristics of the real atmosphere (clouds, water vapor, ice crystals, etc.) in the terahertz band to the greatest extent possible, serving as a reliable input for all subsequent advanced inversion algorithms. The calibration brightness temperature digital sequence is transmitted via data bus to the information processing unit of the airborne terahertz radiometer. This unit is responsible for processing the calibrated digital signal and outputting a valid detection signal. It incorporates a digital signal processing chip (DSP) to perform filtering, smoothing, and feature extraction on the sequence. For example, it eliminates residual noise through mean filtering and supplements missing data through linear interpolation, ultimately outputting a valid terahertz detection signal. The brightness temperature value of this signal is converted to 255K (corresponding to the radiation brightness temperature characteristics of cloud water icing). Further processing optimizes the signal quality, ensuring that the output valid detection signal accurately characterizes the state of cloud water icing.

[0112] Furthermore, through two-stage correction and integrated processing, the original radiation signal, containing multiple interferences, was converted into a precise terahertz effective detection signal, completing the entire dynamic calibration process. This effective detection signal can accurately reflect the true state of cloud water icing on the flight platform, providing reliable data support for subsequent icing early warning and meteorological analysis.

[0113] like Figure 2 The diagram shown is a functional block diagram of a dynamic calibration system for cloud water ice detection of an airborne terahertz radiometer provided in an embodiment of the present invention.

[0114] The cloud ice detection dynamic calibration system 100 of the airborne terahertz radiometer described in this invention can be installed in an electronic device. Depending on the functions implemented, the airborne terahertz radiometer cloud ice detection dynamic calibration system 100 may include a data acquisition module 101, an influence function construction module 102, an equivalent ice attenuation factor analysis module 103, a reference value correction module 104, a calibration coefficient generation module 105, and a terahertz effective detection signal analysis module 106. The module described in this invention can also be called a unit, referring to a series of computer program segments that can be executed by the processor of an electronic device and perform a fixed function, stored in the memory of the electronic device.

[0115] In this embodiment, the functions of each module / unit are as follows:

[0116] The data acquisition module 101 is used to control the reference load switching device of the airborne terahertz radiometer, aligning it with the hot load and the warm load in sequence according to a preset cycle, and collecting the calibration matching parameters of the hot load and the warm load, as well as the temperature and humidity data of the external environment outside the flight platform.

[0117] The influence function construction module 102 is used to construct the influence function of antenna icing on the preset terahertz frequency band radiation efficiency based on the simulation data and historical experimental data of the antenna feeder sub-unit in the airborne terahertz radiometer.

[0118] The equivalent icing attenuation factor analysis module 103 is used to perform icing analysis on the external environmental temperature and humidity data using the influence function, and obtain the equivalent icing attenuation factor of the antenna in the current terahertz detection path.

[0119] The reference value correction module 104 is used to correct the target source radiation reference value stored in the airborne terahertz radiometer based on the equivalent icing attenuation factor to obtain the icing correction term, and to analyze the environmental compensation term of the external environment temperature and humidity data, and to analyze the system response term of the airborne terahertz radiometer according to the calibration matching parameters.

[0120] The calibration coefficient generation module 105 is used to synthesize the icing correction term, the environmental compensation term, and the system response term with the target source radiation reference value to generate the calibration coefficient of the original radiation signal on the flight platform.

[0121] The terahertz effective detection signal analysis module 106 is used in the microwave receiver subsystem of the radiometer to compensate and correct the original radiation signal of the terahertz band received in real time using the calibration coefficient, so as to obtain the terahertz effective detection signal after calibration for cloud water and ice accumulation on the flight platform.

[0122] In detail, the modules in the cloud water ice detection dynamic calibration system 100 of the airborne terahertz radiometer described in this embodiment of the invention adopt the same characteristics as described above during use. Figure 1 The method used is the same as the dynamic calibration method for cloud water ice detection of the airborne terahertz radiometer described in the article, and can produce the same technical effect, so it will not be repeated here.

[0123] In the several embodiments provided by this invention, it should be understood that the disclosed systems and methods can be implemented in other ways. For example, the system embodiments described above are merely illustrative; for instance, the division of modules is only a logical functional division, and other division methods may be used in actual implementation.

[0124] The modules described as separate components may or may not be physically separate. The components shown as modules may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs.

