Helicopter ANFIS self-adaptive vibration isolation control method suitable for plateau environment working condition

Abstract

The invention discloses a helicopter ANFIS self-adaptive vibration isolation control method suitable for a plateau environment working condition, and relates to the technical field of vibration isolation control of rotor aircrafts such as helicopters, rotorcrafts and unmanned aerial vehicles, and the vibration isolation control method is based on helicopter fuselage vibration spectrum characteristic data of different altitudes (1600m, 2500m, 3200m, 3800m, 4500m and the like) of the plateau environment working condition. The method comprises the following steps of: constructing an adaptive fuzzy neural network control system (ANFIS) model according to the change of a working condition, realizing adaptive adjustment of control parameters along with the change of the working condition, and improving the stability and the universality of a vibration isolation effect; and secondly, a vibration response closed-loop control method is provided for a head-up display system (Head-up Display, Hud) additionally installed in a helicopter cockpit, and structural vibration caused by the complex aerodynamic environment of the plateau working condition is suppressed in real time through an ANFIS training nonlinear mapping network.

Description

Technical Field

[0001] This invention relates to the field of vibration isolation and control technology for rotorcraft such as helicopters, rotorcraft, and unmanned aerial vehicles, specifically to an ANFIS adaptive vibration isolation control method for helicopters suitable for high-altitude environmental conditions. Background Technology

[0002] Helicopter vibration isolation control technology mainly includes two categories: passive vibration isolation and traditional active vibration isolation. Passive vibration isolation relies on structures such as rubber vibration dampers and hydraulic dampers to absorb some vibration energy by changing stiffness and damping parameters. However, in high-altitude environments, these vibration isolation methods are affected by reduced air density, increased rotor load fluctuations, and changes in power output, leading to parameter mismatch in the vibration dampers. This significantly reduces the vibration reduction effect of passive vibration dampers, making it difficult to adapt to the complex environmental excitation effects at different altitudes.

[0003] While active vibration isolation technology detects vibration signals and performs real-time control using sensors, its control algorithms often employ fixed-parameter proportional-integral-differential (PID) controllers or linear quadratic (LQR) controllers. These controller algorithms are highly dependent on the system model. In high-altitude environments with low air pressure, low temperatures, and nonlinear aerodynamic coupling, it is difficult to accurately obtain model parameters, leading to control hysteresis, over-response, or energy amplification, resulting in unstable vibration isolation performance.

[0004] To address the aforementioned technical bottlenecks, existing improvement schemes based on fuzzy control or neural networks, while enhancing the adaptive capabilities of control systems to some extent, are largely limited to verification under single altitude or laboratory simulation conditions. They lack systematic adaptation analysis for different altitude environments on plateaus and fail to fully consider the imaging stability requirements of helicopter head-up displays (HUDs) under high-frequency vibration conditions. Therefore, traditional improvement schemes still cannot achieve continuous adaptive vibration isolation control under nonlinear vibration environments at different altitudes on plateaus.

[0005] In view of this, the present invention proposes an ANFIS adaptive vibration isolation control method for helicopters suitable for high-altitude environmental conditions, so as to effectively solve the above-mentioned problems existing in the prior art. Summary of the Invention

[0006] In view of this, the technical problem to be solved by the present invention is to propose an ANFIS adaptive vibration isolation control method for helicopters suitable for high-altitude environmental conditions, so as to solve the problem that the vibration amplitude of helicopter fuselage is significantly increased in high-altitude environments, which seriously affects flight safety and the stability of the added structure or system. At the same time, it addresses the technical defects of traditional vibration isolation control algorithms, which have a significant decrease in vibration isolation effect under the complex aerodynamic and dynamic coupling effect in high-altitude conditions.

