A double FBG icing and deicing detection system embedded between the primer and topcoat of an aircraft and a signal processing method thereof

Abstract

The application discloses a double FBG ice accumulation and deicing detection system embedded between an airplane primer and a topcoat and a signal processing method thereof, and belongs to the technical field of airplane state monitoring. The method comprises the following steps: embedding a double FBG sensor structure containing a temperature compensation grating and a strain sensing grating in an airplane wing coating and performing temperature-strain decoupling to obtain a pure strain component; performing intelligent filtering processing on the component to obtain a smooth strain signal; based on the signal and a change rate thereof, in combination with a preset threshold, dynamically identifying ice layer growth, peeling, stability or residual state; and finally, quantitatively evaluating a deicing effect based on sequence similarity analysis. A corresponding system comprises a sensing module, a signal demodulation module and a data processing and decision module. Through in-coating embedded sensing and multi-stage signal processing, the application realizes high-precision and dynamic monitoring of the airplane ice accumulation and deicing process, and effectively solves the problems of temperature interference and noise influence.

Description

Technical Field

[0001] This invention relates to the field of aircraft condition monitoring technology, specifically to a dual FBG icing and de-icing detection system embedded between the aircraft primer and topcoat, and its signal processing method. Background Technology

[0002] Aircraft icing is a significant safety hazard during ground operations and takeoff. It not only alters the aerodynamic properties of the wing surface, increasing drag and reducing lift, but also causes structural interference to control surfaces, sensor mounting areas, and engine air intakes, potentially leading to mechanical failures. Therefore, real-time and accurate identification of icing conditions is a crucial prerequisite for ensuring flight safety.

[0003] Currently, icing monitoring on aprons mainly relies on technologies such as visual inspection, infrared imaging, capacitive sensors, and single FBG sensors. Visual inspection depends on the experience of maintenance personnel, and its accuracy is unstable and carries high operational risks under adverse weather conditions such as nighttime, heavy snowfall, and low-temperature frost. Infrared imaging, while enabling non-contact detection, is easily affected by ambient background light, changes in the smoothness of the ice surface, and residual de-icing fluid, making it difficult to effectively distinguish between ice and water accumulation. Capacitive sensors judge based on changes in dielectric constant, but their measurement response exhibits significant errors under conditions such as strong electromagnetic interference and changes in dielectric properties due to freezing temperatures. The single FBG sensor method uses Bragg wavelength drift to determine changes in ice load. Although it has advantages such as corrosion resistance and the ability to be embedded in the grating, the grating is sensitive to both temperature and strain, making it difficult to distinguish between "thermal effects" and "stress effects" from wavelength drift, leading to inaccurate icing assessments and becoming a key bottleneck restricting its accurate application.

[0004] Current technologies for icing detection primarily focus on static icing identification, i.e., determining the presence of ice layers. However, they lack real-time monitoring methods for dynamic information such as ice peeling behavior during de-icing, assessment of de-icing fluid effectiveness, and the thoroughness of de-icing. In actual operation, insufficient de-icing leads to residual ice buildup, affecting takeoff safety, while excessive de-icing results in energy waste and increased economic costs. Therefore, there is an urgent need for a new aircraft icing and de-icing detection technology that can simultaneously address the temperature-strain cross-sensitivity issue, suppress measurement data noise interference, and provide real-time monitoring capabilities for de-icing effectiveness. This would meet the application requirements of aviation operations for high reliability, visualization, and intelligent monitoring. Summary of the Invention

[0005] The purpose of this invention is to provide a dual FBG icing and de-icing detection system and its signal processing method embedded between the aircraft primer and topcoat. By changing the sensor from the traditional "surface attachment" to "coating embedding", and combining dual grating decoupling technology and intelligent filtering algorithm, the system can achieve accurate separation of temperature-strain components, and at the same time, can monitor the ice layer thickness change and de-icing status throughout the entire process.

