Helicopter cable fault detection system and method

Abstract

The invention provides a helicopter cable fault detection system and method, and belongs to the technical field of helicopter electrical system detection. The system comprises a positioning and matching device and a multi-modal signal analysis device, the positioning and matching device determines the relative position with a to-be-detected aviation plug through a dual-spectrum camera, generates a contact coordinate matrix and a contact and probe mapping table, injects a test signal to identify the contact material so as to obtain a topology result, and detects the contact pressure; and the multi-modal signal analysis device is used for matching parameters based on the topology result and the mapping table, setting a detection signal, positioning the contact, applying a signal in combination with pressure, acquiring a response signal, establishing a mapping table, and analyzing the response signal to generate a contact and fault result mapping table. According to the invention, the whole process automation from aviation plug positioning, signal detection to fault analysis is realized, the system can complete the core links of contact coordinate generation, signal mapping, fault determination and the like without manual intervention, and the detection efficiency is greatly improved.

Description

Technical Field

[0001] This invention relates to the field of helicopter electrical system testing technology, specifically to a helicopter cable fault detection system and method. Background Technology

[0002] In existing helicopter cable fault detection, traditional methods rely on manual point-by-point testing, which has low automation, low efficiency, and is prone to misjudgment due to unstable contact pressure. At the same time, existing equipment mostly supports single parameter detection and cannot simultaneously analyze multi-modal faults such as continuity, insulation, and crosstalk. It is also difficult to adapt to working conditions such as confined spaces, complex layouts, and plug deformation. Fault diagnosis does not combine topological information such as contact material and functional classification, resulting in problems of single detection dimension and insufficient accuracy. There is an urgent need for an automated detection system that integrates positioning matching and multi-modal signal analysis to solve the above-mentioned defects. Summary of the Invention

[0003] The technical problem to be solved by the present invention is to provide a helicopter cable fault detection system and method to address the shortcomings of the prior art.

[0004] The technical solution of the present invention to solve the above-mentioned technical problems is as follows: a helicopter cable fault detection system, including a positioning and matching device and a multi-modal signal analysis device; The positioning and matching device is used for: The relative position of the aircraft plug to be tested is determined by the image of the aircraft plug to be tested and the contact distribution template acquired by the dual-spectrum camera, a contact coordinate matrix is ​​generated, and a contact-probe mapping table is generated based on the integrated probe array and the contact coordinate matrix. Inject test signals to identify contact materials and obtain plug topology identification results; Contact pressure status data is obtained by detecting the contact pressure at the contact point using a force sensor; The multimodal signal analysis device is used for: Based on the plug topology identification result and the multimodal detection parameters corresponding to the contact and probe mapping table, a multimodal detection signal is set based on the multimodal detection parameters; Using the contact point coordinate matrix as a spatial reference, the position of each designated contact point is located. Combined with the contact pressure state data, the multimodal detection signal of the probe in the probe array is applied to multiple designated contact points to obtain the response signal corresponding to each designated contact point. A contact point and response signal mapping table is established through the mapping relationship between contact points and probes in the contact point and probe mapping table. The response signal is subjected to contact fault analysis to obtain fault results, and a contact-fault result mapping table is generated based on the fault results and the contact-response signal mapping table.

[0005] Another technical solution of the present invention to solve the above-mentioned technical problems is as follows: A helicopter cable fault detection method, comprising the following steps: The relative position of the aircraft plug to be tested is determined by the image of the aircraft plug to be tested and the contact distribution template acquired by the dual-spectrum camera, a contact coordinate matrix is ​​generated, and a contact-probe mapping table is generated based on the integrated probe array and the contact coordinate matrix. Inject test signals to identify contact materials and obtain plug topology identification results; Contact pressure status data is obtained by detecting the contact pressure at the contact point using a force sensor; Based on the plug topology identification result and the multimodal detection parameters corresponding to the contact and probe mapping table, a multimodal detection signal is set based on the multimodal detection parameters; Using the contact point coordinate matrix as a spatial reference, the position of each designated contact point is located. Combined with the contact pressure state data, the multimodal detection signal of the probe in the probe array is applied to multiple designated contact points to obtain the response signal corresponding to each designated contact point. A contact point and response signal mapping table is established through the mapping relationship between contact points and probes in the contact point and probe mapping table. The response signal is subjected to contact fault analysis to obtain fault results, and a contact-fault result mapping table is generated based on the fault results and the contact-response signal mapping table.

[0006] The beneficial effects of this invention are: by using a positioning and matching device and a multimodal signal analysis device, the entire process from aircraft plug positioning and signal detection to fault analysis is automated. The system can complete core steps such as contact coordinate generation, signal mapping, and fault determination without human intervention, which greatly improves detection efficiency. At the same time, the collaborative work of multiple devices ensures the continuity from positioning to fault result output, reduces human operation errors, and improves the reliability of detection results. Attached Figure Description

[0007] Figure 1 This is a block diagram of a helicopter cable fault detection system provided in an embodiment of the present invention; Figure 2 A flowchart of a helicopter cable fault detection method provided in an embodiment of the present invention; Figure 3 This is a schematic diagram of the positioning and matching device provided in an embodiment of the present invention; Figure 4 A schematic diagram illustrating the adaptive scaling principle of a probe provided in an embodiment of the present invention.

[0008] In the attached diagram, the component names represented by each label are as follows: 1. Probe array, 2. OLED display, 3. Wireless communication module. Detailed Implementation

[0009] The principles and features of the present invention are described below with reference to the accompanying drawings. The examples given are only for explaining the present invention and are not intended to limit the scope of the present invention.

[0010] Example 1: Figure 1 As shown, this embodiment of the invention provides a helicopter cable fault detection system, including a positioning and matching device and a multimodal signal analysis device; The positioning and matching device is used for: The relative position of the aircraft plug to be tested is determined by the image of the aircraft plug to be tested and the contact distribution template acquired by the dual-spectrum camera, a contact coordinate matrix is ​​generated, and a contact-probe mapping table is generated based on the integrated probe array and the contact coordinate matrix. Inject test signals to identify contact materials and obtain plug topology identification results; Contact pressure status data is obtained by detecting the contact pressure at the contact point using a force sensor; The multimodal signal analysis device is used for: Based on the plug topology identification result and the multimodal detection parameters corresponding to the contact and probe mapping table, a multimodal detection signal is set based on the multimodal detection parameters; Using the contact point coordinate matrix as a spatial reference, the position of each designated contact point is located. Combined with the contact pressure state data, the multimodal detection signal of the probe in the probe array is applied to multiple designated contact points to obtain the response signal corresponding to each designated contact point. A contact point and response signal mapping table is established through the mapping relationship between contact points and probes in the contact point and probe mapping table. The response signal is subjected to contact fault analysis to obtain fault results, and a contact-fault result mapping table is generated based on the fault results and the contact-response signal mapping table.

