Manned flying saucer cabin door control method and system

Abstract

The invention provides a manned flying saucer cabin door control method and system, and relates to the technical field of aircraft control. The method comprises the steps that acceleration signals of a plurality of positions of a cabin door are obtained through an acceleration sensor array; through decoupling the acceleration signal, estimating to obtain a vibration mode coordinate of the cabin door; generating a control instruction according to the vibration mode coordinate; the sealing pressure and the locking state of the cabin door are monitored, and when the monitoring result is normal, an actuator is controlled to apply control force to the cabin door structure by outputting a control instruction; and when the monitoring result is abnormal, the control instruction is adjusted, and by outputting the adjusted control instruction, the actuator is controlled to apply control force on the cabin door structure or stop outputting the control instruction. According to the method, while vibration is restrained, possible damage to core functions such as sealing and locking is actively avoided, and the safety and reliability of the system in the severe environment are remarkably improved.

Description

Technical Field

[0001] The present invention relates to the technical field of aircraft control, and in particular to a method and system for controlling a cabin door of a manned flying saucer. Background Art

[0002] A manned UFO was carrying out a flight mission at low altitude, relatively low above the ground. Atmospheric flow at this altitude is influenced by factors such as ground topography, objects, and local heat sources, resulting in nonlinear, time-varying turbulence. Turbulence manifests as random fluctuations in airflow velocity and direction, characterized by a distribution of energy spectra across a wide frequency range and weak spatial correlation. These pulsating airflows act on the UFO's outer surface, generating a randomly varying pressure distribution, known as turbulent aerodynamic forces. The UFO's hatch, a key opening in the hull and its covering structure, bears direct weight from these turbulent aerodynamic forces on its outer surface. Due to the weak spatial correlation of turbulence near the ground, the aerodynamic forces acting at different locations on the hatch may be asynchronous and have varying frequency characteristics, resulting in complex and non-uniform dynamic loads on the hatch.

[0003] A door typically consists of an outer skin, internal reinforcements, a frame, seals, a locking mechanism, hinges or slides, and an actuation mechanism. The door panel itself is constructed of composite or metal, providing mass and elasticity. The frame secures the door panel to the opening in the aircraft fuselage. The seal ensures airtightness when the door is closed, and its elastic properties also affect the door's dynamic response. The locking mechanism provides a mechanical lock, ensuring the door remains securely closed during flight. Hinges or slides connect the door to the fuselage, allowing it to open and close. The actuation mechanism operates the door electrically or hydraulically. These components are connected together by connectors and attached to the aircraft fuselage structure. This combined structure exhibits multiple natural frequencies and vibration modes. When the frequency components of the turbulent aerodynamic forces acting on the door are close to the door's natural frequencies, or when the aerodynamic forces vary significantly, resonance or forced vibration of the door can be excited. The heterogeneous nature of near-ground turbulence means that these vibrations can simultaneously excite multiple vibration modes of the door, including bending, torsion, local panel vibrations, and small overall motions relative to the fuselage. The door's mass distribution, structural stiffness, characteristics of its connections, and damping levels all contribute to the potential for these vibrations to have large amplitudes.

[0004] During extended flight in persistent near-ground turbulence, the constant vibration of the hatch door can cause a range of problems. For example, high-frequency vibration can lead to increased friction and wear between the seal and the door frame, compromising the long-term reliability of the seal and even causing a reduction in airtightness. Vibration can also cause changes in the clearance or impact of the locking mechanism, compromising locking reliability and potentially leading to fatigue damage and the risk of locking failure. The continuous stress cycling of the hatch structure accelerates material fatigue and shortens the service life of door components (such as the skin, stiffeners, and connectors). Furthermore, door vibration can generate structure-borne noise that is transmitted into the cabin, impacting the occupant's riding experience. To mitigate these adverse effects, it is necessary to effectively suppress the door's vibration in near-ground turbulence.

[0005] Active vibration control is a technique that senses the vibration state of a structure in real time and applies a reaction force to counteract vibration. Acceleration feedback control is a common approach to achieving active vibration suppression. Accelerometers are placed at key locations on the structure to measure its acceleration signals. The control system calculates control forces based on these signals and drives actuators to apply these forces to reduce vibration. However, in the specific scenario of a complex structure like a manned spacecraft hatch, which faces non-uniform turbulent loads and simultaneously needs to maintain critical functions such as sealing and locking, traditional single-point or simple multi-point vibration control methods are unable to effectively address multimodal coupled vibrations. They also fail to fully consider the potential adverse effects of the vibration control process on the hatch's critical functions, posing the risk of compromising sealing or locking reliability while suppressing vibration. Therefore, how to accurately sense and effectively suppress the multimodal coupled vibrations of a manned spacecraft hatch in a near-Earth turbulent environment, while simultaneously ensuring the safety of the hatch's sealing and locking functions, has become a pressing technical challenge. Summary of the Invention

[0006] The purpose of the present invention is to provide a manned flying saucer hatch control method and system, which accurately suppresses multiple vibration modes through active control, and integrates key function monitoring and linkage with control strategies. While suppressing vibration, it actively avoids possible damage to core functions such as sealing and locking, thereby significantly improving the safety and reliability of the system in harsh environments.

[0007] In a first aspect, the present invention provides a method for controlling a door of a manned flying saucer, comprising the following steps: After arranging acceleration sensors at multiple locations on the hatch door to form an acceleration sensor array, acceleration signals at multiple locations on the hatch door are obtained through the acceleration sensor array; By decoupling the acceleration signal, the vibration modal coordinates of the cabin door are estimated; Generate control instructions based on vibration mode coordinates; Monitor the sealing pressure and locking status of the hatch door. When the monitoring results are normal, the actuator is controlled to exert control force on the hatch door structure by outputting control instructions. When the monitoring results are abnormal, the control instructions are adjusted and the actuator is controlled to exert control force on the hatch door structure by outputting the adjusted control instructions, or the output of control instructions is stopped.

[0008] The manned flying saucer hatch door control method provided by the present invention can accurately capture the multi-point dynamic response of the hatch through distributed sensing in a specific environment where a manned flying saucer encounters random time-varying and spatially non-uniform turbulent loads during near-ground flight, causing multi-modal coupled vibrations of the hatch structure. It can effectively suppress multi-modal vibrations using active control, and can simultaneously monitor the sealing status and locking status of the hatch and adjust the control strategy when necessary to avoid the vibration suppression process from adversely affecting the key functions of the hatch.