[0006] To achieve the above objectives, the technical solution adopted by the present invention is as follows: A method for processing icing and de-icing detection signals of a dual-FBG (aircraft primer and topcoat) system embedded between the aircraft primer and topcoat includes the following steps: S1: In the icy region of the aircraft wing, a dual FBG sensor structure containing a temperature compensation grating and a strain sensing grating is embedded between the primer layer and the topcoat layer, and the measurement signal of the dual FBG sensor structure is subjected to temperature-strain decoupling calculation to obtain the pure strain component characterizing the ice load. S2: Perform data preprocessing and intelligent filtering on the pure strain components to obtain a smoothed strain signal; S3: Based on the smoothed strain signal and its first derivative value, dynamic comparison and state identification are performed in combination with a preset threshold. S4: Based on the results of state identification, sequence similarity analysis is used to quantitatively evaluate the de-icing effect.

[0007] In a specific implementation scheme, in step S1, the temperature-strain decoupling calculation specifically involves: obtaining the change in the center wavelength of the temperature compensation grating and the change in the center wavelength of the strain sensing grating, and calculating the pure strain component based on the difference between the two.

[0008] In one specific implementation, in step S2, the intelligent filtering process employs an adaptive state estimation algorithm.

[0009] In one specific implementation scheme, the adaptive state estimation algorithm is a Kalman filter algorithm, and the process noise covariance and measurement noise covariance of the Kalman filter algorithm are adjusted online based on the statistical characteristics of the pure strain component or the smoothed strain signal.

[0010] In one specific implementation scheme, in step S3, the preset thresholds include: a threshold for the absolute value of strain obtained statistically based on the ice-free state, a positive threshold for the rate of strain change, and a negative threshold for the rate of strain change.

[0011] In one specific implementation, the rules for state identification in step S3 include: When the smoothed strain signal is greater than the absolute value threshold of the strain and its first derivative is greater than the positive threshold of the strain change rate, it is determined that the ice layer is growing. When the smoothed strain signal is greater than the absolute value threshold of the strain and its first derivative is less than the negative threshold of the strain change rate, it is determined that the ice layer is peeling off. When the smoothed strain signal is greater than the absolute value threshold of the strain, and its first derivative value is between the negative threshold and the positive threshold of the strain change rate, it is determined that the ice growth has stopped. When the smoothed strain signal is less than the absolute strain value threshold and its first derivative value is between the negative and positive thresholds of the strain change rate, the ice layer is determined to be stable or has residue.

[0012] In a specific implementation scheme, step S4, the sequence similarity analysis method includes: calculating the derivative dynamic time regularization distance, regularization path deviation, and correlation coefficient between the strain characteristic sequence to be evaluated and the ice-free benchmark sequence, and fusing them to generate a cleanliness assessment index.

[0013] In one specific implementation scheme, the cleanliness assessment index is compared with a preset numerical range to output a quantitative assessment result.

[0014] Another object of the present invention is to provide a dual FBG icing and de-icing detection system embedded between the aircraft primer and topcoat, the system comprising: The sensing module includes a fiber optic sensing array embedded in the surface of an aircraft wing in a coating manner, and comprising a temperature compensation grating and a strain sensing grating. A signal demodulation module, connected to the sensing module, is used to acquire the center wavelength data of each grating in real time; The data processing and decision module is connected to the signal demodulation module and is used to execute the steps in the signal processing method.

[0015] In one specific implementation scheme, in the dual FBG icing and de-icing detection system embedded between the aircraft primer and topcoat, the fiber optic sensor array of the sensing module is arranged in a grid pattern, and the arrangement density in the aerodynamically critical areas of the wing is higher than that in non-critical areas.

[0016] In summary, the present invention has the following beneficial technical effects: This invention achieves effective separation of temperature and strain components through an embedded double FBG structure design, thereby improving measurement accuracy. This invention combines multi-level signal processing and dynamic criteria to monitor the entire process of ice formation, growth, and peeling in real time. This invention introduces a quantitative evaluation mechanism, providing a basis for judging the thoroughness of de-icing operations, which helps to improve flight safety and maintenance efficiency. Attached Figure Description

[0017] Figure 1 This is a schematic diagram of a dual FBG ice buildup and de-icing detection system embedded between the aircraft primer and topcoat, provided by an embodiment of the present invention. Figure 2 A flowchart illustrating a method for processing icing and de-icing detection signals of a dual FBG embedded between the aircraft primer and topcoat, provided in an embodiment of the present invention; In the picture: 1-Aircraft topcoat, 2-Aircraft primer, 3-Strain sensing grating, 4-Temperature compensation grating, 5-Aircraft skin, 6-Fiber optic demodulator, 7-Computer. Detailed Implementation

[0018] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention. Furthermore, the technical features involved in the various embodiments of this invention described below can be combined with each other as long as they do not conflict with each other.