[0011] In the above embodiments, the positioning and matching device and the multimodal signal analysis device realize the full automation of the process from aviation plug positioning and signal detection to fault analysis. The system can complete the core links such as contact coordinate generation, signal mapping and fault determination without human intervention, which greatly improves the detection efficiency. At the same time, the collaborative work of multiple devices ensures the continuity from positioning to fault result output, reduces human operation error and improves the reliability of detection results.

[0012] Preferably, the process of determining the relative position of the aircraft plug to be tested with the image of the aircraft plug to be tested acquired by the dual-spectrum camera and the contact distribution template, generating a contact coordinate matrix, and generating a contact-probe mapping table based on the integrated probe array and the contact coordinate matrix includes: Visible light and infrared images of the aircraft plug to be inspected are acquired simultaneously using a dual-spectrum camera. The visible light and infrared images are then fused to obtain a fused image. Key plug features are extracted from the fused image, and then matched with the contact distribution template of a standard plug model to obtain the image matching result. A device coordinate system is established with the integrated probe array reference point as the origin, and a plug local coordinate system is established with the plug geometric center as the origin. Based on the image matching results, the transformation matrix of the two coordinate systems is calculated through a perspective transformation algorithm to determine the three-dimensional relative position of the probe array and the aircraft plug to be inspected. Based on the transformation matrix, each contact point is transformed from the local coordinate system of the plug to the coordinate system of the device. Combined with the sub-pixel level positioning of the actual contact points in the fused image, a contact point coordinate matrix containing the three-dimensional coordinates of each contact point is generated. The three-dimensional coordinates of each probe are determined based on the device coordinate system. Through a spatial coordinate comparison algorithm, the three-dimensional coordinates of each probe in the probe array are matched one by one with the three-dimensional coordinates of each contact in the contact coordinate matrix. A contact-probe mapping table is generated based on the matching results.

[0013] Specifically, in this embodiment, such as Figures 3-4 As shown, probe array 1 is integrated into the head of the positioning and matching device and includes multiple sets of self-adaptive probes. Each probe is made of shape memory alloy and is telescopically adjustable (0.5mm~5mm stroke) to adapt to aviation plugs with different apertures. Each probe in probe array 1 has a built-in micro motor to drive its extension and retraction, and the contact pressure threshold is set to 0.5N to avoid damaging the plug.

[0014] Visual positioning: The positioning and matching device integrates a dual-spectrum camera (acquiring visible light + infrared images) in its head, and identifies the plug outline and thermal characteristics of metal contacts through image recognition.

[0015] Combining the SLAM algorithm (Simultaneous Localization and Mapping), the relative position of the device and the plug is calculated in real time (accuracy ±0.1mm), and a dynamic guide arrow is displayed on the OLED display screen 2.

[0016] Electromagnetic adsorption control: The head edge of the positioning and matching device is embedded with a ring electromagnet array, which generates a local magnetic field (the intensity is adjustable from 0.1T to 1.2T) when energized.

[0017] When the camera detects that the plug has entered the effective range (≤5cm), the electromagnet is activated, and the device is tightly attached to the plug surface by magnetic attraction (compatible with metal / composite plugs).

[0018] In this embodiment, the probe array 1 also has anti-offset redundancy function and probe adaptive matching function: The anti-offset redundancy function is as follows: 1) A flexible silicone buffer layer is set in the central area of ​​the electromagnet to avoid hard contact that could damage the plug surface.

[0019] 2) If the plug is made of non-magnetic material (such as carbon fiber), activate the vacuum adsorption mode: the head micro air pump generates negative pressure (-50kPa), which is temporarily fixed by the annular sealing ring.

[0020] like Figure 4 As shown, the probe adaptive matching function is as follows: a. Multi-degree-of-freedom probe array 1) Probe structure: Each probe is individually packaged and contains: Shape memory alloy telescopic rod (Ni-Ti material, which can extend and retract 0.5mm to 5mm in response to temperature changes).

[0021] Piezoelectric ceramic micro-motion platform (X / Y axis ±2mm offset, to compensate for pin position deviation).

[0022] Spherical contacts with nano-silver coating (reduce contact resistance and adapt to oxide layer surface).

[0023] 2) Dynamically adjust the process: Coarse matching: Based on a preset plug model database (such as the MIL-DTL-38999 series), the corresponding contact distribution template is called to drive the probe group to the theoretical position.

[0024] Fine calibration: After the probe contacts the socket, the contact pressure is fed back by a triaxial miniature force sensor (range 0N~5N, resolution 0.01N). If the pressure is <0.3N, the piezoelectric ceramic fine-tuning probe angle is triggered until the pressure stabilizes at 0.5N±0.1N.

[0025] 3) Fault tolerance and security mechanisms Overload protection: If the pressure of a probe is greater than 1.5N (possibly due to plugging or misalignment of the socket), the probe will retract immediately and an alarm will be triggered on the interface.

[0026] Self-cleaning function: After the probe is reset, it automatically releases a high-voltage electrostatic pulse (5kV / 1ms) to remove oxides or contaminants from the contact surface.

[0027] The contact mapping algorithm based on the above probe array is as follows: a. Topology identification and dynamic modeling Signal injection method: 1) The control module sends low-voltage pulses (5V / 10mA) to two randomly selected probes, and uses a time domain reflectometer (TDR) to detect the reflected waveform of the signal and construct a diagram of the internal wire connection relationship of the plug.

[0028] 2) Combining graph neural networks (GNN), the plug type is identified within 10ms by comparing the measured topology with the standard topology library (containing 300+ helicopter cable models).

[0029] b. Contact-Probe Map Generation If no matching model is found in the database, the self-learning mode will be activated: 1) The probe array traverses the plug surface in a honeycomb scanning pattern to record the coordinates and electrical characteristics (resistance, capacitance, etc.) of each contact.

