Helicopter unmanned attitude control method and system based on gyro mechanical stabilization

Abstract

The present application relates to the technical field of helicopter posture control, and discloses a gyro-mechanical stabilization-based unmanned helicopter posture control method and system, which can obtain precession characteristic parameters of a gyro-mechanical stabilization unit in real time and non-invasively by injecting a weak detection precession torque disturbance signal in an autonomous flight state and analyzing the posture response of the helicopter, identify the drift generated in long-term operation of the gyro-mechanical stabilization unit, and adaptively adjust control parameters according to the drift, so as to effectively compensate the influence of the drift on the posture control accuracy and significantly improve the flight stability and task execution reliability of the unmanned helicopter.

Description

Technical Field

[0001] This application relates to the field of helicopter attitude control technology, and more specifically, to a method and system for unmanned helicopter attitude control based on gyromechanical stabilization. Background Technology

[0002] When performing high-precision tasks, such as inspecting high-rise buildings and power line inspections, unmanned helicopters (UHHs) require extremely high accuracy in maintaining their attitude. To cope with complex airflow environments, many UHHs integrate gyro-mechanical stabilization units (GMUs), which provide rapid attitude compensation through mechanical inertia, assisting the electronic flight control system in improving stability. However, during long-term operation, the physical performance of the GMU may gradually decline. For example, uneven wear of the support bearings can lead to an irregular increase in internal friction of the gyroscope, which in turn affects the mechanical linkage system, causing minute gaps or misalignments when connecting the swashplate. This decline in physical performance is not a sudden failure but rather a slow, cumulative process that is difficult to detect initially by conventional fault detection systems.

[0003] When unmanned helicopters perform precision hovering missions, inaccuracies and delays in the aforementioned mechanical path can cause slight deviations between the swashplate's actual mechanical response and electronic control commands. For example, when encountering local airflow disturbances, the electronic flight control system issues corrective commands, but due to gaps and delays in the mechanical linkages, the swashplate's actual movement may be delayed or incomplete, leading to attitude overshoot. This overshoot can rapidly worsen initially low-amplitude oscillations into larger and more violent attitude swings. If, at this point, the ground operator, based on incomplete information, misjudges the electronic control system's insufficient response and increases the control weight, they may issue stronger and faster commands to the already delayed mechanical system, further exacerbating the overshoot and reverse swings, leading to increased system oscillations and even pushing the system close to the brink of loss of control.

[0004] The combined effects of this mechanical performance degradation, electronic control system failure, and human intervention severely weaken the core stabilization function of the gyro-mechanical stabilization unit. This ultimately causes the helicopter's attitude control system to fall into divergent oscillations, making it unable to maintain precise hovering and seriously threatening flight safety and mission execution. Current technology lacks a method to accurately identify the precession drift of the gyro-mechanical stabilization unit during long-term operation in real time and adaptively adjust control parameters accordingly to effectively address this complex failure.

[0005] To address the aforementioned issues, existing technologies urgently need improvement. Summary of the Invention

[0006] This application discloses a method and system for unmanned helicopter attitude control based on gyromechanical stabilization, which aims to solve the problem that the physical performance of the gyromechanical stabilization unit deteriorates during long-term operation of unmanned helicopters, resulting in reduced attitude control accuracy and even system oscillation, threatening flight safety and mission execution.

[0007] The technical solution of this application is as follows: Firstly, this application discloses a method for unmanned helicopter attitude control based on gyro-mechanical stabilization, specifically including: When the helicopter is in autonomous flight without pilot intervention, the detected precession torque disturbance signal is superimposed with the normal precession torque command to generate a composite precession torque command, which is then applied to the helicopter's gyromechanical stabilization unit. The gyromechanical stabilization unit generates attitude stabilizing torque through the gyro precession effect. The amplitude of the detected precession torque disturbance signal is limited to not interfering with the precession effect of the normal precession torque command, and the spectrum of the detected precession torque disturbance signal is within the precession response frequency band of the gyromechanical stabilization unit. Acquire attitude response data of the helicopter to compound precession torque commands; Separate and extract the precession response component corresponding to the detected precession torque disturbance signal from the attitude response data; wherein, the precession response component contains the current precession characteristic parameters of the gyro mechanical stabilization unit; The current precession characteristic parameters are compared with the reference precession characteristic parameters of the gyro mechanical stabilization unit in the initial state to identify the precession characteristic drift generated by the gyro mechanical stabilization unit during long-term operation. The control parameters of the normal precession torque command are adjusted according to the precession characteristic drift to compensate for the impact of precession characteristic drift on the accuracy of autonomous attitude control. Generate attitude control commands based on the adjusted control parameters and output them.

[0008] Secondly, this application also discloses an unmanned helicopter attitude control system based on gyromechanical stabilization, comprising: The signal superposition module is used to superimpose the detected precession torque disturbance signal with the normal precession torque command to generate a composite precession torque command. The amplitude of the detected precession torque disturbance signal is limited to not interfere with the precession effect of the normal precession torque command, and the spectrum of the detected precession torque disturbance signal is within the precession response frequency band of the gyro mechanical stabilization unit. The drive output module is used to apply compound precession torque commands to the helicopter's gyro-mechanical stabilization unit; The acquisition module is used to obtain the helicopter's attitude response data to the compound precession torque command; The signal separation module is used to separate and extract the precession response component corresponding to the detected precession torque disturbance signal from the attitude response data; wherein, the precession response component contains the current precession characteristic parameters of the gyro mechanical stabilization unit; The drift identification module is used to compare the current precession characteristic parameters with the reference precession characteristic parameters of the gyro mechanical stabilization unit in the initial state, and to identify the precession characteristic drift generated by the gyro mechanical stabilization unit during long-term operation. The parameter compensation module is used to adjust the control parameters of the normal precession torque command according to the precession characteristic drift, so as to compensate for the impact of the precession characteristic drift on the accuracy of autonomous attitude control. The instruction generation module is used to generate attitude control instructions based on the adjusted control parameters and output them.

[0009] This application discloses an unmanned helicopter attitude control method and system based on gyromechanical stabilization. By cleverly superimposing the detected precession torque disturbance signal with the normal precession torque command during autonomous helicopter flight, and limiting the amplitude and spectrum of the disturbance signal, the gyromechanical stabilization unit is effectively excited without interfering with normal flight. Subsequently, by accurately acquiring and analyzing the helicopter's attitude response data, the precession response component containing the current precession characteristic parameters is separated and extracted. By comparing these real-time parameters with reference parameters, this application can accurately identify the precession characteristic drift generated by the gyromechanical stabilization unit during long-term operation. Accordingly, the system can adaptively adjust the control parameters of the normal precession torque command, thereby effectively compensating for the impact of precession characteristic drift on the accuracy of autonomous attitude control, and generating adjusted attitude control commands.

[0010] This method overcomes the problem of existing technologies being unable to accurately identify the performance degradation of gyro-mechanical stabilization units in real time, avoiding the reduced attitude control accuracy and system oscillations caused by the combined effects of decreased mechanical performance, electronic control systems, and human intervention. Through active detection and adaptive compensation, this application can effectively address the risks of attitude overshoot, increased oscillations, and even loss of control caused by the deterioration of the physical performance of gyro-mechanical stabilization units when unmanned helicopters perform high-precision missions. Therefore, this application significantly improves the attitude maintenance accuracy, flight safety, and mission execution capabilities of unmanned helicopters in complex airflow environments, and has significant practical application value. Attached Figure Description

[0011] Figure 1 This is a flowchart illustrating a method for unmanned helicopter attitude control based on gyro-mechanical stabilization provided in an embodiment of the present invention. Figure 2 This is a schematic diagram of the structure of an unmanned helicopter attitude control system based on gyro mechanical stabilization provided in an embodiment of the present invention. Detailed Implementation

[0012] The technical solutions of this application will now be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are merely some embodiments of this application, and not all embodiments. The components of this application described and shown in the accompanying drawings can generally be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of this application provided in the accompanying drawings is not intended to limit the scope of the claimed application, but merely to illustrate selected embodiments of this application. All other embodiments obtained by those skilled in the art based on the embodiments of this application without inventive effort are within the scope of protection of this application.

[0013] It should be noted that similar reference numerals and letters in the following figures indicate similar items; therefore, once an item is defined in one figure, it does not need to be further defined and explained in subsequent figures. Furthermore, in the description of this application, terms such as "first," "second," etc., are used only to distinguish descriptions and should not be construed as indicating or implying relative importance.