[0019] A complete process for processing signals for detecting icing and de-icing in a dual-FBG (airplane primer and topcoat) system embedded between the aircraft primer and topcoat includes the following steps: S1: In the icy region of the aircraft wing, a dual FBG sensor structure containing a temperature compensation grating and a strain sensing grating is embedded between the primer layer and the topcoat layer, and the measurement signal of the dual FBG sensor structure is subjected to temperature-strain decoupling calculation to obtain the pure strain component characterizing the ice load. S2: Perform data preprocessing and intelligent filtering on the pure strain components to obtain a smoothed strain signal; S3: Based on the smoothed strain signal and its first derivative value, dynamic comparison and state identification are performed in combination with a preset threshold. S4: Based on the results of state identification, sequence similarity analysis is used to quantitatively evaluate the de-icing effect.

[0020] The overall flow of the signal processing method of the present invention is as follows: Figure 2 As shown, the main steps include: In step S1, when deploying FBGs in areas prone to icing on the aircraft, a "grid-based deployment + zoned densification" strategy is adopted. A high-density grid layout with a grid spacing of 25–50 mm is used in the leading edge area and flap / aileron hinge area; a medium-density grid with a grid spacing of 50–100 mm is used in the middle of the wing surface; and a sparse grid of 100–200 mm is used in non-critical aerodynamic areas to reduce fiber length and construction work.

[0021] Each grid node is equipped with a pair of dual FBG sensing units, including a temperature-compensated grating. With strain grating The spacing between the two is 5–20 mm to ensure temperature field consistency. All gratings are C-band reflective FBGs, with center wavelengths arranged at intervals of 0.8–1.5 nm to meet the spectral resolution of the demodulator (typical resolution 1 pm) and avoid grating spectral line overlap. This allows for the acquisition of two-dimensional strain field distributions in each region.

[0022] During the sensor deployment phase, the aircraft wing surface must first undergo a pre-treatment process involving coating. Once the primer has cured to a stage suitable for embedding, fiber optic cable laying proceeds according to the grid layout diagram. (Strain grating) The temperature-compensated grating needs to be located within the actual strain transfer region of the grid node. By using flexible resin coating to isolate the fiber from the substrate, the fiber needs to be laid naturally, adhering to the primer layer without stretching. Subsequently, stress relief tubes are laid along the fiber channel, and a relaxation ring with a minimum bending radius of 3 times is reserved at the outlet to prevent the fiber from being pulled by the topcoat application or aircraft vibration.

[0023] After the optical fiber and primer are embedded and fixed, the topcoat encapsulation and structural curing stage begins. The topcoat is applied according to aircraft painting standards, ensuring complete coverage of the optical fiber path and grating area, with the optical fiber positioned within the primer and topcoat layers. Topcoat curing requires temperature control within the allowable range for both the optical fiber and resin to prevent permanent drift. Temperature grating coating resin can be cured at room temperature or low temperature. Strain grating bonding must ensure effective transfer of ice layer stress and temperature-induced coating deformation, avoiding thermal stress or tensile damage to the grating throughout the process.

[0024] After completing temperature-strain decoupling and ensuring the system is in normal operating condition, baseline noise testing is required under steady-state conditions without ice or load to obtain the statistical threshold parameters needed for the icing criterion. This invention preferably selects a stable ambient temperature and a wing surface free of frost, water accumulation, and stress disturbances, continuously acquiring the strain signal ε(t) and its first derivative signal. The sampling duration is preferably no less than 1 hour, and the sampling frequency is preferably 10–100Hz. This is to cover the typical range of variations in air disturbances, temperature fluctuations, and the demodulator's noise floor.

[0025] The mean value of the continuously acquired strain sequences is calculated. with standard deviation The noise standard deviation of its first derivative sequence is calculated. These parameters are used for subsequent threshold settings.