[0125] Furthermore, the functional modules in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or in the form of hardware plus software functional modules.

[0126] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the present invention can be implemented in other specific forms without departing from the spirit or essential characteristics of the present invention.

[0127] Therefore, the embodiments should be regarded as exemplary and non-limiting in all respects. The scope of the invention is not limited to the foregoing description, and all variations within the meaning and scope of equivalents falling within the protection scope are intended to be included in the invention.

[0128] The embodiments of this application can acquire and process relevant data based on artificial intelligence technology. Artificial intelligence (AI) refers to the theories, methods, technologies, and application systems that use digital computers or machines controlled by digital computers to simulate, extend, and expand human intelligence, perceive the environment, acquire knowledge, and use that knowledge to obtain optimal results.

[0129] Furthermore, it is clear that the word "comprising" does not exclude other units or steps, and the singular does not exclude the plural. Multiple units or systems stated in a system claim may also be implemented by a single unit or system through software or hardware. The terms "first," "second," etc., are used to indicate names and do not indicate any specific order.

[0130] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention.

Claims

1. A dynamic calibration method for cloud water icing detection using an airborne terahertz radiometer, characterized in that, The method includes: The reference load switching device of the airborne terahertz radiometer is controlled to align with the hot load and warm load in sequence according to a preset cycle, and the calibration matching parameters of the hot load and warm load, as well as the temperature and humidity data of the external environment outside the flight platform are collected. Based on the simulation data and historical experimental data of the antenna feeder unit in the airborne terahertz radiometer, a function is constructed to show the influence of antenna icing on the radiation efficiency of the preset terahertz frequency band on the flight platform. Using the aforementioned influence function, icing analysis is performed on the external environmental temperature and humidity data to obtain the equivalent icing attenuation factor of the antenna in the current terahertz detection path; The target source radiation reference value stored in the airborne terahertz radiometer is corrected based on the equivalent icing attenuation factor to obtain the icing correction term. The environmental compensation term of the external ambient temperature and humidity data is analyzed, and the system response term of the airborne terahertz radiometer is analyzed based on the calibration matching parameters. The icing correction term, the environmental compensation term, and the system response term are combined with the target source radiation reference value to generate the calibration coefficient of the original radiation signal on the flight platform. In the microwave receiver subsystem of the radiometer, the calibration coefficient is used to compensate and correct the original radiation signal of the terahertz band received in real time, so as to obtain an effective terahertz detection signal after calibration for cloud water ice accumulation on the flight platform.

2. The dynamic calibration method for cloud water icing detection using an airborne terahertz radiometer as described in claim 1, characterized in that, The process of sequentially aligning the thermal load and warm load according to a preset cycle, and collecting calibration matching parameters of the thermal load and warm load, as well as external environmental temperature and humidity data outside the flight platform, includes: During the first period of the preset cycle, the antenna of the airborne terahertz radiometer is pointed at the sky and a preset number of first raw voltage signals of sky radiation are continuously collected. At the same time, the temperature and humidity data of the external environment outside the flight platform are collected based on the temperature and humidity sensor. During the second time period of the preset cycle, the solenoid valve controlling the airborne terahertz radiometer drives the thermal load to move to the center of the antenna field of view and continuously collects a preset number of thermal load radiation signals. During the third time period of the preset cycle, the antenna of the airborne terahertz radiometer is pointed at the sky again, and a preset number of second raw voltage signals of sky radiation are continuously collected. During the fourth time period of the preset cycle, the solenoid valve controlling the airborne terahertz radiometer drives the warm load to move to the center of the antenna field of view and continuously collects a preset number of warm load radiation signals. The first original voltage signal, the thermal load radiation signal, the second original voltage signal, and the warm load radiation signal are combined into calibration matching parameters.

3. The dynamic calibration method for cloud water icing detection using an airborne terahertz radiometer as described in claim 1, characterized in that, The function for constructing the effect of antenna icing on the preset terahertz frequency band radiation efficiency on the flight platform, based on simulation data and historical experimental data of the antenna feeder unit in the airborne terahertz radiometer, includes: Extract the simulated radiation pattern dataset of multiple discrete frequency points of the target band of the antenna feeder unit under the preset target state; Extract the measured radiation efficiency attenuation dataset from the historical experimental data recorded by the antenna feeder unit under different controllable ice thickness conditions; The normalized ratio of the main lobe gain in the measured radiation efficiency attenuation dataset and the simulated radiation pattern dataset is fitted to a bivariate nonlinear attenuation function with ice thickness and ambient temperature as independent variables. The binary nonlinear attenuation function is physically constrained and verified based on the preset terahertz wave penetration depth. Once the verification is successful, the binary nonlinear attenuation function is output as the function that reflects the effect of antenna icing on the radiation efficiency of the preset terahertz frequency band.