[0030] 2) Generate a contact point distribution matrix using a sparse coding algorithm, store it as a new model, and upload it to the cloud for sharing.

[0031] Based on the above embodiments, the process of human-machine collaborative verification of probe contact with touch points can be visualized through an OLED screen: AR-assisted confirmation: The OLED screen overlays an augmented reality (AR) interface, marking matched touch points with semi-transparent highlighted frames, prompting the operator to manually verify key points (such as the flight control system power cord).

[0032] Audio-visual feedback: After matching is complete, the device vibrates and emits a prompt sound, and the green LED ring light stays on.

[0033] Main unit: Carbon fiber shell (weight <500g), built-in lithium battery, control module, and wireless communication module 3 (Bluetooth / Wi-Fi). The circuit design uses a low-power STM32 chip to control signal generation and acquisition, and is equipped with isolation circuits to prevent high-voltage interference.

[0034] In the above embodiments, dual-spectral image fusion technology is used to combine visible light (geometric contour) and infrared (thermal feature) information, which enhances the extraction accuracy of key plug features and solves the problem of feature blurring under a single spectrum. The coordinate system transformation is achieved through perspective transformation algorithm, combined with sub-pixel level positioning, which ensures the three-dimensional spatial accuracy of the contact point coordinate matrix (±0.1mm), laying the foundation for accurate matching between the probe and the contact point; The spatial coordinate comparison algorithm enables automated matching between the probe array and the contacts, avoiding the tediousness and errors of manual alignment, and adapting to the contact layout of different plug models.

[0035] Preferably, the injection of test signals to identify contact materials to obtain plug topology identification results includes: Based on the common material characteristics of the aviation plug to be tested, a suitable test signal is selected from the system's preset signal library, and the test signal is injected into a number of pre-specified probes in the probe array. After the designated probe carrying the test signal comes into contact with the contact of the aviation plug to be tested, the signal acquisition module of the probe array acquires the response signal of each contact to the test signal in real time. Based on the collected response signals, the material type of each contact point is identified through a material feature matching algorithm; Based on the material type of each contact and the signal transmission relationship between the contacts, the contact connection topology of the aviation plug to be tested is constructed, and the contact connection topology is compared with the standard topology template of the corresponding plug model in the system database to obtain the plug topology recognition result.

[0036] In this embodiment, based on the collected response signals, the material type of each contact point is identified through a material feature matching algorithm, specifically: Response signal preprocessing: The response signals of each contact point collected by the probe array (such as the attenuation rate, reflection coefficient, conductivity and other data of the test signal) are preprocessed, including noise reduction (such as wavelet thresholding), signal smoothing (such as moving average filtering) and normalization (mapping the signal feature values ​​to the range of 0~1), eliminating environmental interference (such as electromagnetic noise) and signal fluctuations caused by contact instability, and extracting stable signal features.

[0037] Material feature parameter extraction: From the preprocessed response signal, feature parameters strongly correlated with the material's physical properties are extracted, including: Signal attenuation characteristics: The attenuation rate of the test signal at a specific frequency (different materials have different conductivity, resulting in different attenuation rates; for example, the attenuation rate of gold-plated materials is lower than that of copper). Reflection coefficient: The intensity of the reflected signal of the test signal on the contact surface (related to the surface impedance of the material; the reflection coefficient of silver alloy is higher than that of copper). Temperature coefficient: Combined with the thermal response of infrared images, the temperature change rate of the contact point during signal transmission is extracted (different materials have different thermal conductivity, resulting in different temperature coefficients).

[0038] Call the material feature standard library: retrieve the preset standard feature templates of common contact materials (such as copper, gold plating, silver alloy, etc.) from the system database. The templates contain the characteristic parameter ranges of each material under the same test signal (such as attenuation rate range, reflection coefficient reference value, etc.) as the matching reference.

[0039] Feature matching and material determination: Using Euclidean distance or cosine similarity algorithms, the extracted features of the touch point to be detected are compared with standard feature templates to calculate the similarity value. If a material template has the highest similarity to the feature to be detected and exceeds a preset threshold (e.g., similarity ≥ 90%), then the touch point is determined to be of that material type. If the similarity of all templates is below the threshold, the system marks them as "unknown material" and triggers a secondary detection (re-collecting and analyzing after adjusting the test signal parameters).

[0040] Through the above process, accurate identification of contact material based on response signal characteristics is achieved, providing a foundation for subsequent plug topology construction and multimodal detection parameter matching.

[0041] In the above embodiments, the material identification technology based on test signals can accurately distinguish the contact material (such as copper, gold plating, etc.) by matching response signal features, providing a basis for subsequent multimodal detection parameter adaptation; By combining material type and signal conduction relationship to construct topology structure and comparing it with standard template, the internal connection relationship of plug can be quickly identified, solving the problem of inefficiency of traditional manual topology inspection; It is applicable to topology testing of different plug models, has strong versatility, and is especially suitable for aviation plugs with complex contact layouts.

[0042] Preferably, the step of setting a multimodal detection signal based on the multimodal detection parameters corresponding to the plug topology identification result and the contact-probe mapping table includes: Key information is extracted from the topology identification results, including contact material type, contact function classification, connection relationship between contacts, and standard electrical characteristics of the corresponding plug model. Import the historical response signal characteristics of the corresponding probes for each contact point based on the contact point and probe mapping table, and determine the detection priority and signal adaptation requirements of each contact point based on the probe and historical response signal characteristics. Based on the detection priority and signal adaptation requirements of each contact, corresponding multi-mode detection parameters are matched from the preset parameter library. The multi-mode detection parameters include: continuity detection parameters, insulation detection parameters, and crosstalk detection parameters. The multimodal detection signal is set according to the multimodal detection parameters. The multimodal detection signal includes a bipolar pulse pair corresponding to the on / off detection parameters, a stepped DC signal corresponding to the insulation detection parameters, and a high-frequency swept sine wave corresponding to the crosstalk detection parameters.

[0043] Specifically, in this embodiment, the multimodal detection parameters are designed as follows: 1) On / off detection signal: Bipolar pulse pair (BPP) is used.

[0044] Positive pulse +Vp (pulse width tp) Negative pulse -Vp (pulse width tp).

[0045] Function: To offset the contact potential difference and improve the signal-to-noise ratio (SNR>60dB) of weak signals.

[0046] Input logic: Automatically select Vp (copper: 5V; gold-plated: 3V) based on the contact material (copper / gold-plated).