[0026] When ice accumulates on a part of the aircraft or during de-icing, the wavelength of the FBG sensor at that location changes, and the strain change is decoupled through dual FBG.

[0027] A dual-FBG structure is embedded between the aircraft primer and topcoat. The temperature-compensating grating, encapsulated in a low-elasticity resin layer, is used solely to measure the effect of temperature changes on the FBG wavelength. The strain-sensing grating, cured and bonded to the primer layer, senses the strain in the coating caused by changes in ice stress. The two FBGs are arranged adjacently on the same optical fiber. By comparing the difference in their wavelength changes, the temperature and strain components can be separated, thus eliminating interference from ambient temperature fluctuations on strain measurements.

[0028] The center reflection wavelength (Bragg wavelength) of the FBG satisfies: in, The original center wavelength of the FBG For effective refractive index, As the grating period, temperature and strain both affect the effective refractive index and the grating period, thus causing changes in the center wavelength.

[0029] Change in the center wavelength of FBG satisfy: in, These are the strain sensitivity coefficient and the temperature sensitivity coefficient, respectively. This refers to the axial strain and temperature changes of the fiber Bragg grating.

[0030] When ice buildup occurs or is cleared, the data measured by the strain-sensing grating and temperature-compensated grating are used for strain analysis. Temperature decoupling, the decoupling of temperature and strain, can be expressed as: in, The change in the center wavelength of the strain sensing grating. This represents the change in the center wavelength of the temperature-compensated grating. The calculation yields the pure strain component caused solely by ice load. This effectively eliminates temperature interference.

[0031] In step S2, the original strain signal may be affected by interference (such as light source jitter, instantaneous reflection of the connector, etc.) due to potential interference to the fiber optic demodulator. It may contain outliers and baseline drift. First, outliers are removed using the 3σ principle, where... The mean of the original strain sequence is given. Let its standard deviation be . If a point satisfies: If it is identified as an outlier, it will be replaced by interpolation using the average value of the preceding and following points. The FBG strain signal will slowly shift up and down (the entire curve shifts) in the absence of ice and external force. Common causes include slight shrinkage during coating curing, slow temperature fluctuations due to incomplete decoupling, thermal expansion and contraction of the aircraft wing surface, long-term creep of the adhesive layer, and slow changes in the power of the demodulator light source. These changes are unrelated to strain changes caused by ice formation or de-icing.

[0032] In practical implementation, the strain sequence after outlier removal... Baseline estimation using the moving average method : in, M is the baseline at time t, and M is the length of the sliding window (5-10 seconds). The correction signal was then obtained: , When the first segment of data has less than M points, the cumulative average is used as a substitute.

[0033] To obtain strain and strain rate signals with high signal-to-noise ratio, this invention employs one-dimensional state-space Kalman filtering for high-frequency denoising and smoothing of the strain sequence after outlier removal and baseline correction. The processing steps are as follows.

[0034] in, The actual strain state at time t. For process noise, it represents unmodelable abrupt changes such as coating stress release and ice growth; The strain value is measured by the FBG demodulator. To measure noise, this indicates the demodulator's electronic noise, spectral fluctuations, etc.

[0035] Then, state initialization is performed, setting initial estimates. And the variance P(0) is used to determine the degree of confidence the filter has in the measurements in the initial stage. This is followed by the prediction step. To make prior predictions about the strain at the current moment, and through =P(t-1)+Q propagation prediction error, where Q is the process noise covariance, so that the filter has a reasonable level of uncertainty before entering a new round of measurement.

[0036] In the update step, the gain K(t) is calculated: Where R is the measurement noise covariance.

[0037] The observation residuals are weighted by calculating the gain. in, The strain prior estimate at time t is given. To estimate the error covariance a priori, Let be the process noise covariance.

[0038] The predicted values ​​are corrected to obtain the posterior estimate at the current time. The estimated variance P(t) is then updated. To improve the filter's adaptability in complex environments, this invention introduces an adaptive update strategy for noise covariance. The updates of covariances Q and R satisfy the following formulas: in , The smoothing coefficient is typically specified to be close to 1, which is the final filtered output. The filtered signal is used as a smoothed strain sequence in subsequent decision-making steps. It both suppressed high-frequency noise and maintained the continuity of the ice layer change trend.