4. The dynamic calibration method for cloud water icing detection using an airborne terahertz radiometer as described in claim 3, characterized in that, The step of performing icing analysis on the external environmental temperature and humidity data using the influence function to obtain the equivalent icing attenuation factor of the antenna in the current terahertz detection path includes: The current ambient temperature and relative humidity data of the flight platform are analyzed, and the current ambient temperature and relative humidity data are input into a preset dew point temperature calculation model to obtain the potential amount of liquid water adhering to the antenna reflector under the current environment. The amount of liquid water adhering is converted into the equivalent uniform ice thickness in the current terahertz detection path. The influence function is used to analyze the equivalent uniform ice thickness to obtain the theoretical radiation efficiency loss coefficient of each terahertz channel for antenna icing under the current operating conditions. Electrical matching state correction is applied to the theoretical radiation efficiency loss coefficient to obtain the equivalent icing attenuation factor of the antenna in the current terahertz detection path.

5. The dynamic calibration method for cloud water icing detection using an airborne terahertz radiometer as described in claim 1, characterized in that, The correction of the target source radiation reference value stored in the airborne terahertz radiometer based on the equivalent icing attenuation factor yields an icing correction term, including: Retrieve the cold target reference brightness temperature and hot target reference brightness temperature values ​​calibrated in a standard laboratory environment from the memory of the airborne terahertz radiometer; On the calibration line segment between the cold target reference brightness temperature value and the hot target reference brightness temperature value, a brightness temperature offset proportional to the equivalent ice accumulation attenuation factor is determined; The brightness temperature offset is converted into an equivalent noise temperature increment in the receive link of the airborne terahertz radiometer for antenna icing. The normalization factor characterizing the contribution weight of icing noise is determined based on the ratio between the equivalent noise temperature increment and the preset theoretical noise temperature of the antenna, and the icing correction term is calculated based on the normalization factor and the voltage coefficient of the total circuit in the receiving link.

6. The dynamic calibration method for cloud water icing detection using an airborne terahertz radiometer as described in claim 2, characterized in that, The analysis of the system response term of the airborne terahertz radiometer based on the calibration matching parameters includes: Obtain the first voltage value of the thermal load radiation signal corresponding to the thermal load and the second voltage value of the warm load radiation signal corresponding to the warm load from the calibration matching parameters; Obtain the first physical temperature for the thermal load and the second physical temperature for the warm load; The ratio of the difference between the first voltage value and the second voltage value to the difference between the first physical temperature and the second physical temperature is used as the system gain in the system response term; Calculate the product between the system gain and the second physical temperature, and use the difference between the first voltage value and the product result as the system offset in the system response term.

7. The dynamic calibration method for cloud water icing detection using an airborne terahertz radiometer as described in claim 1, characterized in that, The step of synthesizing the icing correction term, the environmental compensation term, and the system response term with the target source radiation reference value to generate calibration coefficients for the original radiation signal on the flight platform includes: The system response term is algebraically added to the target source radiation reference value to obtain the preliminary gain parameter and the preliminary offset parameter; Identify the gain compensation component and offset compensation component in the environmental compensation item according to the preset temperature and humidity-gain offset correction table, and superimpose the gain compensation component and the offset compensation component onto the initial gain parameter and the initial offset parameter respectively to obtain the intermediate gain quantity and the intermediate offset quantity; The icing correction term is assigned a dynamic weight that is adaptive to the equivalent icing attenuation factor, wherein the dynamic weight is used to adjust the degree of influence of the icing correction term on the intermediate gain and the intermediate offset. Multiply the first correction amount related to gain in the ice accumulation correction term by the dynamic weight to obtain the weighted gain correction amount, and combine the weighted gain correction amount with the intermediate gain amount to generate the system gain coefficient; The second correction amount related to the offset in the ice accumulation correction term is multiplied by the dynamic weight to obtain the weighted offset correction amount, and the weighted offset correction amount is combined with the intermediate offset amount to generate the system offset amount; The system gain coefficient and the system offset together constitute a calibration coefficient for real-time correction of the original radiation signal.