[0014] Traditional unmanned helicopters (UHHs) require extremely high attitude control precision when performing high-precision missions. To cope with complex airflow environments, many UHHs integrate gyromechanical stabilization units (GMUs) to provide rapid attitude compensation through mechanical inertia, assisting the electronic flight control system in improving stability. However, during long-term operation, the physical performance of the GMU may gradually decline. For example, uneven wear of the support bearings can lead to irregular increases in internal friction of the gyroscope, which in turn affects the mechanical linkage system, causing minute gaps or misalignments when connecting the swashplate. This decline in physical performance is not a sudden failure but rather a slow, cumulative process that is difficult to detect initially by conventional fault detection systems. When the UHH performs precision hovering missions, these inaccuracies and delays in the mechanical path can cause slight deviations between the actual mechanical response of the swashplate and the electronic control commands, leading to attitude overshoot. This causes initially low-amplitude oscillations to rapidly worsen into larger, more violent attitude swings, ultimately causing the helicopter's attitude control system to fall into divergent oscillations, making it unable to maintain precise hovering and seriously threatening flight safety and mission execution. Existing technologies lack a method to accurately identify the precession characteristic drift of the gyro mechanical stabilization unit during long-term operation in real time, and to adaptively adjust the control parameters accordingly, in order to effectively deal with this type of complex fault.

[0015] In this regard, refer to Figure 1 , Figure 1 This is a flowchart illustrating a method for unmanned helicopter attitude control based on gyro-mechanical stabilization provided in an embodiment of the present invention, including: S11, when the helicopter is in an autonomous flight state without pilot intervention, the detected precession torque disturbance signal is superimposed with the normal precession torque command to generate a composite precession torque command, which is then applied to the helicopter's gyromechanical stabilization unit; wherein, the gyromechanical stabilization unit generates attitude stabilizing torque through the gyro precession effect, the amplitude of the detected precession torque disturbance signal is limited to not interfering with the precession effect of the normal precession torque command, and the spectrum of the detected precession torque disturbance signal is within the precession response frequency band of the gyromechanical stabilization unit; S12, acquire the helicopter's attitude response data to the compound precession torque command; S13, Separate and extract the precession response component corresponding to the detected precession torque disturbance signal from the attitude response data; wherein, the precession response component contains the current precession characteristic parameters of the gyro mechanical stabilization unit; S14, compare the current precession characteristic parameters with the reference precession characteristic parameters of the gyro mechanical stabilization unit in the initial state, and identify the precession characteristic drift generated by the gyro mechanical stabilization unit during long-term operation; S15, adjust the control parameters of the normal precession torque command according to the precession characteristic drift to compensate for the impact of precession characteristic drift on the accuracy of autonomous attitude control; S16: Generate attitude control commands based on the adjusted control parameters and output them.

[0016] This application, by injecting a weak probe precession torque disturbance signal during autonomous flight and analyzing the helicopter's attitude response, can acquire the precession characteristic parameters of the gyromechanical stabilization unit in real time and non-invasively, thereby identifying the drift generated during long-term operation and adaptively adjusting the control parameters accordingly. This effectively compensates for the impact of drift on attitude control accuracy and significantly improves the flight stability and mission execution reliability of the unmanned helicopter.

[0017] The method proposed in this application is mainly applied to the autonomous flight control scenario of unmanned helicopters, aiming to solve the performance drift problem that may occur in gyromechanical stabilization units during long-term operation. The "gyromechanical stabilization unit" refers to a mechanical device that uses the precession effect of a gyroscope to generate attitude stabilizing torque. Its core lies in the precession torque generated by a high-speed rotating rotor when subjected to an external torque, perpendicular to both the external torque and the rotor axis, thereby achieving passive or active attitude stabilization of the helicopter. This unit is typically connected to the helicopter's swashplate system, and through mechanical linkage or an electro-hydraulic servo system, the precession torque is converted into periodic pitch or collective pitch control of the rotor blades, thereby changing the direction of rotor thrust and achieving attitude adjustment.

[0018] The "precession torque disturbance detection signal" is a carefully designed, amplitude-limited, and frequency-limited weak signal located within the precession response band of the gyro-mechanical stabilization unit. Its main function is to excite the precession response of the gyro-mechanical stabilization unit without significantly interfering with the helicopter's normal flight attitude, thus enabling online detection of its dynamic characteristics. The amplitude of this signal is strictly limited to ensure that it does not imperceptibly interfere with the precession effect of normal precession torque commands, thereby avoiding any impact on the helicopter's flight safety and stability. Simultaneously, its frequency spectrum is designed within the precession response band of the gyro-mechanical stabilization unit to effectively excite the unit's inherent dynamic response, thereby obtaining accurate precession characteristic parameters.

[0019] "Normal precession torque command" refers to the routine control command generated by the helicopter's main flight control system based on the flight mission and current attitude error, used to maintain or change the helicopter's attitude. This command is the core input for the helicopter's autonomous attitude control, and its accuracy and timeliness are directly related to flight performance.

[0020] The "composite precession torque command" is the result of superimposing the detected precession torque disturbance signal and the normal precession torque command. It is applied to the gyro-mechanical stabilization unit to simultaneously achieve attitude control and precession characteristic detection.

[0021] "Attitude response data" refers to the attitude angles, angular rates, and other data collected by the helicopter's attitude sensors (such as the inertial measurement unit, IMU) after receiving a compound precession torque command. These data reflect the actual dynamic response of the gyro-mechanical stabilization unit under the action of the compound command.

[0022] The "precession response component" is the part directly related to the detected precession torque disturbance signal, separated and extracted from the attitude response data. This component includes the current precession characteristic parameters of the gyro mechanical stabilization unit, such as precession response delay and precession response gain, which are key indicators for evaluating the unit's performance status.

[0023] "Precession drift" refers to the change in the precession characteristic parameters of a gyroscope's mechanical stabilization unit relative to its initial or reference state during long-term operation. This drift is usually caused by factors such as mechanical wear, aging, and environmental changes, and can lead to a decline in the performance of the control system.

[0024] "Control parameters" refer to the parameters used to adjust attitude control performance in normal precession torque commands, such as feedforward phase and feedback gain. By adjusting these parameters, the impact of precession characteristic drift on attitude control accuracy can be compensated.

[0025] In the embodiments of this application, when the helicopter is in an autonomous flight state without pilot involvement, the detected precession torque disturbance signal is superimposed with the normal precession torque command to generate a composite precession torque command, which is then applied to the helicopter's gyromechanical stabilization unit. The amplitude of the detected precession torque disturbance signal is limited to not interfering with the precession effect of the normal precession torque command, and the spectrum of the detected precession torque disturbance signal is within the precession response frequency band of the gyromechanical stabilization unit.

[0026] Specifically, the generation of the precession torque disturbance signal can be achieved in several ways. For example, a sinusoidal signal with a specific frequency and amplitude can be pre-designed, or a frequency-sweeping signal can be used to scan the precession response band of the gyro-mechanical stabilization unit. Another approach is to generate a pseudo-random sequence that exhibits randomness in the time domain but has controllable spectral characteristics in the frequency domain, with its energy distributed within the precession response band. Spread spectrum signals can also be used, expanding narrowband signals to broadband to improve signal anti-interference capability and stealth. In practice, the amplitude of the precession torque disturbance signal needs to be strictly calibrated and limited to ensure its impact on the helicopter's normal attitude control is negligible. For example, a maximum permissible amplitude can be determined experimentally or through simulation, ensuring that the attitude deviation caused by the disturbance signal does not exceed a preset safety threshold under the most sensitive flight conditions. Simultaneously, to ensure that the detection signal can effectively excite the response of the gyro-mechanical stabilization unit, its spectrum must match the precession response band of that unit. For example, if the precession response band of the gyroscope's mechanical stabilization unit is 0.5 Hz to 5 Hz, then the frequency components of the probe signal should be mainly concentrated within this range.

[0027] When generating a composite precession torque command, the superposition of the detected precession torque disturbance signal and the normal precession torque command can be achieved by a hardware adder. For example, in a digital control system, the two digital signals can be summed using a software algorithm; in an analog control system, the signals can be superimposed using analog circuits such as operational amplifiers.