[0009] In a second aspect, the present invention provides a manned flying saucer hatch control system, comprising: an acquisition module, configured to arrange acceleration sensors at multiple locations of the hatch to form an acceleration sensor array, and obtain acceleration signals at multiple locations of the hatch through the acceleration sensor array; An estimation module, configured to estimate the vibration modal coordinates of the cabin door by decoupling the acceleration signal; A generation module, used for generating control instructions according to vibration modal coordinates; The control module is used to monitor the sealing pressure and locking status of the hatch. When the monitoring result is normal, the control module outputs a control instruction to control the actuator to exert a control force on the hatch structure. When the monitoring result is abnormal, the control instruction is adjusted and the actuator is controlled to exert a control force on the hatch structure by outputting the adjusted control instruction, or the output of the control instruction is stopped.

[0010] From the above, it can be seen that the manned flying saucer cabin door control method provided by the present invention realizes the effective suppression of multi-modal vibration of the manned flying saucer cabin door in a near-ground turbulent environment through the strategy of distributed sensing, active control and linkage of key functional states. Compared with the scheme that does not consider multi-modality or does not link key function monitoring, the present invention can deal with various vibration forms caused by complex turbulence more comprehensively and accurately, and significantly reduce the vibration amplitude of the cabin door. At the same time, by adjusting the control output when an abnormal sealing or locking state is detected, the vibration suppression process is effectively avoided to cause excessive wear on the seal or adverse impact on the locking mechanism, thereby ensuring the airtightness and locking reliability of the cabin door, extending the service life of the cabin door components, and improving flight safety and occupant comfort.

[0011] Other features and advantages of the present invention will be described in the following description, and in part will become apparent from the description, or will be understood by practicing the embodiments of the present invention. The purposes and other advantages of the present invention can be realized and obtained by the structures particularly pointed out in the written description and the accompanying drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0012] Figure 1 A flow chart of a method for controlling a manned flying saucer hatch provided in an embodiment of the present invention.

[0013] Figure 2 A schematic structural diagram of a manned flying saucer hatch control system provided by an embodiment of the present invention.

[0014] Description of labels: 100, acquisition module; 200, estimation module; 300, generation module; 400, control module. DETAILED DESCRIPTION

[0015] The technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, not all of the embodiments. The components of the embodiments of the present invention generally described and shown in the drawings herein can be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of the present invention provided in the drawings is not intended to limit the scope of the claimed invention, but merely represents selected embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without making creative work are within the scope of protection of the present invention.

[0016] It should be noted that similar reference numerals and letters represent similar items in the following drawings. Therefore, once an item is defined in one drawing, it does not need to be further defined or explained in subsequent drawings. At the same time, in the description of the present invention, the terms "first", "second", etc. are used only to distinguish the description and should not be understood as indicating or implying relative importance.

[0017] Reference Attachment Figure 1 The present invention provides a method for controlling a cabin door of a manned flying saucer, comprising the following steps: After arranging acceleration sensors at multiple locations on the hatch door to form an acceleration sensor array, acceleration signals at multiple locations on the hatch door are obtained through the acceleration sensor array; the locations include the hatch door skin, reinforcement ribs and frame; By decoupling the acceleration signal, the vibration modal coordinates of the cabin door are estimated; Generate control instructions based on vibration mode coordinates; Monitor the sealing pressure and locking status of the hatch door. When the monitoring results are normal, the actuator is controlled to exert control force on the hatch door structure by outputting control instructions. When the monitoring results are abnormal, the control instructions are adjusted and the actuator is controlled to exert control force on the hatch door structure by outputting the adjusted control instructions, or the output of control instructions is stopped.

[0018] An accelerometer array involves installing multiple accelerometers in a specific layout at different locations on the hatch structure to form a sensor network. This can be achieved using piezoelectric accelerometers, microelectromechanical system (MEMS) accelerometers, or fiber optic accelerometers. For example, sensors can be placed on the hatch skin, stiffeners, and frame. This is primarily to obtain dynamic response information at multiple points on the hatch, providing a data foundation for subsequent vibration state analysis. Accelerometer signal decoupling involves using signal processing techniques to separate the coupled acceleration signals obtained from the accelerometer array into components representing the hatch's individual vibration modes. This can be achieved using modal analysis, state-space methods, or frequency-domain decomposition methods. For example, decoupling based on a structural dynamics model or a data-driven model aims to identify and quantify the contributions of each order of the hatch's vibration mode at the current moment, decomposing the complex overall vibration into modal vibrations that can be independently analyzed and controlled. Vibration modal coordinates are parameters derived after decoupling that characterize the vibration state of each mode of the hatch door. These typically include modal displacement, modal velocity, or modal acceleration. They are primarily used to describe the vibration amplitude or phase of the hatch door in a specific mode, providing a basis for generating targeted control commands. Control commands are signals calculated based on the hatch door's vibration modal coordinates to drive the actuator to generate control force. These can be implemented as voltage, current, or digital signals, primarily instructing the actuator how much force to apply and what type of force to offset door vibration. Monitoring the hatch door's sealing pressure and locking status involves using sensors or status detection devices to obtain real-time information on the door's airtightness (indicated by pressure changes) and mechanical locking status after closing. This can be achieved using pressure sensors, proximity switches, or strain sensors. For example, a pressure sensor can be placed near the sealing strip and a status switch can be placed near the lock tongue. This primarily provides status information on key door safety functions, serving as a basis for adjusting vibration control strategies. Adjusting the control instructions means modifying the original control instructions when an abnormality in the hatch sealing or locking status is detected, so as to change the control force output of the actuator. This can be achieved by reducing the instruction amplitude, changing the instruction frequency, or completely stopping the instruction output. Its main purpose is to give priority to protecting the key functions of the hatch when the safety status is abnormal, and to avoid adverse effects caused by the control force.