[0047] 2) Insulation resistance detection signal: Inject a step-ramp DC signal.

[0048] Initial voltage V start =10V, incremented in ΔV = 5V steps, up to a maximum of V. max =50V.

[0049] The slope k = ∆I / ∆V was calculated by least-squares fitting of the relationship between leakage current Ileak and voltage. Riso = 1 / k.

[0050] Crosstalk detection signal: A high-frequency sweep sine wave (10kHz-10MHz, VAC=1VRMS) is applied to the non-tested contact, and coherent detection is performed through the receiving end of the adjacent contact.

[0051] Extracting the amplitude A of the same frequency signal using a lock-in amplifier. coup C coup =A coup / (2πfZ0), where Z0=50.

[0052] In the above embodiments, based on the plug topology results (material, functional classification, etc.) and the probe's historical response characteristics, "personalized matching" of detection parameters is achieved, avoiding the problem of insufficient adaptability of general parameters to different contacts; The targeted design of multi-modal detection parameters (on / off, insulation, crosstalk) and corresponding signals (bipolar pulse pairs, stepped DC, high-frequency sweep) ensures that each type of fault can be accurately detected, improving the comprehensiveness of fault identification. The combination of a preset parameter library and FPGA time-sharing scheduling enables efficient switching and application of detection signals, thereby improving detection speed.

[0053] Preferably, the step of locating the position of each designated contact point using the contact point coordinate matrix as a spatial reference, and applying the multimodal detection signal of the probes in the probe array to multiple designated contacts in combination with the contact pressure state data to obtain the response signal corresponding to each designated contact point includes: The position of the specified contact point is determined by determining the three-dimensional coordinates of each specified contact point in the device coordinate system based on the contact point coordinate matrix. Based on the position of the specified contact point, the corresponding probe in the probe array is moved to the position of the specified contact point. The contact pressure status data between the probe and the specified contact point is monitored in real time by the force sensor. If the pressure deviates from the effective threshold range, the piezoelectric ceramic micro-motion platform is triggered to fine-tune the angle of the corresponding probe or the memory alloy telescopic rod is adjusted to adjust the length until the contact pressure stabilizes within the effective threshold range. Based on the FPGA time-sharing scheduling strategy, multimodal detection signals are applied one by one to multiple designated contacts through the aligned probes; The signal acquisition module of the probe array captures the response signals of each designated contact to the detection signal in real time, and associates the acquired response signals with the corresponding designated contact based on the contact-probe mapping table to establish the contact-response signal mapping table.

[0054] In the above embodiments, the contact coordinate matrix is ​​used as a spatial reference, and the probe is adjusted with multiple degrees of freedom (piezoelectric ceramic fine adjustment, shape memory alloy extension and retraction) to ensure the physical alignment accuracy between the probe and the contact, and solve the problem of poor contact caused by small deformation of the plug or installation deviation. The force sensor monitors the contact pressure in real time and adjusts it dynamically (stabilizing it within the effective threshold range), ensuring the stability of signal transmission and avoiding problems such as too little pressure (weak signal) or too much pressure (damage to the plug). The FPGA time-sharing scheduling strategy enables interference-free application of multimodal signals and precise correlation between response signals and contacts, providing high-quality data for subsequent fault analysis.

[0055] Preferably, the step of performing contact fault analysis on the response signal to obtain fault results includes: The on / off detection of the response signal is specifically performed as follows: The response signal is a bipolar pulse pair response signal obtained through bipolar pulse pair detection. Time-domain processing is performed on the bipolar pulse pair response signal to obtain the rise time of the bipolar pulse pair response waveform. The bipolar pulse response signal is processed in the frequency domain to obtain the fundamental frequency component energy of the response waveform after FFT transformation. Substituting the rise time and fundamental frequency component energy of the bipolar pulse response waveform into the fuzzy logic membership function, the membership degree of the contact open circuit state is calculated. The fuzzy logic membership function is as follows: , in, The membership degree of the contact open circuit state, when If the value is greater than 0.8, it is determined to be an open circuit (output 0); otherwise, it is a conducting circuit (output 1). The rise time of the bipolar pulse response waveform (10%~90% amplitude) is given by t. r >0.2tp It is determined to be in a high-resistivity state. The threshold for determining the rise time is... For the fundamental frequency component energy, if E fund <0.5E total It was determined to be an open circuit. The threshold for determining the fundamental frequency energy. This is an adjustment coefficient used to control the steepness of the membership function.

[0056] In this embodiment, a signal time-division multiplexing control method is used: 1) Allocation (FPGA hard real-time scheduling) The time period (0-10 ms) corresponds to the continuity / disconnection detection mode. The time period (10-30 ms) corresponds to the insulation detection mode. The time period (30-50) ms corresponds to the crosstalk detection mode.

[0057] 2) Conflict avoidance: During continuity detection, the insulation / crosstalk signal source is turned off to prevent superimposed interference.

[0058] In the above embodiments, the continuity detection combines time domain (rise time) and frequency domain (fundamental frequency energy) features to avoid misjudgment caused by a single feature (such as looking only at the time domain) (such as signal delay caused by contact oxidation). Fuzzy logic membership function quantifies the membership degree of the circuit breaker state by fusing two-dimensional features, which improves the robustness of the judgment (such as the fault tolerance of edge cases) and reduces the false judgment rate. It is suitable for contact continuity testing of different materials and contact states, and has strong versatility.

[0059] Preferably, the step of performing contact fault analysis on the response signal to obtain fault results further includes: Insulation detection is performed on the response signal, specifically as follows: The response signal is a leakage current sequence under stepped voltage obtained by stepped DC signal detection. Where n is the voltage step number, Based on the physical properties of insulation resistance, a nonlinear equation relating leakage current and voltage is constructed: , in, This refers to the leakage current in the leakage current sequence under stepped voltage. For the applied step voltage, The insulation resistance calculated for the target. This represents the contact resistance between the probe and the contact point. The dielectric polarization effect coefficient; The parameters in the nonlinear relational equation are based on the Levenberg-Marquardt algorithm. , and The fitting process is as follows: Setting parameters , and The initial value; By minimizing the measured leakage current With predicted leakage current Sum of squared errors insulation resistance Perform multiple iterative calculations, when When the iteration stops, output the fitted insulation resistance. , The insulation resistance obtained by fitting If the insulation resistance is less than the insulation resistance threshold, it is determined to be an insulation failure; otherwise, the insulation performance is determined to be normal.