[0028] Subsequently, the helicopter's attitude response data to the compound precession torque command is acquired. This attitude response data acquisition is typically accomplished using an inertial measurement unit (IMU) onboard the helicopter. The IMU usually contains gyroscopes and accelerometers, capable of measuring the helicopter's angular rate and linear acceleration in real time, and calculating the helicopter's attitude angles (such as pitch, roll, and yaw) using attitude calculation algorithms. This data is output digitally and recorded and stored by the flight control system. For example, the IMU can continuously acquire attitude data at a sampling rate of 100Hz and transmit this data to the onboard data recorder or flight control computer for further processing.

[0029] Next, the precession response component corresponding to the detected precession torque disturbance signal is separated and extracted from the attitude response data. This precession response component contains the current precession characteristic parameters of the gyromechanical stabilization unit. Various methods exist for separating and extracting the precession response component. A common method is to perform a frequency domain transformation on the attitude response data, such as a Fast Fourier Transform (FFT), converting the time-domain data to the frequency domain. In the frequency domain, since the spectrum of the detected precession torque disturbance signal is known and falls within a specific frequency band, techniques such as bandpass filtering can be used to separate the attitude response component corresponding to the frequency of this disturbance signal. Another method is to use correlation detection, performing correlation operations between the attitude response data and a preset detected precession torque disturbance signal, and identifying and extracting the precession response component based on the position and amplitude of the correlation peak. Furthermore, matched filtering techniques can be used, designing a filter matched to the detected precession torque disturbance signal to filter the attitude response data, thereby maximizing the extraction of the response component related to the detected signal. For example, if the detected precession torque disturbance signal is a sine wave of known frequency, a narrowband filter with the same center frequency as the detected signal can be used to filter the attitude response data to obtain the precession response component.

[0030] Then, the current precession characteristic parameters are compared with the reference precession characteristic parameters of the gyromechanical stabilization unit in its initial state to identify precession characteristic drift generated by the gyromechanical stabilization unit during long-term operation. The reference precession characteristic parameters are typically obtained through ground testing or initial flight testing when the gyromechanical stabilization unit leaves the factory or before the helicopter's first use, and are stored in the non-volatile memory of the flight control system. Comparing the current precession characteristic parameters with the reference precession characteristic parameters can be achieved by calculating the difference or ratio between them. For example, the difference between the current precession response delay and the reference precession response delay, and the ratio of the current precession response gain to the reference precession response gain, can be calculated. These differences or ratios reflect the magnitude and direction of the precession characteristic drift.

[0031] Finally, the control parameters of the normal precession torque command are adjusted based on the precession characteristic drift to compensate for the impact of precession characteristic drift on the accuracy of autonomous attitude control. Attitude control commands are then generated and output based on the adjusted control parameters. Various strategies can be employed to adjust the control parameters. For example, if an increased precession response delay is detected, the control command can be issued earlier by adjusting the feedforward phase of the normal precession torque command to offset the delay. If a decrease in precession response gain is detected, the response strength of the control system can be enhanced by increasing the feedback gain of the normal precession torque command to compensate for the gain loss. These adjustments can be achieved through parameter tuning of a proportional-integral-derivative (PID) controller or other advanced control algorithms. The adjusted control parameters are used to generate the final attitude control command, which is then sent to the helicopter's actuators (such as servo motors) to achieve precise attitude control of the helicopter.

[0032] The unmanned helicopter attitude control method based on gyromechanical stabilization proposed in this application works by using a non-intrusive, online detection mechanism to monitor the performance status of the gyromechanical stabilization unit in real time, and adaptively adjust the attitude control strategy accordingly to cope with performance drift that may occur during long-term operation.

[0033] This application further identifies precession drift by comparing the current precession characteristic parameters with reference parameters, and adaptively adjusts the control parameters of the normal precession torque command accordingly. This adaptive compensation mechanism is not found in existing technologies. Traditional control systems typically use fixed parameters. When the performance of the gyromechanical stabilization unit drifts, the controller with fixed parameters cannot effectively cope, leading to a decrease in control accuracy. The adaptive adjustment capability of this application enables the control system to dynamically optimize according to the actual performance changes of the gyromechanical stabilization unit, thereby effectively compensating for the impact of drift on attitude control accuracy and ensuring that the helicopter maintains high-precision attitude control throughout its entire lifespan.

[0034] For example, in precision hovering tasks, if the precession response delay of the gyro mechanical stabilization unit increases, conventional systems may experience attitude overshoot or oscillation. However, the method in this application can detect the increase in delay in real time and adjust the feedforward phase accordingly, issuing control commands in advance, thereby effectively counteracting the delay effect, avoiding attitude overshoot, and significantly improving hovering accuracy and stability.

[0035] In summary, this application overcomes the limitations of existing technologies in monitoring and compensating for performance drift of gyro mechanical stabilization units by introducing an innovative mechanism of online detection, precise parameter extraction, and adaptive compensation. It provides a more robust, accurate, and reliable solution for the autonomous attitude control of unmanned helicopters, significantly improving flight safety and mission execution efficiency.

[0036] In some embodiments described above in this application, a method is proposed to identify the precession characteristic drift of a gyro mechanical stabilization unit by superimposing a precession torque disturbance signal. However, if the frequency component of the precession torque disturbance signal is singular or the excitation method is insufficient, it may not be able to fully and effectively excite the dynamic characteristics of the gyro mechanical stabilization unit in the entire precession response frequency band, thereby affecting the accuracy and robustness of the precession characteristic parameter identification.

[0037] In this regard, this application further proposes that the aforementioned precession torque disturbance signal includes multiple frequency components, the multiple frequency components covering multiple sub-frequency bands within the precession response frequency band of the gyroscope mechanical stabilization unit; the precession torque disturbance signal is one of a swept frequency signal, a pseudo-random sequence, or a spread spectrum signal; the precession torque disturbance signal is superimposed on the normal precession torque command through an adder.

[0038] Specifically, the detected precession torque disturbance signal is designed to have multiple frequency components, which are configured to cover multiple sub-bands within the precession response frequency band of the gyro mechanical stabilization unit. This design aims to ensure sufficient excitation and detection of the response characteristics of the gyro mechanical stabilization unit at different frequencies. The detected precession torque disturbance signal can be understood as a dynamic excitation signal, the form of which can be selected according to actual application requirements and system characteristics. For example, it can be a swept-frequency signal, where the signal frequency changes continuously or abruptly within a certain range; it can also be a pseudo-random sequence, such as a pseudo-random binary sequence (PRBS), which has spectral characteristics similar to white noise but is deterministic and repeatable; or it can be a spread-spectrum signal, which improves anti-interference capability and signal detection robustness by extending a narrow-band signal to a wide bandwidth. In practical applications, the detected precession torque disturbance signal is superimposed on the normal precession torque command through an adder to generate a composite precession torque command, which is then applied to the gyro mechanical stabilization unit.

[0039] The proposed solution employs a probed precession torque disturbance signal containing multiple frequency components and covering multiple sub-bands, enabling a more comprehensive excitation of the dynamic response of the gyro-mechanical stabilization unit (GMU) across the entire precession response frequency band. When using a swept-frequency signal, the precession response of the GMU at different frequencies can be systematically detected, thereby acquiring its frequency response characteristics. When using a pseudo-random sequence, its wideband spectral characteristics and good autocorrelation properties allow for simultaneous excitation of multiple frequency points, and precession response components can be efficiently extracted from noise through methods such as correlation detection. When using a spread-spectrum signal, its inherent anti-interference capability helps improve the signal-to-noise ratio of the probed signal in complex flight environments, ensuring accurate extraction of the precession response components. By superimposing the probed precession torque disturbance signal with the normal precession torque command using an adder, the probed signal can be effectively injected into the input of the GMU without significantly interfering with normal attitude control, thus achieving online identification of precession characteristic parameters.

[0040] In some preferred embodiments, as a specific implementation, the detected precession torque disturbance signal can be designed as a swept-frequency sinusoidal signal, whose frequency linearly scans from the lower limit (e.g., 0.1 Hz) to the upper limit (e.g., 10 Hz) of the precession response band of the gyromechanical stabilizer, with a scan period set to several seconds, to ensure sufficient excitation of the gyromechanical stabilizer without affecting the normal flight attitude of the helicopter. The amplitude of this swept-frequency signal is precisely controlled to meet the constraint of not interfering with the precession effect of the normal precession torque command. Specifically, the swept-frequency signal is generated by a digital signal generator and superimposed in real time with the normal precession torque command output by the attitude controller through an adder. The superimposed composite precession torque command is then sent to the drive mechanism of the gyromechanical stabilizer, such as a servo motor, to generate the corresponding precession torque. In another embodiment, the detected precession torque disturbance signal can be a pseudo-random binary sequence (PRBS), whose sequence length and clock frequency are selected so that its spectral characteristics can uniformly cover the precession response band of the gyromechanical stabilizer. The PRBS signal is also superimposed on the normal precession torque command through an adder, and the precession response component is efficiently extracted from the attitude response data through subsequent related detection algorithms.