[0019] The core innovation of this application lies in that by combining vibration state estimation based on multi-point sensing and modal decoupling with real-time monitoring of the key safety status of the cabin door (sealing and locking), it achieves effective suppression of the multi-modal vibration of the manned flying saucer cabin door. At the same time, it can adaptively adjust or stop the control output when the safety status is abnormal, achieving the effect of suppressing vibration while ensuring the reliability of the cabin door sealing and locking.

[0020] Specifically, the method first uses an array of accelerometers placed at multiple locations on the door, including the skin, stiffeners, and frame, to acquire acceleration signals at these locations. These signals contain rich information about the door's overall and local vibrations. These multi-point acceleration signals are then decoupled, breaking down the coupled acceleration response into independent responses of each vibration mode. This allows the system to estimate the door's modal coordinates, enabling it to identify the dominant modes and their amplitudes. Based on the estimated modal coordinates, control commands are generated to drive the actuators to apply control forces. These commands are based on an understanding of the door's modal vibration states and are designed to offset or reduce these modal vibrations by applying reaction forces. Simultaneously, the system continuously monitors the door's sealing pressure and locking status to determine whether key safety functions are functioning properly. If the monitoring results indicate normal sealing and locking, the system outputs control commands based on the modal information, directing the actuators to apply control forces to the door structure for standard vibration suppression. However, if monitoring results indicate a drop in seal pressure or an abnormal locking state, the system adjusts control commands accordingly, such as reducing or halting control force output to prevent further interference or damage to the already abnormal seal or locking mechanism. This entire process forms a closed-loop control system that, through the synergy of perception, analysis, control, and safety monitoring, suppresses cabin door vibration and ensures the safety of its critical functions.

[0021] As a preferred embodiment, the solution of this application is implemented as follows: MEMS accelerometers are installed at key locations on the hatch skin, stiffeners, and frame to form an accelerometer array. The acquired acceleration signals are input into an embedded processor. The processor executes a signal processing algorithm, such as a modal decomposition algorithm based on a structural dynamics model, to decouple the acceleration signals and estimate the modal coordinates of the first few major vibration modes of the hatch. Based on these modal coordinates, the processor runs a control algorithm, such as a linear quadratic regulator (LQR) algorithm, to calculate the control force required to be applied to the hatch and convert it into a control command to drive a piezoelectric or hydraulic actuator. Simultaneously, a pressure sensor is installed near the hatch seal and a proximity switch is installed at the lock tongue to monitor the hatch seal pressure and locking status in real time. The monitoring signals are also input into the processor. If the monitored sealing pressure falls below a threshold or the locking switch status is abnormal, the processor will adjust the control command according to a preset safety policy, such as multiplying the amplitude of the control command by a factor less than 1 or directly setting the control command to zero. The adjusted control command is then output to the actuator.

[0022] Through the above-mentioned scheme, this application, through multi-point sensing and modal decoupling, can effectively identify and quantify the multi-modal coupled vibration state of the manned UFO hatch under complex loads, thereby generating targeted control instructions to suppress the hatch vibration. At the same time, the introduction of real-time monitoring of the hatch sealing pressure and locking status, and the adjustment or cessation of control output when the monitoring results are abnormal, effectively avoids the potential adverse effects of the vibration control process on the key safety functions of the hatch, and improves the safety and reliability of the hatch in the near-ground turbulent environment.

[0023] In some embodiments, the step of estimating the vibration modal coordinates of the cabin door by decoupling the acceleration signal includes: According to the position distribution of the acceleration sensor and the acceleration signal, an acceleration signal matrix is constructed; Performing fast Fourier transform on each acceleration signal in the acceleration signal matrix to obtain a frequency domain acceleration signal matrix; Perform short-time Fourier transform analysis based on a sliding time window on the frequency domain acceleration signal matrix, calculate the power spectrum density in each time window, and obtain the power spectrum density matrix; According to the power spectrum density matrix, identify the main frequency components in the current time window; The identified main frequency components are matched with the pre-established cabin door modal frequency database to determine the main vibration mode of the cabin door at the current moment and obtain the main vibration mode information; Based on the determined main vibration mode information, the acceleration signal matrix is decoupled and the vibration mode coordinates of the cabin door are estimated.

[0024] Constructing an acceleration signal matrix involves structured processing of multi-channel acceleration signals collected from different locations on the hatch, forming a data set that facilitates unified analysis and processing. Performing a fast Fourier transform on the acceleration signals converts the time-domain signal into a frequency-domain representation, revealing the signal's frequency composition. Short-time Fourier transform analysis based on a sliding time window is a time-frequency analysis technique that analyzes how the signal's frequency components vary over time by sliding a finite time window over the signal and performing a Fourier transform within each window. This technique is particularly important for analyzing non-stationary or transient signals. Calculating the power spectral density quantifies the signal's energy distribution at different frequencies. The power spectral density matrix contains energy information at different sensor locations, at different times, and at different frequencies. Identifying the primary frequency components involves extracting frequency points with significant energy from the power spectral density matrix. These frequency points typically correspond to the structure's natural frequencies or forced vibration frequencies. Matching the identified primary frequency components with a pre-established hatch modal frequency database correlates the real-time monitored frequency information with the structure's inherent dynamic characteristics, thereby determining which vibration modes are primarily exciting the hatch at the current moment. A pre-established database of hatch door modal frequencies, obtained through finite element analysis or modal experiments, contains frequency information for each of the hatch door's natural modes. Decoupling the acceleration signal matrix based on the identified primary vibration modal information uses this identified modal information as prior knowledge or guidance to decompose the coupled sensor acceleration signals into their individual modes. This allows for an estimate of the time-varying amplitude of each mode, i.e., the modal coordinates.