[0060] In the above embodiments, a stepped DC signal and a nonlinear fitting model are introduced, taking into account contact resistance (probe and contact) and dielectric polarization effect (insulating material characteristics), which solves the problem of insufficient accuracy caused by the traditional linear model ignoring interference factors; The Levenberg-Marquardt algorithm achieves insulation resistance by minimizing the sum of squared errors (termination condition < 1 μA²). The high-precision calculations ensure the accuracy of insulation condition determination; Strict threshold comparison (with insulation resistance judgment threshold) can accurately identify insulation failure faults, providing a reliable basis for safety protection in high-voltage scenarios.

[0061] Preferably, the step of performing contact fault analysis on the response signal to obtain fault results further includes: Crosstalk detection is performed on the response signal, specifically as follows: The response signal is a sweep response amplitude sequence obtained by high-frequency sweep sine wave detection. Where m is the number of sweep points, The frequency of the high-frequency swept sine wave. Based on the frequency characteristics of crosstalk signals, a sparse basis matrix is ​​constructed to characterize the sparse distribution features of the crosstalk coupling coefficients. , Based on the orthogonal matching pursuit (OMP) algorithm and the sparse basis matrix The crosstalk coupling coefficient is calculated from the sweep frequency response amplitude sequence and expressed as: , in, This represents the sweep frequency response amplitude at the corresponding frequency. For random downsampling matrix, The crosstalk coupling coefficient is... This is the error threshold; Based on crosstalk coupling coefficient By combining the frequency parameters of the high-frequency swept sine wave, the crosstalk coupling coefficient between each pair of contacts is calculated: , in, The amplitude of the coupling signal between contact i and contact j. For the corresponding frequency, Characteristic impedance; The calculated crosstalk coupling coefficient between the two contacts The crosstalk coupling coefficient is compared with the threshold. If it is greater than the threshold, it is judged as a crosstalk exceeding the standard fault; otherwise, it is judged as a normal crosstalk characteristic.

[0062] In the above embodiments, a high-frequency swept sine wave (10kHz~10MHz) is used to cover a wide frequency range, which can capture crosstalk characteristics under different contact spacing and is suitable for contact interference detection in complex layouts. By combining sparse basis matrices with the orthogonal matching pursuit (OMP) algorithm, the crosstalk coupling coefficient matrix can be efficiently reconstructed using compressed sensing technology, reducing the amount of data collected and improving computational efficiency. The coupling coefficient calculation formula based on characteristic impedance (Z0=50Ω) ensures the accuracy of crosstalk exceedance judgment and can effectively identify the risk of signal interference between contacts.

[0063] Preferably, the multimodal signal analysis device is further configured to: after obtaining the fault result, perform a health assessment on the aircraft plug to be tested based on the fault result, specifically: The obtained fault results are constructed into a three-dimensional feature vector. , The membership degree for the contact open circuit state. For insulation resistance, This is the crosstalk coupling coefficient matrix. For each contact point, the comprehensive fault probability corresponding to the three-dimensional feature vector is calculated using a weighted summation formula: , in, , and for , and The corresponding weight, where k is the contact number. The threshold for determining insulation resistance. Determine the threshold for the crosstalk coupling coefficient; Calculate the overall health score of the aviation connector based on the combined failure probability of all contacts: , Where N is the total number of contacts. The importance weight of touch point k, When S≥80, the patient is considered healthy. When 60 ≤ S < 80, it is considered a mild abnormality. When S < 60, it is judged as a serious fault.

[0064] Specifically, the on / off resistance threshold =10、 =100M、 =5pF.

[0065] In the above embodiments, the on / off state is integrated ( ),insulation( Crosstalk The three-dimensional feature vector enables the fusion of multi-dimensional fault information, avoiding the one-sidedness of evaluating a single fault type. The health score quantifies the overall condition of the plug, and the classification (healthy / minor abnormality / serious fault) provides an intuitive and scientific basis for maintenance decisions, improving the pertinence and efficiency of maintenance.

[0066] Based on the above embodiments, during the data transmission process between the positioning and matching device and the multimodal signal analysis device, wireless transmission and data encryption are employed, and results are generated using a dynamic template method. (1) System architecture and core modules The hierarchical data stream processing architecture enables efficient transmission of detection results and intelligent report generation, including a data compression and encryption layer, an adaptive transmission layer, a multimodal reporting engine, and an interactive visualization layer.

[0067] a. Input: ① On / off state matrix M open ② Insulation resistance matrix M iso ③ Crosstalk coupling matrix M coup ④ Health score S∈[0,100] ⑤ Contact topology mapping table T map (Including physical coordinates and function labels) b. Output: ① Encrypted compressed data packets (DTX) (wireless transmission) ②Multi-format test reports (PDF / Excel / AR-HUD) ③ Dynamic knowledge graph (fault root cause reasoning and maintenance suggestions) (2) Data compression and secure transmission algorithms a. Hybrid Entropy Coding Compression (HEC) ① Input: Original matrix M open , M iso , M coup ②Algorithm: Sparsity treatment: For M iso With M coup Perform threshold filtering (preserving R) iso <1G, C coup (Unhealthy data > 1pF), generating a sparse matrix. , '.

[0068] By using the improved CSR (Compressed Sparse Row) format to store the coordinates and values ​​of non-zero elements, the compression rate is increased to 92%.

[0069] Context-adaptive arithmetic coding: Constructing a two-level Markov model: Level 1 model: Analyze the fault correlation of contact function labels (power / signal / ground).

[0070] Second-level model: Predicts the probability distribution of the current failure mode based on historical maintenance records. Entropy encoding of sparse matrices based on dynamic probability tables further improves the compression rate by 35%.

[0071] ③ Output: Compressed data stream Dcompressed (98% smaller than the original data volume).

[0072] b. Quantum Noise Confusion Encryption (QNCE) ① Input: Dcompressed ②Algorithm: Key generation: The hardware fingerprint Hpuf is extracted using the device's built-in PUF (Physically Unclonable Function) chip and combined with the current timestamp t to generate a dynamic key. , The encryption process is as follows: D compressed The blocks are divided into 64-byte units, each of which is XORed and obfuscated with a quantum random number sequence (generated by the chip TRNG).

[0073] Superimposed chaotic mapping scrambling: (r=3.999, x0=K) dynamic mod1) The data blocks are cyclically shifted according to a chaotic sequence.