[0007] To achieve the above objectives, the present invention provides the following technical solution: an ANFIS adaptive vibration isolation control method for helicopters suitable for high-altitude environmental conditions, comprising: S1. On the VehicleSim–Simulink co-simulation platform, a helicopter fuselage dynamics model and a head-up display system (HUD) vibration response model are established. The helicopter fuselage vibration data at altitudes of 1000-5000 meters are obtained through simulation calculations. The vibration data includes Y-axis and Z-axis vibration signals. S2, perform FFT processing on the Y-axis vibration signal obtained in S1, and extract the Y-axis principal vibration frequency as... Combined with the Z-axis principal vibration frequency is A mapping information table is constructed through an iterative optimization process. This iterative optimization process aims to minimize the camera's offset motion near its equilibrium position, and defines an evaluation criterion function. and The damping coefficient C value is adjusted in combination to find the optimal value; S3. Construct an adaptive fuzzy neural network control system (ANFIS). The ANFIS consists of a 5-layer structure, with the vibration frequency vector as the input variable. The output is the damping coefficient C. Each input variable uses three generalized Bell membership functions. The parameters are updated through a hybrid learning algorithm. In the forward propagation stage, the hybrid learning algorithm uses the least squares method to estimate the linear parameters of the output layer, and in the back propagation stage, it uses the gradient descent method to update the center and width parameters of the membership functions. S4. An FFT module is set between the helicopter vibration signal and the adaptive fuzzy neural network control system (ANFIS) to extract the spectral components of the Y-axis and Z-axis vibration signals in real time and input them into the trained adaptive fuzzy neural network control system (ANFIS). The adaptive fuzzy neural network control system (ANFIS) outputs the optimal damping value C in each fixed sampling period and inputs the damping value into the semi-active control system connecting the fuselage and the dynamic platform base. The semi-active control system uses a spring damper system with a variable damping coefficient C(t) as the vibration isolation carrier. S5, the vertical vibration displacement of the Hud system The input parameters are incorporated into the Adaptive Fuzzy Neural Network Control System (ANFIS), combined with the root mean square value of the fuselage acceleration. Principal vibration frequency Energy ratio index of 25Hz band and the energy ratio index of the 12.62Hz frequency band This enables real-time vibration isolation control of the Hud system's display imaging process.

[0008] Preferably, the 5-layer structure of the Adaptive Fuzzy Neural Network Control System (ANFIS) in S3 specifically includes: Level 1: Each node outputs the membership degree value of the generalized Bell membership function corresponding to the input variable. The specific expression is as follows:

[0009]

[0010]

[0011] In the formula, This represents the membership degree value of the membership function corresponding to the input variable. , b , c These are respectively represented as the width, slope, and center parameter of the membership function; This is expressed as the principal vibration frequency along the Y-axis; This is expressed as the Z-axis principal vibration frequency; Layer 2: Apply fuzzy operators to multiply the input signals to obtain the weight of each rule. The specific expression is as follows:

[0012]

[0013] In the formula, Let be the excitation intensity of the i-th fuzzy rule; Represented as input variables The corresponding i-th fuzzy subset; i represents the index of the fuzzy rule or fuzzy subset; Layer 3: The implication method is used to calculate the standard weights, and the specific expression is as follows: , i=1,…,n(3) In the formula, Represented as the first The incentive intensity of a fuzzy rule; This represents the normalized standard weights.

[0014] Layer 4: Fuzzy set aggregation is achieved through fuzzy if-then rules of the Takagi-Sugeno type, and the output expression is:

[0015] In the formula, This represents the output level of a Takagi-Sugeno type fuzzy rule. It is typically a linear combination of the input variables used to calculate the final damping control quantity; in The fuzzy corresponding to the three Takagi-Sugeno types The rule, specifically the formula expression, is as follows:

[0016]

[0017]

[0018] In the formula, p i q i r i Represented as a result parameter; Layer 5: Defuzzification is achieved through weighted averaging, and the damping coefficient C is output. The specific expression is as follows:

[0019] In the formula, N represents the total number of nodes in the 3rd layer; it is used to average all weighted outputs when calculating the final output damping coefficient C.

[0020] Preferably, the membership function described in S3 is the generalized Bell membership function, and its specific mathematical expression is as follows:

[0021]

[0022]

[0023] In the formula, This represents the membership degree value of the membership function corresponding to the input variable. , b , c These are respectively represented as the width, slope, and center parameter of the membership function; This is expressed as the principal vibration frequency along the Y-axis; This is expressed as the Z-axis principal vibration frequency; The output of the adaptive fuzzy neural network control system (ANFIS) described in S3 is the damping control command. It uses a Sugeno-type linear output function, with the expression:

[0024] In the formula, These are linear parameters, automatically calculated using the least squares method during training.