[0025] The reason why the present application can accurately and robustly estimate the vibration modal coordinates of the cabin door is that it adopts a multi-stage, multi-domain combined signal processing and modal identification method. First, multi-point acceleration signals are acquired through an acceleration sensor array, and an acceleration signal matrix is constructed, which is the basic data for structural dynamics analysis. Subsequently, the overall frequency characteristics of the signal are preliminarily analyzed by fast Fourier transform. Furthermore, the short-time Fourier transform and power spectral density analysis based on a sliding time window are used to capture the change of the cabin door vibration frequency over time, which is crucial for identifying time-varying or transient excited modes caused by turbulence. By identifying the main frequency components in the time-frequency domain, the currently active vibration mode can be located more accurately. By matching these real-time identified frequencies with a pre-established modal frequency database, the abstract frequency information can be associated with the specific structural vibration mode, thereby determining the main vibration mode of the current cabin door. It is precisely because of the ability to accurately identify the current dominant vibration mode that the subsequent decoupling process based on this modal information can more effectively decompose the coupled sensor signals into individual active modes, thereby improving the accuracy and robustness of modal coordinate estimation in complex load and noise environments. This strategy, which combines time-domain, frequency-domain, and time-frequency-domain analysis with modal identification, fully utilizes the rich information contained in the acceleration signal, overcomes the limitations of simple decoupling methods when dealing with complex multi-modal coupled vibrations, and provides reliable modal state perception for subsequent precise vibration control.

[0026] Through the above method, the present application can effectively extract the multi-modal coupled vibration state information of the cabin door in a near-ground turbulent environment from the complex acceleration signal. This method combines the advantages of time domain, frequency domain and time-frequency domain analysis, and introduces modal identification and matching links, so that under unfavorable conditions such as non-uniform dynamic loads, noise and modal density, it can still accurately and robustly estimate the main vibration modes of the cabin door and their corresponding modal coordinates. As a result, it provides a reliable input for the subsequent generation of precise control instructions based on modal coordinates, thereby improving the vibration control effect of the manned UFO cabin door in a near-ground turbulent environment and helping to ensure the sealing and locking reliability of the cabin door.

[0027] In some embodiments, the step of decoupling the acceleration signal matrix based on the determined primary vibration mode information to estimate the vibration mode coordinates of the cabin door includes: Subtracting the modal acceleration signal reconstructed based on the determined main vibration modal information from the acceleration signal matrix to obtain a residual acceleration signal matrix to represent the vibration information of the unidentified mode; According to the energy of the residual acceleration signal matrix, the decoupling parameters are adjusted to reduce the impact of modal identification error on modal decoupling; The acceleration signal matrix is weightedly decoupled using the adjusted decoupling parameters to obtain the modal coordinates of each vibration mode. The weight is determined according to the recognition confidence of the corresponding mode.

[0028] The residual acceleration signal matrix refers to the signal matrix remaining after removing the identified main modal contributions from the original acceleration signal, and represents the vibration information that has not been fully identified or identified inaccurately. The energy of the residual acceleration signal matrix refers to the sum of the energies or average energy of each signal in the residual acceleration signal matrix. It is an indicator to measure the intensity of unidentified or inaccurately identified vibration information. It can be achieved by calculating the mean square value, sum of squares or integral square of the signal. The decoupling parameters refer to the set of parameters used to separate the mixed acceleration signal into each modal coordinate. For example, it can be the inverse or pseudo-inverse of the modal matrix, state space model parameters, etc. It can be achieved by using a modal matrix constructed based on the results of structural modal analysis, or parameters obtained through system identification methods. Weighted decoupling refers to assigning different weights to each mode according to the identification confidence of the mode during the decoupling process to adjust the degree of influence of each mode on the final modal coordinate estimation result. The weight is determined according to the recognition confidence of the corresponding mode, which refers to the weight value used in weighted decoupling. Its size is positively correlated with the recognition confidence of the corresponding mode. The recognition confidence can be calculated based on multiple indicators such as the signal-to-noise ratio, modal shape matching, and frequency matching.

[0029] Based on the above technical features, the modal decoupling method of the present application works as follows: After identifying the main vibration modal information, the signal reconstructed based on this identified modal information is first subtracted from the original acceleration signal matrix to separate the residual acceleration signal matrix. The purpose of this step is to explicitly extract the parts of the original signal that were not fully identified or were identified inaccurately. This residual information reflects the shortcomings of the modal identification process and provides a basis for subsequent evaluation and improvement of the decoupling process. Next, by calculating the energy of the residual acceleration signal matrix, the degree of modal identification error or omission can be quantified. The magnitude of the residual energy directly indicates the intensity of the vibration information that was not fully captured. Then, based on the magnitude of the residual energy, the parameters used for the decoupling calculation are dynamically adjusted. This adjustment mechanism enables the decoupling process to adapt to the uncertainty of modal identification. For example, when the residual energy is large, the decoupling parameters can be adjusted to increase the robustness to unmodeled dynamics and avoid large deviations in modal coordinate estimation due to inaccurate identification information. Finally, when decoupling the acceleration signal matrix using the adjusted decoupling parameters, a weighted processing based on the confidence level of each modal identification is introduced. Modes with high identification confidence are assigned higher weights to their corresponding decoupling results, making a greater contribution to the final modal coordinate estimation; modes with low identification confidence are assigned lower weights, and their interference with the estimation results is suppressed. The entire process forms an adaptive and weighted decoupling mechanism. Based on preliminary modal identification and decoupling, the accuracy of modal coordinate estimation in the presence of modal identification uncertainty is significantly improved by analyzing residuals, adjusting parameters, and introducing confidence weighting. This optimization is particularly important in specific scenarios where complex turbulent loads make modal identification difficult. It can more accurately perceive the true vibration state of the cabin door and provide a reliable basis for subsequent precise control.

[0030] By adopting the above technical solution, the present application can achieve the following technical effects: by separating the residual information from the original signal, the uncertainty of modal identification can be quantified. The decoupling parameters are adaptively adjusted according to the residual energy to enhance the robustness of the decoupling process to unidentified or inaccurately identified modes. The introduction of weighting based on recognition confidence makes the high-confidence modal estimation more reliable and the influence of low-confidence modals is weakened. These measures work together to significantly improve the accuracy and reliability of the estimation of each vibration modal coordinate in the presence of errors in modal identification. More accurate modal coordinate estimation can provide more precise state information for subsequent vibration control, thereby improving the overall effect of active vibration control and effectively suppressing the multi-modal coupled vibration of the cabin door under complex turbulent loads.