[0074] ③ Output: Ciphertext data packet D tx It meets the quantum attack resistance requirements of the national cryptographic standard SM4.

[0075] (3) Adaptive wireless transmission protocol a. Multi-band dynamic frequency hopping (MBDFH) ① Input: D tx Current channel quality metrics (CQI) – packet loss rate and signal-to-noise ratio. ②Algorithm: Frequency band selection strategy: Real-time scanning of the 2.4GHz / 5.8GHz / 60GHz frequency bands to construct the channel state matrix Q. (Bandwidth, interference, attenuation, multipath delay).

[0076] The optimal frequency band is dynamically selected using a multi-armed slot machine (MAB) model to maximize throughput. : , Where h is the channel gain and τ_delay is the delay penalty factor.

[0077] Fragmented redundant transmission: Divide the Dtx into N data slices, and attach a redundant fountain code (RaptorQ) packet to each slice.

[0078] The receiving end dynamically requests retransmissions based on channel quality to ensure complete data recovery even with a 30% packet loss rate.

[0079] ③ Output: Complete encrypted data received by the terminal .

[0080] (4) Multimodal intelligent report generation engine a. Semantic Network-Driven Reporting Structure Ratio (SN-Report) ① Input: Decrypted M open , M iso , M coup T map ②Algorithm: Fault entity extraction: Based on a predefined aviation electrical ontology library (such as SAE-AS50881), the BiLSTM-CRF model is used to identify the subsystem (flight control / navigation / power) to which the fault contact belongs.

[0081] Output the set of faulty entities E= .

[0082] Finally, dynamic template generation: Construct a conditional random field (CRF) driven template selector to adaptively select the report skeleton based on the fault type and quantity: Single point of failure Concise report (focusing on topology diagram and maintenance steps).

[0083] Multi-point associated failure In-depth analysis report (including simulation of fault propagation path).

[0084] ③ Output: Semantically enhanced intermediate report structure Rsemantic.

[0085] b. Visualization rendering of multiphysics coupling ① Input: R semantic Terminal device type (mobile phone / tablet / AR glasses, etc.) ②Algorithm: 3D electrical field reconstruction: M iso With M coup Mapping the data to a 3D plug model (i.e., SLAM data), the electric field distribution in the weak insulation zone is calculated using the finite element method (FEM). (x, y, z).

[0086] Monte Carlo ray tracing is used to generate heatmaps, which are then overlaid onto the AR display interface.

[0087] Adaptive rendering strategy: Mobile: The interactive topology map is implemented using WebGL 2.0, which supports gesture zooming and fault point drill-down.

[0088] AR glasses: Spatial anchoring of report content is achieved through SLAM-Visual Inertial Odometry (VIO) fusion, and historical maintenance records automatically pop up when maintenance personnel look at a certain contact point.

[0089] ③ Output: Interactive visual report R adapted for the device visual .

[0090] (5) Knowledge graph-driven decision support That is, KG-Reasoner. ① Input: E (Collection of Fault Entities), Model Maintenance Manual (PDF text) ②Algorithm: Heterogeneous knowledge fusion: The BERT-wwm model was used to extract maintenance specifications from the manual and construct a rule graph Grule.

[0091] Perform subgraph isomorphic matching between the current fault E and the historical case library (stored in the graph database Neo4j) to find the top k cases with the highest similarity.

[0092] Probabilistic inference engine: Define Bayesian network nodes as follows: contact material aging (P=0.3), vibration causing loosening (P=0.4), moisture infiltration (P=0.2), etc.

[0093] The maximum posterior probability root cause is calculated using Markov chain Monte Carlo (MCMC) sampling, and the confidence level Proot∈[0,1] is generated.

[0094] ③ Output: Repair suggestion list A = {a_1(action, tools, P_success), ..., a_m}.

[0095] (6) Implementation of verification Taking the avionics system testing of the Black Hawk helicopter as an example a. Input data: ①Mopen: Contact 17 is open (1 0 jump) ②Miso: Riso = 85M between contacts 29-43 ③Tmap: Contact 17 belongs to the flight control bus, and contact 29 belongs to the fire control power supply line. b. Output report: ① PDF report: Highlight the 3D location of the faulty contact and the electric field distribution cloud map.

[0096] Repair recommendation: Replace plug J27 (89% confidence level), using a torque screwdriver (5.5 N•m).

[0097] ②AR visualization: After the maintenance personnel put on HoloLens, a red flashing frame was superimposed on the surface of the faulty plug. When looking at contact 17, three similar fault records were displayed.

[0098] ③ Knowledge graph update: The system automatically associates this failure mode with the "vibration stress accumulation" root cause node, triggering a preventative checklist for the same batch of helicopters.

[0099] Example 2: As Figure 2 As shown, the present invention also provides a method for detecting helicopter cable faults, comprising the following steps: S1. The relative position of the aircraft plug to be tested with the image of the aircraft plug to be tested and the contact distribution template acquired by the dual-spectrum camera is determined, a contact coordinate matrix is ​​generated, and a contact-probe mapping table is generated based on the integrated probe array and the contact coordinate matrix. S2. Inject test signals to identify contact materials and obtain plug topology identification results; S3. The contact pressure status data is obtained by detecting the contact pressure at the contact point using a force sensor; S4. Based on the plug topology identification result and the multimodal detection parameters corresponding to the contact and probe mapping table, set the multimodal detection signal based on the multimodal detection parameters; S5. Using the contact point coordinate matrix as a spatial reference, locate the position of each designated contact point. Combine the contact pressure state data to apply the multimodal detection signal of the probe in the probe array to multiple designated contact points to obtain the response signal corresponding to each designated contact point. Establish a contact point and response signal mapping table through the mapping relationship between contact points and probes in the contact point and probe mapping table. S6. Perform contact fault analysis on the response signal to obtain fault results, and generate a contact-fault result mapping table based on the fault results and the contact-response signal mapping table.

[0100] The workflow for helicopter cable fault detection is as follows: 1. Align the head of the device with the aviation plug to be tested, and the probe will automatically attach and complete the contact matching.

[0101] 2. Trigger one-key detection: The device scans all contacts with pulse signals to detect continuity, insulation resistance, and crosstalk.

[0102] 3. The results are wirelessly transmitted to the terminal (such as a display screen) and a test report is generated (supports PDF / Excel format).