[0041] Specifically, the steps described above for separating and extracting the precession response component corresponding to the detected precession torque disturbance signal from the attitude response data can be further refined. The steps for separating and extracting the precession response component corresponding to the detected precession torque disturbance signal from the attitude response data include: The attitude response data is subjected to frequency domain transformation, correlation detection, or matched filtering to separate the in-phase and quadrature components corresponding to the detected precession torque disturbance signal. The current precession characteristic parameters of the gyroscope mechanical stabilization unit are calculated based on the in-phase component and the quadrature component; wherein the precession characteristic parameters include precession response delay and precession response gain.

[0042] Various techniques can be used to process attitude response data to separate precession response components.

[0043] For example, frequency domain transformation refers to converting time-domain attitude response data to the frequency domain using mathematical methods such as Fourier transform, in order to identify and extract the response components corresponding to the frequency of the detected precession torque disturbance signal. Correlation detection involves calculating the cross-correlation function between the attitude response data and the known detected precession torque disturbance signal to detect and extract the response component with the highest correlation to the disturbance signal. Matched filtering involves designing a filter that matches the waveform of the detected precession torque disturbance signal, passing the attitude response data through this filter to maximize the signal-to-noise ratio and effectively extract the precession response components. Through these processes, the in-phase and quadrature components corresponding to the detected precession torque disturbance signal can be separated, and these components contain the phase and amplitude information of the precession response.

[0044] Furthermore, the precession characteristic parameters specifically include precession response delay and precession response gain. Precession response delay refers to the time lag required for the gyromechanical stabilization unit to generate a corresponding attitude response after receiving a precession torque command, reflecting the dynamic response speed of the system. Precession response gain refers to the ratio between the amplitude of the attitude response of the gyromechanical stabilization unit and the amplitude of the input torque under a specific precession torque input, reflecting the strength and efficiency of its response. These precession characteristic parameters can be accurately calculated by analyzing the separated in-phase and quadrature components. For example, the precession response delay can be calculated by analyzing the phase relationship between the in-phase and quadrature components, combined with the known phase of the detected precession torque disturbance signal; the precession response gain can be determined by analyzing the ratio between the combined amplitude of the in-phase and quadrature components and the amplitude of the detected precession torque disturbance signal.

[0045] The proposed solution effectively identifies and separates the weak precession response component caused by the probed precession torque disturbance signal from complex attitude response data by performing frequency domain transformation, correlation detection, or matched filtering. These processing methods utilize known characteristics (such as frequency and waveform) of the probed precession torque disturbance signal to accurately extract the target signal from noise and non-precession responses. Specifically, the phase and amplitude information of the precession response can be accurately analyzed by separating the in-phase and quadrature components. The precession response delay can be calculated by analyzing the phase relationship between the in-phase and quadrature components and combining it with the known phase of the probed precession torque disturbance signal. The precession response gain can be determined by analyzing the ratio between the combined amplitude of the in-phase and quadrature components and the amplitude of the probed precession torque disturbance signal. Therefore, the current precession characteristic parameters of the gyro mechanical stabilization unit, namely the precession response delay and precession response gain, can be accurately calculated and quantified.

[0046] This application further proposes the steps for identifying the precession characteristic drift of the gyroscope mechanical stabilization unit during long-term operation, specifically including: A recursive estimation method with a forgetting factor is used to estimate the current value and trend of the precession characteristic drift based on the current and historical precession characteristic parameters.

[0047] Specifically, the recursive estimation method with a forgetting factor is an algorithm that can update the estimation results in real time based on newly acquired data and gradually reduce the influence of old data. The "forgetting factor" is a parameter between 0 and 1, used to control the weight of historical data in the estimation process. When the forgetting factor is close to 1, the influence of historical data is greater; when the forgetting factor is smaller, the system responds faster to new data and "forgets" old data more quickly, allowing the estimator to better adapt to slow changes in system parameters. The purpose of this method is to improve the tracking ability and estimation accuracy of dynamically changing precession drift. Current precession parameters refer to parameters extracted from the attitude response data by the signal separation module that reflect the current precession characteristics of the gyro mechanical stabilization unit, such as precession response delay and precession response gain. Historical precession parameters refer to precession parameter data acquired and stored at previous time points; these data, together with the current parameters, constitute the data basis for estimating drift. Estimating the current value and trend of precession drift means that by using a recursive estimation method, we can not only obtain the real-time value of the drift, but also analyze its rate and direction of change over time, thus providing more comprehensive information for subsequent control parameter adjustments.

[0048] This application's solution, by introducing a recursive estimation method with a forgetting factor, overcomes the limitations of traditional simple comparison methods in identifying precession drift under dynamic and noisy conditions. Through this technical solution, the application significantly improves the accuracy and robustness of identifying precession drift in gyromechanical stabilization units. Compared to simple parameter comparison, the recursive estimation method with a forgetting factor better adapts to the slow changes and random disturbances in precession parameters, effectively filtering out noise and thus obtaining a more accurate current drift value and trend. This allows for more precise and timely adjustment of control parameters for subsequent normal precession torque commands, effectively compensating for the impact of precession drift on the accuracy of autonomous attitude control and ensuring the stability and reliability of helicopter attitude control during long-term operation. Furthermore, the recursive estimation characteristic reduces the demand for computing resources and storage space, improving the system's real-time processing capabilities.

[0049] In some preferred embodiments, a specific example is given below. Suppose that during autonomous helicopter flight, the precession characteristic parameters (e.g., precession response delay and precession response gain) of the gyromechanical stabilization unit may slowly drift due to factors such as temperature changes and component wear. To accurately identify this drift, the system can employ a Kalman filter or a least-squares recursive algorithm as a recursive estimation method with a forgetting factor. Specifically, in each detection cycle, after acquiring the current precession characteristic parameter, this parameter is input into the recursive estimator. The estimator combines the estimation results from previous moments with a preset forgetting factor (e.g., set to 0.95) to calculate the current estimated value of the precession characteristic drift. The forgetting factor ensures that newer data has a greater influence on the estimation results, while older data is gradually "forgotten," allowing the estimator to track the dynamic changes in drift in real time. For example, if the precession response delay gradually increases from a baseline value, this recursive estimation method can smoothly identify this increasing trend and provide an accurate current delay drift value, while also estimating its drift rate. In this way, stable and accurate drift estimation results can be obtained even in the presence of measurement noise.

[0050] In some embodiments described above, this application proposes adjusting the control parameters of the normal precession torque command based on precession characteristic drift to compensate for the impact of drift on the accuracy of autonomous attitude control. However, in actual implementation, precession characteristic drift may manifest in various forms, such as changes in precession response delay or precession response gain. Failure to specifically adjust these different types of drift may result in poor compensation effects or even introduce new control errors, thereby affecting the accuracy and stability of helicopter attitude control.

[0051] In response, this application further proposes a more refined control parameter adjustment strategy, which aims to adjust the control parameters of the normal precession torque command in a targeted manner according to the specific manifestation of precession characteristic drift.

[0052] The steps described above for adjusting the control parameters of the normal precession torque command based on the precession characteristic drift include: Adjust the feedforward phase of the normal precession torque command based on the precession response delay change corresponding to the precession characteristic drift. Adjust the feedback gain of the normal precession torque command based on the change in precession response gain corresponding to the precession characteristic drift.

[0053] Specifically, precession response delay refers to the time lag required for the gyromechanical stabilization unit (GMU) to generate the corresponding attitude stabilizing torque after receiving a precession torque command. When the precession response delay changes during long-term operation, it means that its response speed to control commands has changed. To compensate for this delay change, the feedforward phase of the normal precession torque command needs to be adjusted. The feedforward phase can be understood as a phase advance or lag introduced in advance when the command is issued. Its purpose is to keep the actual response of the GMU synchronized with the expected response in time, thereby offsetting the impact of the delay change. For example, if an increase in precession response delay is detected, the feedforward phase needs to be advanced appropriately to ensure that the attitude stabilizing torque can be generated in a timely manner.