[0031] In some embodiments, the step of performing weighted decoupling on the acceleration signal matrix using the adjusted decoupling parameters to obtain the modal coordinates of each vibration mode includes: Obtain the peak value of power spectrum density of each mode; The recognition confidence of each mode is calculated based on the ratio of the power spectrum density peak of each mode to the preset noise threshold, and the modal confidence matrix is constructed according to the arrangement of the acceleration signal matrix. The higher the ratio, the higher the confidence. According to the modal confidence matrix, a normalized exponential function is used to calculate the weighted coefficients of the corresponding modes of each accelerometer. A weighted decoupling matrix is constructed to highlight the contribution of high-confidence modes to the modal coordinate estimation and suppress the interference of low-confidence modes. The weighted coefficients are proportional to the confidence level. Using the adjusted decoupling parameters and weighted decoupling matrix, the acceleration signal matrix is weighted decoupled to obtain the modal coordinates of each vibration mode. The calculation formula is: modal coordinates = weighted decoupling matrix * adjusted decoupling parameters * acceleration signal matrix.

[0032] The power spectral density peak refers to the maximum energy value corresponding to a specific modal frequency in the signal's power spectral density plot. It can be calculated by performing a Fourier transform on the signal. The preset noise threshold refers to a pre-set energy level used to distinguish signal energy from background noise. It can be determined based on the sensor noise characteristics or the ambient noise level. Recognition confidence is a quantitative assessment of the reliability of the recognition results of a specific modality. It can be calculated using the ratio of the power spectral density peak to the noise threshold or other statistical methods. The modal confidence matrix is a matrix that arranges the recognition confidence of each modality according to the structure of the acceleration signal matrix. Its dimensions are related to the acceleration signal matrix. The normalized exponential function is a mathematical function used to map the recognition confidence to weighting coefficients and ensure that these coefficients are normalized. This can be achieved by taking the exponential form of e and dividing by the sum. The weighting coefficient is the weight assigned to different modes or different sensor signals in the decoupling calculation. Its size reflects the reliability of the corresponding modal recognition. A weighted decoupling matrix is a matrix composed of calculated weighting coefficients, used to weight the acceleration signal matrix during the decoupling process. Weighted decoupling is the process of adjusting the input signal or decoupling parameters according to preset weights during modal decoupling calculations.

[0033] This solution addresses the issue of improving estimation accuracy when decoupling the acceleration signal matrix using adjusted decoupling parameters to estimate modal coordinates. A weighted decoupling method based on modal identification confidence is proposed. First, the peak power spectral density of each mode is obtained, providing basic data for subsequent quantification of modal identification reliability. The peak power spectral density reflects the energy intensity of a specific mode in the signal; generally, the stronger the mode, the easier it is to accurately identify. Next, the identification confidence of each mode is calculated based on the ratio of the peak power spectral density of each mode to a preset noise threshold, and a modal confidence matrix is constructed. This step quantifies the reliability of modal identification. By comparing it with the noise threshold, modes with signal energy exceeding the noise level can be distinguished. A higher ratio indicates a clearer modal signal, less noise interference, and a more reliable identification result. Constructing the modal confidence matrix maps this reliability information to the structure of the acceleration signal matrix, paving the way for subsequent weighted processing. Then, based on the modal confidence matrix, a normalized exponential function is used to calculate the weighting coefficients for each accelerometer mode, constructing a weighted decoupling matrix. This is a key step in achieving weighted decoupling. By using the identification confidence to calculate the weighting coefficients and employing a normalized exponential function, modes with high identification confidence receive larger weights, contributing more to the estimation of the modal coordinates during the decoupling process, while modes with low identification confidence receive smaller weights, suppressing their interference. The design of weighting coefficients proportional to the confidence ensures that the more reliably identified modes receive higher weights in the decoupling calculation. The weighted decoupling matrix organizes these weights and applies them to the decoupling process of the entire acceleration signal matrix. Finally, using the adjusted decoupling parameters and the weighted decoupling matrix, the acceleration signal matrix is weightedly decoupled to obtain the modal coordinates of each vibration mode. By applying the previously calculated weighted decoupling matrix to the decoupling calculation, weighted processing based on modal identification confidence is achieved. The calculation formula "modal coordinates = weighted decoupling matrix * adjusted decoupling parameters * acceleration signal matrix" clarifies the specific mathematical implementation of this weighted decoupling. This weighted decoupling method can more effectively utilize the differences in the reliability of different modal identifications, making the estimation of modes with high identification confidence more accurate, while reducing the negative impact of modes with low identification confidence on the overall estimation, thereby improving the estimation accuracy of each vibration modal coordinate. This scheme is combined with the step of adjusting the decoupling parameters based on the energy of the residual acceleration signal matrix. The former focuses on improving the accuracy of the estimation of identified modal coordinates, while the latter focuses on reducing the impact of unidentified modes on the decoupling. The two work synergistically to improve the overall accuracy and robustness of the modal coordinate estimation.

[0034] In a specific embodiment, the power spectral density peaks corresponding to the identified primary modal frequencies can be first extracted from the power spectral density matrix. For example, if three primary modes are identified, the power spectral density peaks of these three modes at each sensor location are obtained. Next, a preset noise threshold is set, such as the noise equivalent power spectral density level of an acceleration signal determined from the sensor manual or experimental data. The ratio of the power spectral density peak of each mode at each sensor location to the noise threshold is then calculated. This ratio can serve as a preliminary indicator of identification confidence. These ratios are arranged according to the structure of the acceleration signal matrix (e.g., rows represent sensors, and columns represent time or frequency points) to construct a modal confidence matrix. For example, if the acceleration signal matrix has M rows and N columns, the modal confidence matrix can also have M rows and N columns, where each element represents the identification confidence for the corresponding sensor location and mode (if multimodality is considered). Subsequently, a normalized exponential function is used to calculate the corresponding weighting coefficient for each confidence value in the modal confidence matrix. For example, the function form w_ij = exp(k*confidence_ij) / sum(exp(k*confidence_ij)) can be used; where k is an adjustable parameter that controls the steepness of the weight change with confidence, confidence_ij is an element in the modal confidence matrix, and sum is the sum of all relevant weights for normalization. The calculated weighting coefficients are constructed into a weighted decoupling matrix. The structure of this matrix should match the requirements of the subsequent decoupling calculations, for example, it can be a diagonal matrix or a more complex structure, depending on the decoupling algorithm. Finally, the decoupling parameter matrix obtained in the previous step based on the energy of the residual acceleration signal matrix is multiplied by the constructed weighted decoupling matrix, and then multiplied by the acceleration signal matrix to obtain the modal coordinates of each vibration mode. This calculation process can be performed by a digital signal processor or embedded system.