[0103] Taking a 100-pin connector as an example: a. Input data: ① The contact point coordinate matrix Pos

[100] [3] (x, y, z coordinates). ②Preset thresholds Rth=10, Riso=100M, Ccoup=5pF.

[0104] b. Output results: ① Fault report: Contact 23 open circuit (Ropen=∞), insulation failure between contacts 45-67 (Riso=15M). ② Health score S=82.3 (Main deduction item: Deterioration of insulation of contact points in flight control system). c. Real-time performance: Total detection time is 50ms (meeting the "100 needles completed in 10s" target).

[0105] It should be noted that, in this document, relational terms such as first and second, etc., are used only to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the terms "comprises," "comprising," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that includes a list of elements includes not only those elements but also other elements not explicitly listed, or elements inherent to such process, method, article, or apparatus.

[0106] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working process of the above-described apparatus and unit can be referred to the corresponding process in the foregoing method embodiments, and will not be repeated here.

[0107] In the several embodiments provided in this application, it should be understood that the disclosed apparatus and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative. For instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed.

[0108] The units described as separate components may or may not be physically separate. The components shown as units 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 units can be selected to achieve the purpose of the embodiments of the present invention, depending on actual needs.

[0109] Furthermore, the functional units 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 as a software functional unit.

[0110] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods of the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0111] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A helicopter cable fault detection system, characterized in that, Includes a positioning and matching device and a multimodal signal analysis device; The positioning and matching device is used for: The relative position of the aircraft plug to be tested is determined by the image of the aircraft plug to be tested and the contact distribution template acquired by the dual-spectrum camera, a contact coordinate matrix is ​​generated, and a contact-probe mapping table is generated based on the integrated probe array and the contact coordinate matrix. Inject test signals to identify contact materials and obtain plug topology identification results; Contact pressure status data is obtained by detecting the contact pressure at the contact point using a force sensor; The multimodal signal analysis device is used for: Based on the plug topology identification result and the multimodal detection parameters corresponding to the contact and probe mapping table, a multimodal detection signal is set based on the multimodal detection parameters; Using the contact point coordinate matrix as a spatial reference, the position of each designated contact point is located. Combined with the contact pressure state data, the multimodal detection signal of the probe in the probe array is applied to multiple designated contact points to obtain the response signal corresponding to each designated contact point. A contact point and response signal mapping table is established through the mapping relationship between contact points and probes in the contact point and probe mapping table. The response signal is subjected to contact fault analysis to obtain fault results, and a contact-fault result mapping table is generated based on the fault results and the contact-response signal mapping table.

2. The helicopter cable fault detection system according to claim 1, characterized in that, The relative position of the aircraft plug to be tested to the aircraft plug is determined by the image of the aircraft plug to be tested and the contact distribution template acquired by the dual-spectrum camera, generating a contact coordinate matrix, and generating a contact-probe mapping table based on the integrated probe array and the contact coordinate matrix, including: Visible light and infrared images of the aircraft plug to be inspected are acquired simultaneously using a dual-spectrum camera. The visible light and infrared images are then fused to obtain a fused image. Key plug features are extracted from the fused image, and then matched with the contact distribution template of a standard plug model to obtain the image matching result. A device coordinate system is established with the integrated probe array reference point as the origin, and a plug local coordinate system is established with the plug geometric center as the origin. Based on the image matching results, the transformation matrix of the two coordinate systems is calculated through a perspective transformation algorithm to determine the three-dimensional relative position of the probe array and the aircraft plug to be inspected. Based on the transformation matrix, each contact point is transformed from the local coordinate system of the plug to the coordinate system of the device. Combined with the sub-pixel level positioning of the actual contact points in the fused image, a contact point coordinate matrix containing the three-dimensional coordinates of each contact point is generated. The three-dimensional coordinates of each probe are determined based on the device coordinate system. Through a spatial coordinate comparison algorithm, the three-dimensional coordinates of each probe in the probe array are matched one by one with the three-dimensional coordinates of each contact in the contact coordinate matrix. A contact-probe mapping table is generated based on the matching results.

3. The helicopter cable fault detection system according to claim 1, characterized in that, The injected test signal identifies the contact material to obtain the plug topology identification result, including: Based on the common material characteristics of the aviation plug to be tested, a suitable test signal is selected from the system's preset signal library, and the test signal is injected into a number of pre-specified probes in the probe array. After the designated probe carrying the test signal comes into contact with the contact of the aviation plug to be tested, the signal acquisition module of the probe array acquires the response signal of each contact to the test signal in real time. Based on the collected response signals, the material type of each contact point is identified through a material feature matching algorithm; Based on the material type of each contact and the signal transmission relationship between the contacts, the contact connection topology of the aviation plug to be tested is constructed, and the contact connection topology is compared with the standard topology template of the corresponding plug model in the system database to obtain the plug topology recognition result.

4. The helicopter cable fault detection system according to claim 1, characterized in that, The step of setting a multimodal detection signal based on the multimodal detection parameters corresponding to the plug topology identification result and the contact-probe mapping table includes: Key information is extracted from the topology identification results, including contact material type, contact function classification, connection relationship between contacts, and standard electrical characteristics of the corresponding plug model. Import the historical response signal characteristics of the corresponding probes for each contact point based on the contact point and probe mapping table, and determine the detection priority and signal adaptation requirements of each contact point based on the probe and historical response signal characteristics. Based on the detection priority and signal adaptation requirements of each contact, corresponding multi-mode detection parameters are matched from the preset parameter library. The multi-mode detection parameters include: continuity detection parameters, insulation detection parameters, and crosstalk detection parameters. The multimodal detection signal is set according to the multimodal detection parameters. The multimodal detection signal includes a bipolar pulse pair corresponding to the on / off detection parameters, a stepped DC signal corresponding to the insulation detection parameters, and a high-frequency swept sine wave corresponding to the crosstalk detection parameters.