[0039] In step S3, firstly, the strain sequence after intelligent filtering is processed... Calculate the first derivative , in, The sampling period.

[0040] Baseline data were collected under ice-free steady-state conditions, and the standard deviation of the strain sequence was statistically analyzed. and the standard deviation of the derivative sequence Based on this, a threshold is set: Strain threshold: ,in It is a positive integer multiple coefficient (usually taken as 2~4); Positive threshold for strain rate of change: This corresponds to the growth of the ice layer; Negative threshold for strain rate of change: This corresponds to the peeling of the ice layer; among which It is a positive coefficient (usually taken as 2~3).

[0041] The filtered strain value is compared with the first derivative of the strain and a threshold to determine the current ice accumulation or de-icing status.

[0042] During the judgment process, when and At that time, it was determined that the ice layer was growing; and At that time, it was determined that the ice layer was peeling or melting; and At that time, it is determined that ice growth has stopped; and At that time, it was determined that the ice layer was stable or that a small amount of ice remained.

[0043] In step S4, after completing the filtering and dynamic determination steps, this invention sets up a de-icing status determination module. This module uses the following sequence similarity analysis method to quantitatively evaluate the de-icing effect: further, residual ice is identified in the strain sequence after de-icing. To overcome the problems of traditional Euclidean distance DTW being sensitive to baseline drift and having poor robustness of single threshold decision, this embodiment proposes a fusion decision method based on derivative weighted dynamic time warping (DDTW) and path deviation analysis.

[0044] This method introduces a feature sequence Q(t), which integrates the strain amplitude and its first derivative characteristics, to enhance the ability to capture subtle morphological distortions caused by ice layers. in, Weighting coefficient (value 0.5) and The strain and its derivative sequence were measured separately using optical fibers. and The mean and standard deviation of the strain sequence are given. and The mean and standard deviation of the derivative sequence are given by the baseline sequence. With the sequence to be tested Perform the analysis.

[0045] Calculated using dynamic time warping algorithm and Shortest matching distance The recursive formula is: in, For sequence The first i point and The cumulative distance between the first j points, ultimately , The sequence length is given.

[0046] During this process, the optimal regularization path is recorded. ,in Indicates the baseline sequence number Point and the sequence to be tested Point alignment.

[0047] This invention innovatively introduces regular path deviation (PDA) as a discrimination parameter. In the ice-free state, due to the consistent system response, the regular path tends to be close to the diagonal. The damping and stiffness changes caused by residual ice will lead to a phase shift in the response, resulting in path bending. The PDA calculation is as follows: PDA characterizes the phase hysteresis of the signal response.

[0048] Subsequently, the Pearson correlation coefficient was used to calculate the overall correlation between the two sequences, and its formula is as follows: in, and They are respectively and The zero-mean sequence.

[0049] To eliminate the random errors of a single indicator, a dimensionless de-icing cleanliness fusion index (CCI) is constructed. This index integrates the Pearson correlation coefficient (r) and the DWT distance. and path deviation PDA in, and For empirical normalization constants, , , As a weighting factor, satisfying The CCI value is divided into three intervals for output: Completely cleaned area When de-icing is complete, the system outputs a "no ice" signal. Trace Residue Area If the surface is found to have residual de-icing fluid or small ice fragments, initiate the next round of continuous monitoring and verification. Areas where de-icing is incomplete If significant ice accumulation is detected, the system will automatically trigger an alarm and identify the corresponding grid area.