[0054] Precession response gain refers to the magnitude of the attitude stabilizing torque generated by the gyromechanical stabilization unit (GMU) after receiving a unit precession torque command. A change in the precession response gain of the GMU indicates a change in the efficiency or strength of its attitude stabilizing torque generation. To compensate for this gain change, the feedback gain of the normal precession torque command needs to be adjusted. Feedback gain is a parameter in the control system used to adjust the output response strength, aiming to ensure that the GMU can generate the required amplitude of attitude stabilizing torque under different operating conditions. In practical applications, if a decrease in precession response gain is detected, the feedback gain needs to be appropriately increased to enhance the output of the attitude stabilizing torque, and vice versa.

[0055] This application's solution subdivides precession drift into two specific forms: precession response delay variation and precession response gain variation. By specifically adjusting the feedforward phase and feedback gain of the normal precession torque command, it achieves precise compensation for the precession drift of the gyro-mechanical stabilization unit. Since the precession response delay directly affects the system's time response characteristics, adjusting the feedforward phase can effectively correct for attitude control lag or lead caused by delay variations, ensuring that the application of attitude stabilizing torque remains synchronized with the helicopter's attitude change requirements. Simultaneously, changes in precession response gain directly affect the output intensity of the attitude stabilizing torque. Adjusting the feedback gain can precisely compensate for insufficient or excessive torque output caused by gain variations, thereby maintaining the accuracy and stability of attitude control. This divide-and-conquer adjustment strategy makes the compensation mechanism more targeted and effective, avoiding the suboptimal compensation effect that may result from a single adjustment method.

[0056] The above technical solution enables refined and targeted compensation for the precession drift of the gyro-mechanical stabilization unit. Compared to simply adjusting control parameters, this application can adjust the feedforward phase and feedback gain separately based on the specific drift of the precession response delay and precession response gain, thereby more accurately offsetting the impact of drift on attitude control accuracy. This significantly improves the robustness and accuracy of autonomous attitude control of helicopters during long-term operation, effectively avoiding the decline in attitude control performance caused by the degradation of the gyro-mechanical stabilization unit, and ensuring the safety and reliability of unmanned flight.

[0057] In some preferred embodiments, a specific example is given below. Suppose that during long-term helicopter operation, analysis of attitude response data reveals a drift in the precession characteristic parameters of the gyromechanical stabilization unit. Specifically, the current precession response delay has increased by 5 milliseconds compared to the baseline value, while the precession response gain has decreased by 10% compared to the baseline value.

[0058] According to the scheme of this application, in the case of a 5-millisecond increase in precession response delay, the control system calculates the amount of feedforward phase that needs to be advanced. For example, if the center frequency of the precession response band of the gyro mechanical stabilization unit is 10 Hz, then a 5-millisecond delay is equivalent to 0.05 cycles, requiring the feedforward phase of the normal precession torque command to be advanced by 0.05 * 360 degrees = 18 degrees. By adjusting the feedforward phase, it can be ensured that the attitude stabilization torque generated by the gyro mechanical stabilization unit after receiving the command can be advanced by 18 degrees, thereby offsetting the attitude response lag caused by the increased delay.

[0059] Simultaneously, in response to a 10% decrease in precession response gain, the control system will adjust the feedback gain of the normal precession torque command accordingly. For example, if the original feedback gain is K, the new feedback gain will be adjusted to K / (1 - 0.1) = K / 0.9, an increase of approximately 11.1%. By increasing the feedback gain, the insufficient attitude stabilizing torque output caused by the decrease in gain of the gyro mechanical stabilization unit can be compensated, ensuring that a sufficiently large stabilizing torque can be generated under the same command.

[0060] Through this targeted adjustment, even if the precession characteristics of the gyromechanical stabilization unit drift, the helicopter's autonomous attitude control system can maintain high-precision attitude control performance, effectively addressing the performance degradation caused by long-term operation.

[0061] In some embodiments described above, the precession characteristic parameters of the gyromechanical stabilizer (GMS) are compared with the reference precession characteristic parameters of the GMS in its initial state to identify precession drift during long-term operation. However, the precession characteristic parameters of the GMS are affected not only by factors such as long-term wear and tear but also by external conditions such as ambient temperature and vibration. If these environmental factors are not compensated for, directly comparing the initial reference precession characteristic parameters may lead to inaccurate identification of precession drift, or even misjudging characteristic fluctuations caused by environmental changes as drift, thereby affecting the accuracy of subsequent attitude control command adjustments. Therefore, this application further proposes a scheme to perform environmental correction on the reference precession characteristic parameters before identifying precession drift in the GMS, in order to improve the accuracy of drift identification.

[0062] Specifically, before the step of identifying the precession characteristic drift of the gyroscope's mechanical stabilization unit during long-term operation, the following steps are also included: Obtain the current temperature and vibration parameters; Based on the temperature parameters and vibration parameters, as well as the preset temperature-precession characteristic mapping model and vibration-precession characteristic mapping model, the reference precession characteristic parameters are corrected for temperature and vibration respectively to obtain the corrected reference precession characteristic parameters corresponding to the current temperature parameters and vibration parameters.

[0063] Specifically, obtaining current temperature and vibration parameters refers to real-time monitoring of the helicopter's ambient temperature and the vibration level of the gyro-mechanical stabilization unit (GMU) during operation using onboard sensors. Temperature parameters can be obtained through temperature sensors installed near the GMU or inside the fuselage, while vibration parameters can be measured using vibration sensors (e.g., accelerometers) to measure the amplitude and frequency of vibration within the GMU or its supporting structure. These parameters are dynamically changing and need to be acquired in real-time or periodically to reflect the actual operating environment of the GMU. The preset temperature-precession characteristic mapping model and vibration-precession characteristic mapping model are mathematical models established through experimental calibration before the GMU leaves the factory or during periodic maintenance. These models describe how the precession characteristic parameters (such as precession response delay and precession response gain) of the GMU change under different temperature and vibration conditions. For example, the temperature-precession characteristic mapping model can be a function or lookup table, taking the current temperature as input and outputting the corresponding precession characteristic correction; similarly, the vibration-precession characteristic mapping model takes the current vibration parameters as input and outputs the corresponding correction. These models allow for the correction of the reference precession characteristic parameters of the gyroscope mechanical stabilization unit in its initial state based on current temperature and vibration parameters, thereby obtaining a corrected reference precession characteristic parameter that better matches the current actual working environment.

[0064] This application's solution addresses the potential deviation of reference precession characteristic parameters under different environmental conditions by incorporating environmental factors before identifying precession drift. Through this technical solution, the application effectively eliminates the interference of ambient temperature and vibration on the identification of precession characteristic parameters of the gyro-mechanical stabilization unit, thereby significantly improving the accuracy and reliability of precession drift identification. This avoids misjudgments caused by environmental factors, ensuring that subsequent adjustments to normal precession torque command control parameters are based on actual equipment drift, rather than environmental fluctuations. This not only improves the accuracy and stability of unmanned helicopter attitude control but also extends the effective service life of the gyro-mechanical stabilization unit and reduces maintenance costs.

[0065] In some preferred embodiments, it is assumed that during an autonomous flight mission, the helicopter takes off from the ground and climbs to a high altitude, with the ambient temperature dropping from 20°C to -10°C, while the vibration levels generated by the engine and rotor also change. Before identifying the precession characteristic drift of the gyromechanical stabilization unit, the current temperature parameter is first obtained as -10°C using an onboard temperature sensor, and the current vibration parameter is obtained using a vibration sensor. Subsequently, the system calls pre-stored temperature-precession characteristic mapping models and vibration-precession characteristic mapping models. For example, the temperature-precession characteristic mapping model might be a function f(T) used to calculate the correction amount for precession response delay and gain at temperature T; the vibration-precession characteristic mapping model might be a function g(V) used to calculate the correction amount at vibration level V. The baseline precession characteristic parameters under the initial state (e.g., calibrated at 25°C in a low-vibration environment) are input into these models, and combined with the current -10°C temperature parameter and vibration parameter, the corrected baseline precession characteristic parameters under the current high-altitude, low-temperature, and specific vibration environment are calculated. For example, if the low temperature causes an increase in precession response delay, the model will adjust the baseline delay value accordingly. Then, the current precession characteristic parameters obtained by detecting precession torque disturbance signals are compared with these corrected reference parameters. This allows for a more accurate identification of the actual precession characteristic drift generated by the gyromechanical stabilization unit during this flight, in addition to environmental influences. Based on this, subsequent control parameter adjustments will be more precise, ensuring the helicopter's attitude control accuracy in complex environments.