8. The dynamic calibration method for cloud water icing detection using an airborne terahertz radiometer as described in claim 7, characterized in that, The dynamic weighting of the icing correction term, which is adaptive to the equivalent icing attenuation factor, includes: A dynamic weight decision-maker for multi-channel changes is constructed based on a pre-defined neural network, wherein the dynamic weight decision-maker learns a non-linear mapping relationship from the equivalent ice accumulation attenuation factor to the weight coefficient. The equivalent icing attenuation factor of the current flight cycle on the flight platform is input into the dynamic weight decision-maker to obtain the instantaneous value of the equivalent icing attenuation factor and the rate of change of the equivalent icing attenuation factor per unit time in the adjacent calibration cycles of the current flight cycle. The instantaneous value and the rate of change per unit time are combined into a two-dimensional feature vector, and the two-dimensional feature vector is used as the input layer data of the dynamic weight decision-maker. The pre-stored multilayer perceptron model in the dynamic weight decision-maker is invoked to transform the two-dimensional feature vector sequentially through at least one hidden layer with a nonlinear activation function; The activation function of the output layer in the dynamic weight decision-maker is used to linearly map the transformed neuron activity value to obtain the initial weight value. The initial weight values ​​are constrained by physical characteristics using a preset limiting function to obtain dynamic weights.

9. The dynamic calibration method for cloud water icing detection using an airborne terahertz radiometer as described in claim 1, characterized in that, The process of compensating and correcting the original terahertz frequency band radiation signal received in real time using the calibration coefficient to obtain an effective terahertz detection signal calibrated for cloud water icing on the flight platform includes: Identify the operating interval between the intermediate frequency amplification link and the analog-to-digital conversion operation of the microwave receiver subsystem, and within the operating interval, use the difference between the voltage value corresponding to the original radiated signal and the system offset in the calibration coefficient as the first-level scale for the gain fluctuation of the receiver link. The ratio of the first-level scale to the system gain coefficient in the calibration coefficient is used as the second-level scale for noise floor drift in the terahertz band. The signal after the second-level scaling is used as a calibrated brightness temperature digital sequence that reflects the true atmospheric radiation; The calibration brightness temperature digital sequence is transmitted to the information processing unit of the airborne terahertz radiometer to obtain an effective terahertz detection signal calibrated for cloud water ice accumulation on the flight platform.

10. A dynamic calibration system for cloud water icing detection using an airborne terahertz radiometer, characterized in that, A method for dynamic calibration of cloud water icing detection using an airborne terahertz radiometer as described in any one of claims 1-9, the system comprising: The data acquisition module is used to control the reference load switching device of the airborne terahertz radiometer, aligning it with the hot load and warm load sequentially according to a preset cycle, and collecting the calibration matching parameters of the hot load and warm load, as well as the temperature and humidity data of the external environment outside the flight platform. The influence function construction module is used to construct the influence function of antenna icing on the preset terahertz frequency band radiation efficiency based on the simulation data and historical experimental data of the antenna feeder sub-unit in the airborne terahertz radiometer. The equivalent icing attenuation factor analysis module is used to perform icing analysis on the external environmental temperature and humidity data using the influence function, and to obtain the equivalent icing attenuation factor of the antenna in the current terahertz detection path. The reference value correction module is used to correct the target source radiation reference value stored in the airborne terahertz radiometer based on the equivalent icing attenuation factor to obtain the icing correction term, and to analyze the environmental compensation term of the external environment temperature and humidity data, and to analyze the system response term of the airborne terahertz radiometer according to the calibration matching parameters. The calibration coefficient generation module is used to synthesize the icing correction term, the environmental compensation term, and the system response term with the target source radiation reference value to generate the calibration coefficient of the original radiation signal on the flight platform. The terahertz effective detection signal analysis module is used in the microwave receiver subsystem of the radiometer to compensate and correct the original radiation signal of the terahertz band received in real time using the calibration coefficient, so as to obtain the terahertz effective detection signal after calibration for cloud water and ice accumulation on the flight platform.

Images