[0050] This embodiment achieves high-precision dynamic monitoring of wing icing formation, peeling, and de-icing processes by embedding a dual fiber Bragg grating (FBG) sensing unit between the aircraft wing primer and topcoat. Combined with multi-level signal processing methods including temperature-strain decoupling, outlier removal, sliding baseline correction, adaptive Kalman filtering, threshold determination, and sequence similarity analysis, it enables precise dynamic monitoring of wing icing formation, peeling, and de-icing. The thermodynamically compatible and stress-sensitive arrangement of the temperature-compensated grating and strain grating effectively eliminates the influence of environmental temperature disturbances on strain measurements. The zoned densification and gridded arrangement strategy ensures high-resolution sensing of ice loads in key areas, achieving accurate mapping of local and global two-dimensional strain fields. Dynamic determination based on mean-standard deviation statistical thresholds can identify ice growth, peeling, and stabilization states in real time. Furthermore, through joint analysis with the correlation coefficient and dynamic time warping distance of the ice-free baseline sequence, it quantitatively identifies residual ice and local stress anomalies after de-icing, significantly improving the reliability and robustness of de-icing completion determination. The embedded fiber optic structure balances aerodynamic integrity and stress transfer efficiency, achieving high system integration and engineering feasibility. This embodiment not only significantly improves the accuracy and real-time performance of aircraft icing and de-icing monitoring, but also reduces the amount of de-icing fluid used and flight safety risks, thus having significant engineering application value and promising prospects for promotion.

[0051] Another embodiment of the present invention provides a dual FBG icing and de-icing detection system embedded between the aircraft primer and topcoat, corresponding to the above method, the system comprising: The sensing module includes a fiber optic sensing array embedded in the surface of an aircraft wing in a coating manner, and comprising a temperature compensation grating and a strain sensing grating. A signal demodulation module, connected to the sensing module, is used to acquire the center wavelength data of each grating in real time; The data processing and decision module is connected to the signal demodulation module and is used to execute the steps in the signal processing method.

[0052] like Figure 1 As shown, the system of the present invention mainly includes a sensing module, a signal demodulation module, and a data processing and decision module. The sensing module consists of a dual fiber Bragg grating (FBG) sensor embedded between the aircraft primer 2 and the aircraft topcoat 1, including a strain sensing grating 3 and a temperature compensation grating 4. The two gratings are located on the same optical fiber and arranged adjacent to each other, with a preferred spacing of 5~20mm to ensure that they are in similar temperature fields.

[0053] The temperature compensation grating 4 is encapsulated in a low-elasticity flexible resin and isolated from the wing skin 5, sensing only temperature changes; the strain sensing grating 3 is cured and bonded to the aircraft primer 2, directly sensing the coating strain changes caused by ice load. This arrangement changes the sensor from traditional surface mounting to coating embedding, ensuring mechanical coupling between the sensor and the structure while avoiding significant impact on the aircraft's aerodynamic shape and painting process.

[0054] In areas prone to icing on the aircraft, such as the wing leading edge and flap / aileron hinge area, a "grid-based layout + zoned densification" strategy is adopted. The grid spacing is 25-50 mm in critical areas, 50-100 mm in the middle of the wing surface, and 100-200 mm in non-critical areas. Each grid node is equipped with a pair of dual FBG sensing units. All gratings are C-band reflective FBGs, with center wavelengths arranged at intervals of 0.8-1.5 nm to avoid spectral line overlap and facilitate resolution by the fiber optic demodulator.

[0055] During fiber optic cable laying, it must be kept naturally relaxed and without stretching, and stress relief tubes and relaxation rings should be installed at the exit point to prevent damage to the fiber optic cable caused by coating or vibration. After the aircraft topcoat 1 is applied, the fiber optic cable is completely encapsulated within the interlayer between the aircraft primer 2 and the aircraft topcoat 1, forming a stable and reliable embedded sensing structure.

[0056] The signal demodulation module is implemented by the fiber optic demodulator 6, which is responsible for acquiring and analyzing the wavelength offset of each FBG; the data processing and decision module is completed by the computer 7, which analyzes, processes and determines the icing status of the demodulated signal.

[0057] Contents not described in detail in this specification are prior art known to those skilled in the art. It is hereby indicated that the above description is intended to help those skilled in the art understand this invention, but does not limit the scope of protection of this invention. Any equivalent substitutions, modifications, improvements, or simplifications of the above descriptions that do not depart from the essential content of this invention fall within the scope of protection of this invention.