[0035] By obtaining the peak power spectral density of each modal and comparing it with a preset noise threshold, the recognition reliability of each mode can be quantified. Based on this quantified recognition confidence, a normalized exponential function is used to calculate weighting coefficients and construct a weighted decoupling matrix. This gives modes with high recognition confidence greater weight in the decoupling calculation, while effectively suppressing the interference of modes with low recognition confidence. Using the adjusted decoupling parameters and the constructed weighted decoupling matrix to perform weighted decoupling on the acceleration signal matrix, the modal coordinates of each vibration mode can be more accurately estimated, thereby improving the accuracy and reliability of modal coordinate estimation.

[0036] In some embodiments, the step of generating a control instruction according to the vibration modal coordinates includes: Based on the finite element model of the hatch structure, the control force distribution matrix is constructed by determining the transfer function between the actuator position and the door's modal vibration. This matrix represents the control force influence coefficient of each actuator on each mode. The control force required for each mode is calculated based on the control force distribution matrix and vibration mode coordinates. During the calculation process, a control force penalty term is added to reduce the control force output of specific modes that are prone to causing hatch seal failure or loosening. According to the control force required by each mode, combined with the pseudo-inverse matrix of the control force distribution matrix, the control force that each actuator should output is calculated, and the control force is limited to prevent actuator saturation; The calculated control forces are converted into control commands for driving the actuators.

[0037] The control force distribution matrix is a mathematical matrix that characterizes the relationship between the physical forces applied by multiple actuators and the various vibration modal responses of the hatch. It can be constructed using analytical methods based on structural dynamics theory or numerical methods based on finite element analysis results. The transfer function is a mathematical model that describes the dynamic relationship between the system input (actuator force) and output (modal vibration response). It can be determined using methods such as frequency domain analysis, time domain identification, or modal synthesis. The control force penalty term is an additional term introduced when calculating the required control force for a specific mode to reduce the control objective or weight for that mode. This can be achieved using weighting coefficients, penalty factors in cost functions, or objective function modifications. The pseudo-inverse matrix is the generalized inverse matrix of a non-square or singular matrix. It is used to solve the least-squares solution to a system of linear equations. It can be calculated using methods such as singular value decomposition (SVD), QR decomposition, or iterative algorithms. Limiting is the process of limiting the calculated control force to within a preset maximum and minimum range. This can be achieved using methods such as saturation functions, truncation functions, or dead-band functions.

[0038] This solution specifies how to generate control force commands for actuators based on the estimated cabin door vibration modal coordinates. First, based on a deep understanding of the cabin door structure, a mathematical relationship between the physical forces applied by the actuators and the responses of the cabin door's various vibration modes is established through finite element modeling and transfer function analysis, thereby constructing a control force allocation matrix. This matrix clearly characterizes the control force influence coefficient of each actuator on each mode, providing the basis for subsequently translating control requirements in the modal space into actuator outputs in the physical space. This is key to achieving precise multi-actuator control of multiple modes. Second, based on the currently sensed cabin door vibration state (vibration modal coordinates) and the actuator's ability to influence each mode (control force allocation matrix), the theoretical control force required to effectively suppress vibration in each mode is calculated. Furthermore, during the calculation process, a control force penalty term is added to reduce the control force output for specific modes that are prone to causing cabin door seal failure or loosening. This feature identifies specific vibration modes with potential negative impacts on the hatch's critical functions. By adding a penalty term, the control force applied to these modes is intentionally weakened, preventing excessive control from causing unnecessary stress or wear on seals or locking mechanisms. This approach effectively suppresses vibration while ensuring the safety of the hatch's critical functions. Next, after determining the control force required for each mode, the pseudo-inverse of the control force allocation matrix is used to rationally distribute the control force requirements across the modal space to the actuators. The use of the pseudo-inverse matrix helps address mismatches between the number of actuators and the number of modes. After calculating the actuator's required output force, limiting is essential to ensure that the calculated control force remains within the actuator's safe operating range, improving the robustness and reliability of the control system. Finally, the calculated physical control force values are converted into control commands that the actuators can recognize and execute, ensuring that the control system can effectively drive the actuators to apply the required control force to the hatch structure, thereby suppressing hatch vibration. Through the synergistic effect of the above steps, this solution can generate control instructions that can effectively suppress cabin door vibration while taking into account the cabin door sealing and locking safety under complex load and actuator restricted conditions, solving the technical problem of difficulty in generating control instructions that can take into account both cabin door vibration suppression and cabin door safety.

[0039] In a specific embodiment, a precise three-dimensional structural model of the hatch door can be first established using commercial finite element analysis software (such as ANSYS or NASTRAN). Modal and frequency response analyses are then performed to determine the hatch door's natural frequencies, modal shapes, and the responses of each mode when a unit force is applied at the actuator location. Based on these analysis results, a control force allocation matrix can be constructed, with each column corresponding to a mode and each row corresponding to an actuator. The matrix elements represent the control influence coefficient of each actuator on that mode. To calculate the control force required for each mode, modern control theory methods such as linear quadratic regulator (LQR) or H∞ control can be employed, using the vibration modal coordinates as state inputs to calculate the optimal modal control force. For specific modes identified as prone to sealing or locking issues (such as certain low-order bending or torsional modes), a penalty weight for these modal control forces can be added to the LQR cost function, or a damping factor less than 1 can be multiplied by the calculated modal control force. Subsequently, the pseudo-inverse of the control force distribution matrix is calculated using functions provided by a numerical computing library (such as MATLAB or SciPy). The calculated modal control force vector is multiplied on the left by this pseudo-inverse matrix to obtain the physical control force vector that each actuator should output. The calculated actuator control force is saturated and limited to ensure that it does not exceed the maximum output capacity of the actuator. Finally, the limited control force value is converted into a voltage or current signal through a digital-to-analog converter (DAC) or a specific actuator drive interface, and output as the control command to drive the actuator.