5. The helicopter cable fault detection system according to claim 1, characterized in that, The process of locating the positions of each designated contact point using the contact point coordinate matrix as a spatial reference, and applying multimodal detection signals from the probes in the probe array to multiple designated contact points in conjunction with the contact pressure state data to obtain the response signal corresponding to each designated contact point includes: The position of the specified contact point is determined by determining the three-dimensional coordinates of each specified contact point in the device coordinate system based on the contact point coordinate matrix. Based on the position of the specified contact point, the corresponding probe in the probe array is moved to the position of the specified contact point. The contact pressure status data between the probe and the specified contact point is monitored in real time by the force sensor. If the pressure deviates from the effective threshold range, the piezoelectric ceramic micro-motion platform is triggered to fine-tune the angle of the corresponding probe or the memory alloy telescopic rod is adjusted to adjust the length until the contact pressure stabilizes within the effective threshold range. Based on the FPGA time-sharing scheduling strategy, multimodal detection signals are applied one by one to multiple designated contacts through the aligned probes; The signal acquisition module of the probe array captures the response signals of each designated contact to the detection signal in real time, and associates the acquired response signals with the corresponding designated contact based on the contact-probe mapping table to establish the contact-response signal mapping table.

6. The helicopter cable fault detection system according to claim 1, characterized in that, The step of performing contact fault analysis on the response signal to obtain fault results includes: The on / off detection of the response signal is specifically performed as follows: The response signal is a bipolar pulse pair response signal obtained through bipolar pulse pair detection. Time-domain processing is performed on the bipolar pulse pair response signal to obtain the rise time of the bipolar pulse pair response waveform. The bipolar pulse response signal is processed in the frequency domain to obtain the fundamental frequency component energy of the response waveform after FFT transformation. Substituting the rise time and fundamental frequency component energy of the bipolar pulse response waveform into the fuzzy logic membership function, the membership degree of the contact open circuit state is calculated. The fuzzy logic membership function is as follows: ， in, The membership degree of the contact open circuit state, when A value greater than 0.8 indicates an open circuit. The rise time of the bipolar pulse response waveform. The threshold for determining the rise time is... The fundamental frequency component energy, The threshold for determining the fundamental frequency energy. This is an adjustment coefficient used to control the steepness of the membership function.

7. The helicopter cable fault detection system according to claim 1, characterized in that, The step of performing contact fault analysis on the response signal to obtain fault results also includes: Insulation detection is performed on the response signal, specifically as follows: The response signal is a leakage current sequence under stepped voltage obtained by stepped DC signal detection. Where n is the voltage step number, Based on the physical properties of insulation resistance, a nonlinear equation relating leakage current and voltage is constructed: ， in, This refers to the leakage current in the leakage current sequence under stepped voltage. For the applied step voltage, The insulation resistance calculated for the target. This represents the contact resistance between the probe and the contact point. The dielectric polarization effect coefficient; The parameters in the nonlinear relational equation are based on the Levenberg-Marquardt algorithm. , and The fitting process is as follows: Setting parameters , and The initial value; By minimizing the measured leakage current With predicted leakage current Sum of squared errors insulation resistance Perform multiple iterative calculations, when When the iteration stops, output the fitted insulation resistance. , The insulation resistance obtained by fitting If the insulation resistance is less than the insulation resistance threshold, it is determined to be an insulation failure; otherwise, the insulation performance is determined to be normal.

8. The helicopter cable fault detection system according to claim 1, characterized in that, The step of performing contact fault analysis on the response signal to obtain fault results also includes: Crosstalk detection is performed on the response signal, specifically as follows: The response signal is a sweep response amplitude sequence obtained by high-frequency sweep sine wave detection. Where m is the number of sweep points, The frequency of the high-frequency swept sine wave. Based on the frequency characteristics of crosstalk signals, a sparse basis matrix is ​​constructed to characterize the sparse distribution features of the crosstalk coupling coefficients. , Based on the orthogonal matching pursuit (OMP) algorithm and the sparse basis matrix The crosstalk coupling coefficient is calculated from the sweep frequency response amplitude sequence and expressed as: ， in, This represents the sweep frequency response amplitude at the corresponding frequency. For random downsampling matrix, The crosstalk coupling coefficient is... This is the error threshold; Based on crosstalk coupling coefficient By combining the frequency parameters of the high-frequency swept sine wave, the crosstalk coupling coefficient between each pair of contacts is calculated: ， in, The amplitude of the coupling signal between contact i and contact j. For the corresponding frequency, Characteristic impedance; The calculated crosstalk coupling coefficient between the two contacts The crosstalk coupling coefficient is compared with the threshold. If it is greater than the threshold, it is judged as a crosstalk exceeding the standard fault; otherwise, it is judged as a normal crosstalk characteristic.

9. The helicopter cable fault detection system according to any one of claims 1 to 8, characterized in that, The multimodal signal analysis device is further configured to: after obtaining the fault result, perform a health assessment on the aircraft plug to be tested based on the fault result, specifically: The obtained fault results are constructed into a three-dimensional feature vector. , The membership degree for the contact open circuit state. For insulation resistance, This is the crosstalk coupling coefficient matrix. For each contact point, the comprehensive fault probability corresponding to the three-dimensional feature vector is calculated using a weighted summation formula: ， in, , and for , and The corresponding weight, where k is the contact number. The threshold for determining insulation resistance. Determine the threshold for the crosstalk coupling coefficient; Calculate the overall health score of the aviation connector based on the combined failure probability of all contacts: ， Where N is the total number of contacts. The importance weight of touch point k, When S≥80, the patient is considered healthy. When 60 ≤ S < 80, it is considered a mild abnormality. When S < 60, it is judged as a serious fault.

10. A method for detecting faults in helicopter cables, characterized in that, The steps include: The relative position of the aircraft plug to be tested is determined by the image of the aircraft plug to be tested and the contact distribution template acquired by the dual-spectrum camera, a contact coordinate matrix is ​​generated, and a contact-probe mapping table is generated based on the integrated probe array and the contact coordinate matrix. Inject test signals to identify contact materials and obtain plug topology identification results; Contact pressure status data is obtained by detecting the contact pressure at the contact point using a force sensor; Based on the plug topology identification result and the multimodal detection parameters corresponding to the contact and probe mapping table, a multimodal detection signal is set based on the multimodal detection parameters; Using the contact point coordinate matrix as a spatial reference, the position of each designated contact point is located. Combined with the contact pressure state data, the multimodal detection signal of the probe in the probe array is applied to multiple designated contact points to obtain the response signal corresponding to each designated contact point. A contact point and response signal mapping table is established through the mapping relationship between contact points and probes in the contact point and probe mapping table. The response signal is subjected to contact fault analysis to obtain fault results, and a contact-fault result mapping table is generated based on the fault results and the contact-response signal mapping table.

Images