Claims

1. A method for processing signals for detecting icing and de-icing in a dual-FBG array embedded between the aircraft primer and topcoat, characterized in that, Includes the following steps: S1: In the icy region of the aircraft wing, a dual FBG sensor structure containing a temperature compensation grating and a strain sensing grating is embedded between the primer layer and the topcoat layer, and the measurement signal of the dual FBG sensor structure is subjected to temperature-strain decoupling calculation to obtain the pure strain component characterizing the ice load. S2: Perform data preprocessing and intelligent filtering on the pure strain components to obtain a smoothed strain signal; S3: Based on the smoothed strain signal and its first derivative value, dynamic comparison and state identification are performed in combination with a preset threshold. S4: Based on the results of state identification, sequence similarity analysis is used to quantitatively evaluate the de-icing effect.

2. The method for processing signals for detecting icing and de-icing using a dual FBG embedded between the aircraft primer and topcoat, as described in claim 1, is characterized in that... In step S1, the temperature-strain decoupling calculation specifically involves: obtaining the change in the center wavelength of the temperature compensation grating and the change in the center wavelength of the strain sensing grating, and calculating the pure strain component based on the difference between the two.

3. The dual FBG icing and de-icing detection signal processing method embedded between the aircraft primer and topcoat according to claim 1, characterized in that, In step S2, the intelligent filtering process employs an adaptive state estimation algorithm.

4. The method for processing signals for detecting icing and de-icing using a dual FBG embedded between the aircraft primer and topcoat, as described in claim 3, is characterized in that... The adaptive state estimation algorithm is a Kalman filter algorithm. The process noise covariance and measurement noise covariance of the Kalman filter algorithm are adjusted online based on the statistical characteristics of the pure strain component or the smoothed strain signal.

5. The method for processing signals for detecting icing and de-icing using a dual FBG embedded between the aircraft primer and topcoat, as described in claim 1, is characterized in that... In step S3, the preset thresholds include: a threshold for the absolute value of strain obtained statistically based on the ice-free state, a positive threshold for the rate of strain change, and a negative threshold for the rate of strain change.

6. The method for processing signals for detecting icing and de-icing in a dual FBG array embedded between the aircraft primer and topcoat, as described in claim 5, is characterized in that... In step S3, the state recognition rules include: When the smoothed strain signal is greater than the absolute value threshold of the strain and its first derivative is greater than the positive threshold of the strain change rate, it is determined that the ice layer is growing. When the smoothed strain signal is greater than the absolute value threshold of the strain and its first derivative is less than the negative threshold of the strain change rate, it is determined that the ice layer is peeling off. When the smoothed strain signal is greater than the absolute value threshold of the strain, and its first derivative value is between the negative threshold and the positive threshold of the strain change rate, it is determined that the ice growth has stopped. When the smoothed strain signal is less than the absolute strain value threshold and its first derivative value is between the negative and positive thresholds of the strain change rate, the ice layer is determined to be stable or has residue.

7. The method for processing signals for detecting icing and de-icing using a dual FBG embedded between the aircraft primer and topcoat, as described in claim 1, is characterized in that... In step S4, the sequence similarity analysis method includes: calculating the derivative dynamic time regularization distance, regularization path deviation, and correlation coefficient between the strain characteristic sequence to be evaluated and the ice-free benchmark sequence, and then fusing them to generate a cleanliness assessment index.

8. The method for processing signals for detecting icing and de-icing using a dual FBG embedded between the aircraft primer and topcoat, as described in claim 7, is characterized in that... The cleanliness assessment index is compared with a preset numerical range to output a quantitative assessment result.

9. A dual FBG icing and de-icing detection system embedded between the aircraft primer and topcoat for implementing the method of any one of claims 1-8, characterized in that, include: The sensing module includes a fiber optic sensing array embedded in the surface of an aircraft wing in a coating manner, and comprising a temperature compensation grating and a strain sensing grating. A signal demodulation module, connected to the sensing module, is used to acquire the center wavelength data of each grating in real time; The data processing and decision module is connected to the signal demodulation module and is used to execute the steps in the signal processing method.

10. The dual FBG icing and de-icing detection system embedded between the aircraft primer and topcoat according to claim 9, characterized in that, The fiber optic sensing array is arranged in a grid pattern, with a higher density in the aerodynamically critical areas of the wing than in the non-critical areas.

Images