[0040] Through the above-mentioned technical solution, this application is able to generate targeted control instructions based on precise structural models and modal information. In particular, by introducing penalties for specific modes in the control force calculation and limiting the actuator output, this solution can effectively suppress cabin door vibration while reducing potential adverse effects on the cabin door seals and locking mechanism, thereby improving the safety and reliability of the cabin door in complex turbulent environments.

[0041] In certain embodiments, when the monitoring result is abnormal, the control instruction is adjusted and the actuator is controlled to exert a control force on the door structure by outputting the adjusted control instruction, or the output of the control instruction is stopped. Obtain the control force output amplitude and frequency information of the current actuator; The control force adjustment factor is calculated based on the control force output amplitude and frequency information. The calculation formula is: adjustment factor = 1-(amplitude / maximum amplitude)*(frequency / maximum frequency), where the maximum amplitude and maximum frequency are the preset actuator safe operation thresholds; Based on the control force adjustment factor, the original control instruction is corrected to obtain the adjusted control instruction. The correction method is: adjusted control instruction = original control instruction * adjustment factor; Determine whether the adjusted control command amplitude is lower than the minimum control command threshold. If so, stop outputting the control command. If not, send the adjusted control command to the actuator to control the actuator to apply control force on the door structure.

[0042] The control force output amplitude refers to the magnitude of the actual control force currently being output by the actuator. This can be obtained by measuring and analyzing the actuator drive signal, or by acquiring force or current signals from internal sensors within the actuator. The control force output frequency refers to the dynamic rate of change of the actual control force currently being output by the actuator. This can be obtained by performing spectral analysis on the actuator drive signal or force signal. The maximum amplitude and maximum frequency represent the upper limits of the actuator's output control force while ensuring the structural and functional safety of the door. These can be determined through experimental testing or simulation analysis based on the door's structural characteristics, the load-bearing capacity of the seals and locking mechanisms, and flight safety requirements. The control force adjustment factor is a proportional coefficient used to correct the original control command amplitude. Its value typically ranges from 0 to 1 and indicates the degree to which the control force needs to be reduced. The minimum control command threshold is a preset lower limit below which the control command is deemed insufficient to produce effective control action or below which continued application may pose a risk. This value is determined based on the actuator's minimum effective output capability, the response characteristics of the door structure, and the safety margin.

[0043] This solution addresses the issue of how to safely adjust control commands or halt control when abnormalities in hatch door sealing pressure or locking status are detected. It provides a specific adaptive adjustment strategy based on the actuator's current output state. When an abnormality is detected, the system first obtains the actual actuator output control force amplitude and frequency information. This information is used to understand the magnitude and dynamic characteristics of the currently applied control force, which is crucial for assessing the potential impact of the control force on the already abnormal hatch door state. For example, excessive control force amplitude or a frequency close to the hatch door's structural characteristics may exacerbate seal wear or locking mechanism fatigue. Next, a control force adjustment factor is calculated based on the acquired control force output amplitude and frequency information. The calculation formula is: Adjustment Factor = 1 - (Amplitude / Maximum Amplitude) * (Frequency / Maximum Frequency). The maximum amplitude and maximum frequency represent preset actuator safe operating thresholds. This calculation process embodies an adaptive adjustment logic based on the "criticality" of the current control force. When the actuator's output amplitude or frequency increases and approaches the preset safety threshold, the calculated adjustment factor decreases, indicating a higher potential risk of the currently applied control force posing an abnormal condition and requiring a more significant reduction. Conversely, when the amplitude and frequency are lower, the adjustment factor approaches 1, indicating a smaller reduction. This allows the system to dynamically determine the degree of control command adjustment based on the actuator's actual operating state, avoiding blind adjustments. The original control command is then modified based on the calculated control force adjustment factor to produce the adjusted control command. The correction method is: Adjusted control command = Original control command * Adjustment factor. Since the adjustment factor is less than or equal to 1, this correction method directly reduces the amplitude of the original control command proportionally. Based on the adjustment factor calculated in the previous step, high-risk control commands are significantly reduced, while low-risk commands are less reduced. This reduces the control force applied to the door, mitigates the risk of an abnormal condition, helps prevent further deterioration, and ensures safe flight. Finally, the system determines whether the adjusted control command amplitude is below the preset minimum control command threshold. If the force falls below this threshold, the corrected control force is already very weak. Continued application may be ineffective and poses a risk. In this case, the system stops outputting control commands, a safety strategy for extreme abnormal situations. If the force remains above this threshold, the adjusted control command is sent to the actuator, instructing it to apply a weakened but still effective control force on the door structure. This judgment and decision-making process ensures that, in abnormal situations, the system can switch between continuing to apply limited control force to minimize vibration and completely stopping control to maximize safety, based on the magnitude of the corrected control force. This improves the system's safety and robustness, and avoids the potential for uncontrolled vibration caused by simply stopping control.By combining the original control instructions generated based on modal coordinates with the adjustment factors calculated based on the current state of the actuator, this solution can intelligently adjust or stop control based on abnormal conditions and potential risks of actuator output while maintaining a certain level of vibration suppression capability, thereby achieving more reliable hatch control while ensuring the safety of hatch sealing and locking.

[0044] Through the above-mentioned technical means, the present application can adaptively adjust the control instructions or safely stop the control based on the current control force output state of the actuator when the cabin door sealing pressure or locking state is detected to be abnormal. This method avoids simple and crude instruction adjustment or stopping, and reduces the risk of increased cabin door vibration or secondary damage to the abnormal state due to improper control. The system can dynamically weaken the control force according to the potential risk level of the actuator output and stop safely when the control force is too low, thereby improving the control safety in abnormal situations, better ensuring the sealing and locking functions of the manned aircraft cabin door, and meeting the safety flight requirements.

[0045] Reference Attachment Figure 2 The present invention provides a manned flying saucer hatch control system, comprising: An acquisition module 100 is configured to arrange acceleration sensors at multiple locations on the door to form an acceleration sensor array, and then obtain acceleration signals at multiple locations on the door through the acceleration sensor array; An estimation module 200 is configured to estimate the vibration modal coordinates of the cabin door by decoupling the acceleration signal; A generating module 300 is used to generate a control instruction according to the vibration modal coordinates; The control module 400 is used to monitor the sealing pressure and locking status of the hatch. When the monitoring result is normal, the control module 400 controls the actuator to exert control force on the hatch structure by outputting control instructions. When the monitoring result is abnormal, the control instruction is adjusted and the actuator is controlled to exert control force on the hatch structure by outputting the adjusted control instruction, or the output of the control instruction is stopped.