[0066] Traditional gyro-based mechanical stabilization methods for unmanned helicopter attitude control typically involve directly outputting and executing attitude control commands after adjusting control parameters based on identified precession drift. However, even with parameter adjustments, deviations in attitude response during actual flight can occur due to environmental changes, incompletely modeled nonlinear effects, or residual precession drift. If these issues are not addressed, the system may fail to detect and correct these deviations in a timely manner, affecting the long-term accuracy and stability of autonomous attitude control and potentially jeopardizing flight safety during critical flight phases.

[0067] In this regard, this application further proposes that after the above-mentioned steps of generating and outputting attitude control commands based on the adjusted control parameters, it also includes: Obtain the actual attitude data of the helicopter after it executes the attitude control command; The actual attitude data is compared with the expected attitude data to calculate the attitude control deviation; When the attitude control deviation exceeds the preset deviation threshold, the step of separating and extracting the precession response component corresponding to the detected precession torque disturbance signal from the attitude response data is re-executed.

[0068] Specifically, acquiring the actual attitude data of the helicopter after executing attitude control commands refers to collecting attitude information of the helicopter in real time during the execution of attitude control commands through airborne sensors, such as inertial measurement units (IMUs), global positioning system (GPS) receivers, and attitude and heading reference systems (AHRS). This information includes, but is not limited to, pitch angle, roll angle, yaw angle, and their rate of change. This data is used to reflect the helicopter's attitude state during actual flight.

[0069] The process of comparing actual attitude data with expected attitude data to calculate attitude control deviation involves comparing the real-time acquired actual attitude data with the target attitude data desired by the attitude control commands. Expected attitude data is typically generated by the flight control system based on mission planning and control commands. Attitude control deviation can be understood as the difference between the actual and expected attitude, such as angular or angular velocity differences, and its purpose is to quantify the accuracy of current attitude control.

[0070] In practical applications, when the attitude control deviation exceeds a preset deviation threshold, the step of separating and extracting the precession response component corresponding to the detected precession torque disturbance signal from the attitude response data is repeated. The preset deviation threshold is an allowable error range set according to flight mission requirements, helicopter performance, and safety standards. When the deviation between the actual attitude and the expected attitude exceeds this range, it indicates that the current attitude control effect is poor, which may mean that the precession characteristic parameters of the gyro-mechanical stabilization unit have changed or that the previous drift compensation was not accurate enough. At this time, the system will trigger a re-evaluation, that is, to re-identify the precession characteristic parameters to obtain the latest precession characteristic parameters, providing a basis for more accurate compensation in the future.

[0071] This application's solution effectively addresses the issue of unverified control accuracy that may exist in basic solutions by introducing a real-time monitoring and feedback mechanism for attitude control effectiveness. Through this technical solution, the application enables real-time monitoring and verification of the helicopter's autonomous attitude control effectiveness. When attitude control accuracy significantly decreases, the system can promptly detect and automatically trigger the re-identification and compensation process of precession characteristic parameters, effectively preventing further drift of precession characteristic parameters during long-term operation or continuous deterioration of control accuracy due to insufficient initial compensation. This significantly enhances the robustness and adaptability of the helicopter's unmanned attitude control system, ensuring long-term stability and high accuracy of helicopter attitude control under various flight conditions, and improving flight safety and mission reliability.

[0072] In some preferred embodiments, it is assumed that an unmanned helicopter is performing an autonomous hovering mission. According to the basic scheme, the system has identified precession drift in the gyromechanical stabilization unit and adjusted the control parameters of the normal precession torque command, subsequently generating and outputting control commands to maintain the hovering attitude. After the command output, the onboard attitude sensors continuously monitor the helicopter's actual pitch and roll angles. If, due to environmental disturbances or residual precession drift, the deviation between the helicopter's actual pitch or roll angle and the expected hovering attitude (e.g., 0 degrees pitch and 0 degrees roll) continuously exceeds a preset 0.5-degree threshold, the system will immediately determine that the current attitude control effect is unsatisfactory. At this time, the control system will no longer rely solely on previous parameter adjustments but will restart the injection of the precession torque disturbance signal and again separate and extract the precession response component from the attitude response data to obtain the latest precession characteristic parameters. Based on these updated parameters, the system will reassess the precession drift and make more precise adjustments to the control parameters, enabling the helicopter to quickly recover to a stable hovering attitude and maintain high-precision attitude control.

[0073] In some embodiments described above, this application proposes identifying precession drift of the gyromechanical stabilization unit by superimposing a precession torque disturbance signal. However, during implementation, if the precession torque disturbance signal is continuously applied, it may unnecessarily affect the helicopter's attitude stability during certain flight phases or increase the system's processing burden, thereby affecting overall control efficiency and accuracy. Therefore, this application further proposes optimizing the timing of applying the precession torque disturbance signal to minimize interference with normal helicopter flight while ensuring drift detection effectiveness.

[0074] Specifically, the detection of the precession torque disturbance signal is applied during at least one period of time, either during the helicopter's hovering phase, low-speed cruise phase, or within a preset detection window. The "hovering phase" refers to the helicopter's relatively stationary state in the air, where its flight attitude is relatively stable and external disturbances are minimal, making it an ideal time for system characteristic detection. The "low-speed cruise phase" refers to the helicopter's stable flight at a low speed with slow attitude changes, which also helps reduce the impact of the detection signal on flight control. The "preset detection window" can be understood as the system actively activating the detection function within a specific time period based on flight mission, environmental conditions, or maintenance needs. This window can be periodic or triggered on demand. "Applying during at least one period of time" means that one or more periods can be selected for detection based on actual conditions, balancing detection frequency and flight mission requirements.

[0075] The proposed solution effectively avoids introducing additional disturbances during high-dynamic maneuvers or critical flight phases by selectively applying probed precession torque disturbance signals during relatively stable or non-critical flight phases. During hovering and low-speed cruise phases, the helicopter is typically in a relatively stable flight state with low attitude change rates and accelerations. At this time, the attitude response data generated by applying the probe signal can more clearly and accurately reflect the precession characteristics of the gyromechanical stabilization unit, thereby improving the accuracy of precession response component extraction and more precisely identifying precession drift. Furthermore, by applying the probe signal within a preset detection window, the system can perform characteristic detection according to a predetermined maintenance plan or under specific conditions, thus avoiding the introduction of additional disturbances during high-difficulty or high-risk missions. This selective application method allows the probed precession torque disturbance signal to be effectively utilized while minimizing its potential impact on the helicopter's autonomous attitude control performance.

[0076] In some preferred embodiments, when the helicopter is performing a routine patrol mission and hovers at a checkpoint for observation, the system can automatically trigger the application of a precession torque disturbance detection signal. During this hovering phase, the helicopter's attitude is relatively stable, external interference is minimal, and the precession characteristic parameters of the gyromechanical stabilization unit can be accurately measured and analyzed. For example, during a long-distance, low-speed cruise flight, when the helicopter enters a stable, straight flight segment, the system can also periodically or at preset time points apply detection signals to continuously monitor the health status of the gyromechanical stabilization unit. Furthermore, during pre-flight checks or periodic maintenance, a specific detection window can be manually or via the ground control station. Within this window, precession drift detection and parameter adjustments can be performed centrally to ensure the system is in optimal condition before mission execution.

[0077] In some of the embodiments described above in this application, although the impact of drift on the accuracy of autonomous attitude control can be effectively compensated by identifying the precession characteristic drift of the gyro-mechanical stabilization unit and adjusting the control parameters, thereby maintaining the flight stability of the helicopter, this compensation mechanism mainly focuses on real-time performance correction and does not provide the ability to assess and predict the long-term health status of the gyro-mechanical stabilization unit itself. If the health trend of the gyro-mechanical stabilization unit is not grasped in time, once the drift accumulates to a critical point or accelerates, it may lead to compensation failure or even system failure, thereby affecting flight safety and mission reliability.

[0078] To address this, this application further proposes a scheme for health assessment and early warning of the gyro-mechanical stabilization unit, thereby achieving more comprehensive system management and maintenance. The aforementioned unmanned helicopter attitude control method based on gyro-mechanical stabilization also includes: Based on the changing trend of the precession characteristic drift and the preset drift-health score correspondence, the current health score of the gyroscope mechanical stabilization unit is calculated. When the current health score is lower than the preset health threshold, a maintenance prompt message is sent to the ground control station via the airborne communication link.