[0025] Preferably, the parameter update rule described in S3 is expressed as follows:

[0026]

[0027] In the formula, For learning rate, Let be the error function. It is represented as a partial derivative operator and is used in the gradient descent method to calculate the gradient of the error function with respect to the parameters; t This represents the number of iterations for parameter updates; This is represented as the width parameter of the membership function; Preferably, the semi-active control system described in S4 provides the optimal damping level for different vibration conditions by adjusting the damping coefficient C(t) in real time, so as to suppress the Y-axis and Z-axis vibrations of the camera in Hud.

[0028] Preferably, the input parameters of the adaptive fuzzy neural network control system (ANFIS) described in S5 comprehensively characterize the vibration intensity, main frequency characteristics and Hud jitter state of the helicopter in different altitude environments on the plateau, so as to realize the adaptive adjustment of control parameters as the operating conditions change.

[0029] Compared with existing technologies, the ANFIS adaptive vibration isolation control method for helicopters suitable for high-altitude environments provided by this invention has the following advantages: 1. Achieves adaptive vibration isolation control at different altitudes on plateaus: Based on the vibration spectrum characteristics of helicopter fuselages at different altitudes (1600m, 2500m, 3200m, 3800m, 4500m) on plateaus, this invention constructs diverse training samples, enabling the ANFIS controller to learn the nonlinear dynamic characteristics caused by changes in altitude, and achieve adaptive adjustment of control parameters, effectively overcoming the problem of performance degradation of traditional control algorithms at different altitudes.

[0030] 2. Enhance the robustness of vibration isolation systems to nonlinear aerodynamics and coupled vibrations: By introducing fuzzy reasoning and neural network learning mechanisms, this invention can automatically identify changes in system characteristics in high-altitude environments with low air pressure, low-density air, and fluctuating aerodynamic loads, thereby achieving dynamic compensation for nonlinear vibrations and significantly improving the stability and anti-interference capability of vibration isolation control.

[0031] 3. Improve the safety of helicopter flight at high altitudes and the imaging stability of the added Hud: This invention not only effectively controls the vibration of the helicopter fuselage, but also incorporates the vibration response of the added Hud into the feedback loop, realizing real-time vibration isolation of the Hud display imaging process, improving the clarity and reliability of the pilot's visual information, and ensuring flight safety under complex conditions at high altitudes.

[0032] 4. Achieving stable vibration isolation performance in high-altitude environments through online optimization and adjustment of control parameters: This invention utilizes fuzzy inference and neural network learning mechanisms to construct an ANFIS adaptive controller, reducing the dependence on precise dynamic models during real-time control and exhibiting stronger robustness and environmental adaptability. By training with vibration spectrum samples at multiple altitudes, online optimization and dynamic adjustment of control parameters are achieved, ensuring the stability of the system's vibration isolation performance under complex high-altitude environmental conditions. Attached Figure Description

[0033] Figure 1 This is a schematic diagram of the ANFIS adaptive vibration isolation control method for helicopters applicable to high-altitude environments according to the present invention. Figure 2 This is a schematic diagram of the architecture of the Adaptive Fuzzy Neural Network Control System (ANFIS) of the present invention; Figure 3 This is a schematic diagram of ANFIS training at different altitudes according to an embodiment of the present invention; Figure 4 This is a schematic diagram of the vibration spectrum of a helicopter fuselage at different altitudes under plateau environmental conditions, according to an embodiment of the present invention. Figure 5 This is a schematic diagram illustrating the impact of helicopter vibration on the fuselage at different altitudes in a high-altitude environment, according to an embodiment of the present invention. Figure 6 This is a schematic diagram illustrating the impact of helicopter vibration at different altitudes on the Hud fixed camera in a high-altitude environment, according to an embodiment of the present invention. in, Figures 1 to 5 In the diagram, a, b, c, d, and e represent height test diagrams for 1600m, 2500m, 3200m, 3800m, and 4500m, respectively. Detailed Implementation

[0034] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present invention.

[0035] The present invention will now be described in further detail with reference to the accompanying drawings and embodiments.

[0036] For an example, please refer to... Figures 1 to 6 As shown: To address the problems mentioned in the technical solutions, this application provides an ANFIS adaptive vibration isolation control method for helicopters suitable for high-altitude environments. This invention establishes a helicopter fuselage dynamics model and a Hud system vibration response model on the VehicleSim–Simulink co-simulation platform. Through simulation calculations, helicopter fuselage vibration data at different altitudes are obtained: 1600 m, 2500 m, 3200 m, 3800 m, and 4500 m. Based on this, an ANFIS controller is constructed to achieve effective vibration isolation control of the helicopter fuselage.