[0046] In this document, relational terms such as first and second, etc. are used merely to distinguish one entity or operation from another entity or operation, but do not necessarily require or imply any actual relationship or order between these entities or operations.

[0047] The foregoing description is merely an embodiment of the present invention and is not intended to limit the scope of protection of the present invention. Those skilled in the art will readily appreciate that the present invention is susceptible to various modifications and variations. Any modifications, equivalent substitutions, or improvements made within the spirit and principles of the present invention shall be included within the scope of protection of the present invention.

Claims

1. A method for controlling a manned flying saucer hatch, characterized in that: The following steps are involved: After arranging acceleration sensors at multiple locations on the hatch door to form an acceleration sensor array, acceleration signals at multiple locations on the hatch door are obtained through the acceleration sensor array; By decoupling the acceleration signal, the vibration modal coordinates of the cabin door are estimated; Generate control instructions based on vibration mode coordinates; Monitor the sealing pressure and locking status of the hatch door. When the monitoring results are normal, the actuator is controlled to exert control force on the hatch door structure by outputting control instructions. When the monitoring results are abnormal, the control instructions are adjusted and the actuator is controlled to exert control force on the hatch door structure by outputting the adjusted control instructions, or the output of control instructions is stopped.

2. The method for controlling a manned flying saucer door according to claim 1, characterized in that: The acceleration sensors are arranged in the skin, stiffeners and frame of the door.

3. The method for controlling a manned flying saucer door according to claim 1, characterized in that: The steps of estimating the vibration modal coordinates of the cabin door by decoupling the acceleration signal include: According to the position distribution of the acceleration sensor and the acceleration signal, an acceleration signal matrix is constructed; Performing fast Fourier transform on each acceleration signal in the acceleration signal matrix to obtain a frequency domain acceleration signal matrix; Perform a sliding time window analysis on the frequency domain acceleration signal matrix, calculate the power spectrum density in each time window, and obtain the power spectrum density matrix; According to the power spectrum density matrix, identify the main frequency components in the current time window; The identified main frequency components are matched with the pre-established cabin door modal frequency database to determine the main vibration mode of the cabin door at the current moment and obtain the main vibration mode information; Based on the determined main vibration mode information, the acceleration signal matrix is decoupled and the vibration mode coordinates of the cabin door are estimated.

4. The method for controlling a manned flying saucer door according to claim 3, characterized in that: Based on the determined main vibration mode information, the acceleration signal matrix is decoupled and the vibration mode coordinates of the cabin door are estimated. The steps include: Subtracting the modal acceleration signal reconstructed based on the determined main vibration mode information from the acceleration signal matrix to obtain a residual acceleration signal matrix; Adjust the decoupling parameters according to the energy of the residual acceleration signal matrix; The acceleration signal matrix is weighted decoupled using the adjusted decoupling parameters to obtain the modal coordinates of each vibration mode.

5. The method for controlling a manned flying saucer door according to claim 4, characterized in that: The steps of performing weighted decoupling on the acceleration signal matrix using the adjusted decoupling parameters to obtain the modal coordinates of each vibration mode include: Obtain the peak value of power spectrum density of each mode; The recognition confidence of each mode is calculated based on the ratio of the power spectrum density peak of each mode to the preset noise threshold, and the modal confidence matrix is constructed according to the arrangement of the acceleration signal matrix; According to the modal confidence matrix, the weighted coefficients of the corresponding modes of each acceleration sensor are calculated to construct a weighted decoupling matrix; The acceleration signal matrix is weighted decoupled using the adjusted decoupling parameters and weighted decoupling matrix to obtain the modal coordinates of each vibration mode.

6. The method for controlling a manned flying saucer door according to claim 1, characterized in that: The steps of generating control instructions according to the vibration modal coordinates include: By determining the transfer function between the actuator position and each modal vibration of the hatch, the control force distribution matrix is constructed; Calculate the control force required for each mode based on the control force distribution matrix and vibration mode coordinates; According to the control force required by each mode, combined with the pseudo-inverse matrix of the control force distribution matrix, the control force that each actuator should output is calculated; The calculated control forces are converted into control commands for driving the actuators.

7. The method for controlling a manned flying saucer door according to claim 6, characterized in that: In the process of calculating the control force required for each mode, a control force penalty term is added for specific modes that are likely to cause hatch seal failure or loosening of the lock.

8. The method for controlling a manned flying saucer door according to claim 6, characterized in that: After calculating the control force that each actuator should output, the control force is limited to obtain the final control force.

9. The method for controlling a manned flying saucer door according to claim 1, characterized in that: When the monitoring result is abnormal, the control command is adjusted and the actuator is controlled to exert control force on the door structure by outputting the adjusted control command, or the control command is stopped from being output: Obtain the control force output amplitude and frequency information of the current actuator; Calculate the control force adjustment factor according to the control force output amplitude and frequency information; Based on the control force adjustment factor, the original control instruction is corrected to obtain the adjusted control instruction; Determine whether the adjusted control command amplitude is lower than the minimum control command threshold. If so, stop outputting the control command. If not, send the adjusted control command to the actuator to control the actuator to apply control force on the door structure.

10. A manned flying saucer hatch control system, characterized in that: include: an acquisition module, configured to arrange acceleration sensors at multiple locations of the hatch to form an acceleration sensor array, and obtain acceleration signals at multiple locations of the hatch through the acceleration sensor array; An estimation module, configured to estimate the vibration modal coordinates of the cabin door by decoupling the acceleration signal; A generation module, used for generating control instructions according to vibration modal coordinates; The control module is used to monitor the sealing pressure and locking status of the hatch. When the monitoring result is normal, the control module outputs a control instruction to control the actuator to exert a control force on the hatch structure. When the monitoring result is abnormal, the control instruction is adjusted and the actuator is controlled to exert a control force on the hatch structure by outputting the adjusted control instruction, or the output of the control instruction is stopped.

Images