[0079] Specifically, the precession drift trend refers to the rate or acceleration of change of the precession characteristic parameters (such as precession response delay and precession response gain) of the gyroscope mechanical stabilization unit over time. This trend can be estimated by analyzing historical precession drift data. For example, methods such as linear regression, exponential smoothing, or Kalman filtering can be used to predict the future direction of the drift. The drift-health score correspondence can be understood as a pre-established model or lookup table that maps different precession drift trends to a quantified health score. This correspondence can be constructed based on extensive operational data, failure mode analysis, or expert experience, aiming to reflect the degree of impact of the drift trend on the overall health of the gyroscope mechanical stabilization unit. The current health score is a comprehensive indicator used to quantitatively assess the current operating status and potential failure risks of the gyroscope mechanical stabilization unit. When this score is below a preset health threshold, it indicates that the performance of the gyroscope mechanical stabilization unit has significantly deteriorated or there is a high risk of failure, requiring intervention. At this time, maintenance prompts are sent to the ground control station via airborne communication links, such as satellite communication links or data links, so that ground personnel can arrange inspections, repairs or replacements in a timely manner, thereby avoiding potential failures.

[0080] This application's solution addresses the problem of insufficient early warning for potential faults by introducing a monitoring and health assessment mechanism for precession characteristic drift trends. Through this technical solution, predictive maintenance of the gyro-mechanical stabilization unit (GMU) is achieved, significantly improving the overall reliability and safety of the helicopter's unmanned attitude control system. This solution not only compensates for the impact of precession characteristic drift on attitude control accuracy but, more importantly, provides a proactive health management capability, effectively preventing flight accidents or mission interruptions caused by sudden GMU failures. By promptly sending maintenance alerts to the ground control station, maintenance resources can be optimally allocated, reducing unplanned downtime, lowering operating costs, and extending the lifespan of the GMU. Furthermore, this health assessment mechanism provides valuable data support for system upgrades and improvements.

[0081] In some preferred embodiments, a specific example is given below. Assume that during autonomous helicopter flight, the system continuously monitors the precession response delay and precession response gain of the gyro mechanical stabilization unit and calculates its precession characteristic drift. By performing time-series analysis on these drift data, such as using a moving average or exponentially weighted average, the trend of precession characteristic drift can be obtained. For example, if the drift rate of the precession response delay continues to increase, or the drift amplitude of the precession response gain increases significantly in a short period of time, this may indicate that mechanical wear or aging of electronic components within the gyro mechanical stabilization unit is accelerating. The system internally predefines a drift-health score correspondence, which may be a multidimensional lookup table mapping parameters such as drift rate and drift acceleration to a health score of 0-100. For example, when the drift rate of the precession response delay exceeds a certain threshold, the health score will decrease accordingly. Assume the predetermined health threshold is 60. When the system calculates a current health score of 55, which is below the threshold, it immediately sends a maintenance notification to the ground control station via an onboard communication link, such as a satellite data link. This notification includes the helicopter's identification code, the current health score, and a suggested maintenance type (e.g., "Suggest checking the gyro-mechanical stabilization unit"). Upon receiving this information, the ground control station can schedule the helicopter to return for inspection in advance, intervening before the gyro-mechanical stabilization unit actually malfunctions. This effectively avoids the risk of attitude instability during flight and ensures the successful completion of the helicopter mission.

[0082] refer to Figure 2 , Figure 2 This is a schematic diagram of a helicopter unmanned attitude control system based on gyro-mechanical stabilization provided in an embodiment of the present invention, including: The signal superposition module is used to superimpose the detected precession torque disturbance signal with the normal precession torque command to generate a composite precession torque command. The amplitude of the detected precession torque disturbance signal is limited to not interfere with the precession effect of the normal precession torque command, and the spectrum of the detected precession torque disturbance signal is within the precession response frequency band of the gyro mechanical stabilization unit. The drive output module is used to apply compound precession torque commands to the helicopter's gyro-mechanical stabilization unit; The response acquisition module is used to acquire the helicopter's attitude response data to the compound precession torque command; The signal separation module is used to separate and extract the precession response component corresponding to the detected precession torque disturbance signal from the attitude response data; wherein, the precession response component contains the current precession characteristic parameters of the gyro mechanical stabilization unit; The drift identification module is used to compare the current precession characteristic parameters with the reference precession characteristic parameters of the gyro mechanical stabilization unit in the initial state, and to identify the precession characteristic drift generated by the gyro mechanical stabilization unit during long-term operation. The parameter compensation module is used to adjust the control parameters of the normal precession torque command according to the precession characteristic drift, so as to compensate for the impact of the precession characteristic drift on the accuracy of autonomous attitude control. The instruction generation module is used to generate attitude control instructions based on the adjusted control parameters and output them.

[0083] The unmanned helicopter attitude control system based on gyromechanical stabilization proposed in this application achieves online monitoring and adaptive compensation for performance drift of the gyromechanical stabilization unit through modular design. A signal superposition module fuses weak detected precession torque disturbance signals with normal precession torque commands to form a composite command. This composite command is applied to the gyromechanical stabilization unit via the drive output module, stimulating its dynamic response. A response acquisition module acquires the helicopter's attitude response data in real time, and a signal separation module accurately extracts the response components containing precession characteristic parameters. A drift identification module quantifies the precession characteristic drift by comparing current and reference parameters. Finally, a parameter compensation module adjusts the control parameters based on the identified drift, and a command generation module outputs the compensated attitude control command, thereby effectively improving the attitude control accuracy and flight stability of the unmanned helicopter during long-term operation and overcoming the limitations of traditional systems in dealing with the slow decline in mechanical performance.

[0084] The unmanned helicopter attitude control system based on gyro-mechanical stabilization proposed in this application can be implemented in various ways for each module.

[0085] Specifically, the signal superposition module can be a hardware adder. For example, in a digital control system, the summation of two digital signals can be achieved through a dedicated digital signal processor (DSP) or an arithmetic logic unit (ALU) in a microcontroller; in an analog control system, signal superposition can be achieved through analog circuits such as operational amplifiers. Alternatively, the signal superposition module can be a functional module within the flight control software, using programming logic to add the value of the detected precession torque disturbance signal to the value of the normal precession torque command in real time. The amplitude of the detected precession torque disturbance signal is limited to not interfering with the precession effect of the normal precession torque command, and its spectrum is within the precession response frequency band of the gyromechanical stabilization unit. The relevant details have been described in the above embodiments and will not be repeated here.

[0086] The drive output module is responsible for applying compound precession torque commands to the helicopter's gyromechanical stabilization unit (GMU). This module can be a combination of a digital-to-analog converter (DAC) and a power amplifier, converting the digitized compound commands into analog voltage or current signals to drive the servo mechanism of the GMU. For example, the drive output module can be a pulse-width modulation (PWM) controller, converting the compound commands into PWM signals to control the speed or position of the servo motor. In some implementations, the drive output module can also be a direct electro-hydraulic servo valve controller, adjusting the hydraulic oil flow according to the compound commands to control the mechanical movements of the GMU.

[0087] The response acquisition module is used to acquire the helicopter's attitude response data to compound precession torque commands. This module typically consists of an inertial measurement unit (IMU) and its associated data acquisition circuitry. The IMU may contain a three-axis gyroscope and a three-axis accelerometer to measure the helicopter's angular rate and linear acceleration. The response acquisition module can be configured to continuously acquire this sensor data at a fixed sampling rate (e.g., 100Hz or higher), converting the analog signals to digital signals via an analog-to-digital converter (ADC) before transmitting them to the flight control computer. In some implementations, the response acquisition module can also integrate attitude calculation algorithms, directly outputting attitude angle data.

[0088] The signal separation module is used to separate and extract the precession response component corresponding to the detected precession torque disturbance signal from the attitude response data. This module can be a digital filter bank; for example, by designing one or more bandpass filters, it separates the components in the attitude response data that correspond to the frequency of the detected precession torque disturbance signal. Alternatively, the signal separation module can be a correlator, which extracts the precession response component by calculating the cross-correlation function between the attitude response data and a preset detected precession torque disturbance signal. Furthermore, the signal separation module can also be implemented using an adaptive filtering algorithm (such as the Least Mean Square (LMS) algorithm) to dynamically extract the target signal from noise.

[0089] The drift identification module compares the current precession characteristic parameters with the baseline precession characteristic parameters of the gyromechanical stabilization unit in its initial state to identify precession drift. This module can be a comparator that directly calculates the difference or ratio between the current parameter and the baseline parameter. For example, it can be a software algorithm module that quantifies the drift through simple subtraction or division operations. In more complex implementations, the drift identification module can include a state estimator, such as a Kalman filter, to smooth the current precession characteristic parameters and identify significant drifts based on statistical methods.

[0090] The parameter compensation module adjusts the control parameters of the normal precession torque command based on the precession characteristic drift. This module can be a lookup table that pre-stores the control parameter adjustment values ​​corresponding to different drift amounts. For example, when the drift identification module outputs a certain drift value, the parameter compensation module retrieves the corresponding compensation parameter from the lookup table. Alternatively, the parameter compensation module can be an adaptive controller, such as a controller based on fuzzy logic or a neural network, which dynamically calculates and adjusts the control parameters (such as feedforward phase and feedback gain) based on the real-time precession characteristic drift.

[0091] The command generation module generates and outputs attitude control commands based on the adjusted control parameters. This module is the core of the flight control system and can be a proportional-integral-derivative (PID) controller whose parameters (Kp, Ki, Kd) are adjusted by the parameter compensation module. The command generation module calculates the final attitude control command based on the adjusted PID parameters and the current attitude error. In some advanced implementations, the command generation module can be a model predictive controller (MPC) or a low-level quality control (LQR) controller, using an adjusted system model to generate the optimal attitude control command. This command is then sent to the helicopter's actuators, such as servo motors, to achieve precise attitude control of the helicopter.

[0092] The gyromechanical stabilization-based unmanned helicopter attitude control system proposed in this application aims to solve the problem of difficulty in real-time identification and compensation of performance drift of the gyromechanical stabilization unit during long-term operation in traditional unmanned helicopters performing high-precision missions. Traditional systems often use fixed hardware configurations and control parameters, lacking the ability to perceive and adapt to changes in the performance of the gyromechanical stabilization unit itself. When the precession characteristics of the gyromechanical stabilization unit drift due to factors such as uneven wear of the support bearings, or small gaps or misalignments in the mechanical linkage system, traditional systems cannot adjust in time, resulting in decreased attitude control accuracy and potentially even flight instability, seriously threatening flight safety and mission execution.

[0093] The above description is merely an embodiment of this application and is not intended to limit the scope of protection of this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of protection of this application.

Claims

1. A method for unmanned helicopter attitude control based on gyro-mechanical stabilization, characterized in that, include: When the helicopter is in autonomous flight without pilot intervention, the detected precession torque disturbance signal is superimposed with the normal precession torque command to generate a composite precession torque command, which is then applied to the helicopter's gyromechanical stabilization unit. The gyromechanical stabilization unit generates attitude stabilizing torque through gyro precession effect. The amplitude of the detected precession torque disturbance signal is limited to not interfere with the precession effect of the normal precession torque command, and the spectrum of the detected precession torque disturbance signal is within the precession response frequency band of the gyromechanical stabilization unit. Acquire the attitude response data of the helicopter to the compound precession torque command; Separate and extract the precession response component corresponding to the detected precession torque disturbance signal from the attitude response data; wherein, the precession response component includes the current precession characteristic parameters of the gyro mechanical stabilization unit; The current precession characteristic parameters are compared with the reference precession characteristic parameters of the gyro mechanical stabilization unit in the initial state to identify the precession characteristic drift generated by the gyro mechanical stabilization unit during long-term operation. The control parameters of the normal precession torque command are adjusted according to the precession characteristic drift to compensate for the impact of the precession characteristic drift on the accuracy of autonomous attitude control. Generate attitude control commands based on the adjusted control parameters and output them.

2. The unmanned helicopter attitude control method based on gyromechanical stabilization according to claim 1, characterized in that, The detected precession torque disturbance signal contains multiple frequency components, which cover multiple sub-bands within the precession response frequency band of the gyroscope mechanical stabilization unit; the detected precession torque disturbance signal is one of a swept frequency signal, a pseudo-random sequence, or a spread spectrum signal; the detected precession torque disturbance signal is superimposed on the normal precession torque command by an adder.

3. The unmanned helicopter attitude control method based on gyromechanical stabilization according to claim 1, characterized in that, The step of separating and extracting the precession response component corresponding to the detected precession torque disturbance signal from the attitude response data includes: The attitude response data is subjected to frequency domain transformation, correlation detection, or matched filtering to separate the in-phase and quadrature components corresponding to the detected precession torque disturbance signal. The current precession characteristic parameters of the gyroscope mechanical stabilization unit are calculated based on the in-phase component and the quadrature component; wherein the precession characteristic parameters include precession response delay and precession response gain.

4. The unmanned helicopter attitude control method based on gyro-mechanical stabilization according to claim 1, characterized in that, The step of identifying the precession characteristic drift generated by the gyroscope mechanical stabilization unit during long-term operation includes: A recursive estimation method with a forgetting factor is used to estimate the current value and trend of the precession characteristic drift based on the current and historical precession characteristic parameters.

5. The unmanned helicopter attitude control method based on gyromechanical stabilization according to claim 1, characterized in that, The step of adjusting the control parameters of the normal precession torque command according to the precession characteristic drift includes: Adjust the feedforward phase of the normal precession torque command according to the precession response delay change corresponding to the precession characteristic drift; The feedback gain of the normal precession torque command is adjusted according to the change in precession response gain corresponding to the precession characteristic drift.

6. The unmanned helicopter attitude control method based on gyromechanical stabilization according to claim 1, characterized in that, Before the step of identifying the precession characteristic drift generated by the gyroscope mechanical stabilization unit during long-term operation, the method further includes: Obtain the current temperature and vibration parameters; Based on the temperature parameters and vibration parameters, as well as the preset temperature-precession characteristic mapping model and vibration-precession characteristic mapping model, the reference precession characteristic parameters are corrected for temperature and vibration respectively to obtain the corrected reference precession characteristic parameters corresponding to the current temperature parameters and vibration parameters.

7. The unmanned helicopter attitude control method based on gyromechanical stabilization according to claim 1, characterized in that, After the step of generating and outputting attitude control commands based on the adjusted control parameters, the method further includes: Obtain the actual attitude data of the helicopter after it executes the attitude control command; The actual attitude data is compared with the expected attitude data to calculate the attitude control deviation; When the attitude control deviation exceeds the preset deviation threshold, the step of separating and extracting the precession response component corresponding to the detected precession torque disturbance signal from the attitude response data is re-executed.

8. The unmanned helicopter attitude control method based on gyromechanical stabilization according to claim 1, characterized in that, The detection of the precession torque disturbance signal is applied during at least one period of time when the helicopter is hovering, cruising at low speed, or within a preset detection window.

9. The unmanned helicopter attitude control method based on gyromechanical stabilization according to claim 1, characterized in that, Also includes: Based on the changing trend of the precession characteristic drift and the preset drift-health score correspondence, the current health score of the gyroscope mechanical stabilization unit is calculated. When the current health score is lower than the preset health threshold, a maintenance prompt message is sent to the ground control station via the airborne communication link.

10. A helicopter unmanned attitude control system based on gyro-mechanical stabilization, characterized in that, include: The signal superposition module is used to superimpose the detected precession torque disturbance signal with the normal precession torque command to generate a composite precession torque command; wherein, the amplitude of the detected precession torque disturbance signal is limited to not interfere with the precession effect of the normal precession torque command, and the spectrum of the detected precession torque disturbance signal is within the precession response frequency band of the gyro mechanical stabilization unit; A drive output module is used to apply the compound precession torque command to the gyro-mechanical stabilization unit of the helicopter; The response acquisition module is used to acquire the attitude response data of the helicopter in response to the compound precession torque command; A signal separation module is used to separate and extract the precession response component corresponding to the detected precession torque disturbance signal from the attitude response data; wherein the precession response component includes the current precession characteristic parameters of the gyro mechanical stabilization unit; The drift identification module is used to compare the current precession characteristic parameters with the reference precession characteristic parameters of the gyro mechanical stabilization unit in the initial state, and to identify the precession characteristic drift generated by the gyro mechanical stabilization unit during long-term operation. The parameter compensation module is used to adjust the control parameters of the normal precession torque command according to the precession characteristic drift, so as to compensate for the impact of the precession characteristic drift on the accuracy of autonomous attitude control. The instruction generation module is used to generate attitude control instructions based on the adjusted control parameters and output them.

Images