[0037] The semi-active vibration isolation architecture is described in detail below: A spring damper system with a variable damping coefficient C(t) is used as a vibration isolation carrier to reduce the vibration experienced by the camera mounted on the helicopter's HUD. The time-dependent nature of this damping coefficient enables adaptive control, allowing it to dynamically adjust based on vibrations sensed by the helicopter platform. The control system is responsible for real-time adjustment. C The value provides the optimal damping level for different vibration conditions. The primary focus is on the vibration frequency parameter, as the vibrations generated by the fuselage are mainly in the X and Z directions, affecting the Hud along the Y and Z axes. An adaptive system continuously monitors the disturbance frequencies generated by the helicopter in these two axes and calculates the optimal value for each moment. C Due to the inertial coupling and vibration propagation mechanism of the fuselage structure, secondary vibration components will be generated in the Y-axis direction under the main excitation of the Z-axis. In the simulation, the Y-axis response signal is extracted through a multibody dynamics model and then processed by FFT to extract the main frequency. , as the input to the ANFIS control system.

[0038] Given the damping C With frequency and The relationships between them require the establishment of a mapping information table, which is used to map relationships based on a given set of conditions. and Combined methods allow for quick lookup of the optimal damping value. The construction of this information table requires an iterative optimization process, i.e., for... and Continuously adjust all possible combinations C The optimal value is sought. To this end, an evaluation criterion function is defined to measure the optimality of the damping effect in each case, with the core objective of maximizing control over the camera's offset motion near the equilibrium position. The degree of undesirable camera displacement can be quantified by the standard deviation of its position. This criterion function guides the optimization...C The choice of values ​​minimizes camera vibration under the corresponding frequency combinations. After iterative optimization, a reference table is generated as a result of the standard function minimization process, which can efficiently optimize any specific... and Frequency combinations provide the optimal matching damping coefficient. C .

[0039] The vibration isolation control method is constructed as follows: The core objective of the system design is to provide the optimal damping coefficient in real time for different vibration states during helicopter flight, based on a pre-set information table. C Given that this objective requires a data-driven modeling approach, ANFIS was chosen as the solution. The ANFIS method integrates the adaptive learning capabilities of neural networks, that is, it learns and models using a given input / output sample dataset, where the input dataset is a vector of vibration frequencies. , The corresponding target output is the damping coefficient. C ANFIS architecture such as Figure 1 It consists of a 5-layer structure, with each input variable employing 3 generalized Bell membership functions. and (i=1, 2, 3), and 1 output. C .like Figure 2 As shown; The ANFIS model building process involves three steps: First, collecting the input / output datasets required for model training; then, establishing the initial fuzzy inference system structure; and finally, applying the backpropagation gradient learning algorithm for parameter optimization. The reference table obtained through the aforementioned iterative process constitutes the basic dataset required for training. The trained ANFIS system will be integrated into the simulation environment as a Simulink module. This is to meet the module's requirements for vibration frequency input. and To meet the requirements, an FFT module was added between the helicopter vibration signal and the ANFIS module. This module uses Fast Fourier Transform to extract the spectral components of the Y-axis and Z-axis vibration signals in real time. After analyzing the input frequency, ANFIS outputs an optimal damping value at each fixed sampling period. C This value is then input to the semi-active control system that connects the fuselage to the dynamic platform base.

[0040] Step 1: This paper defines three generalized Bell membership functions for each input and determines their parameters. In the first layer, the output of each node is the membership degree value (a, b, c) of the membership function corresponding to the input variable:

[0041]

[0042]

[0043] The second layer applies fuzzy operators, multiplying the input signals, and the result represents the weight of each rule.

[0044] ,

[0045] The third layer uses the implication method, where the output of each node corresponds to a standard weight, which is a number between 0 and 1.

[0046] , i=1,…,n(3) Aggregation occurs at the fourth layer, where fuzzy sets representing the outputs of each rule are combined into a single fuzzy set.

[0047]

[0048] in Corresponding to three Takagi-Sugeno type fuzziness The rules are:

[0049]

[0050]

[0051] Where, p i q i r i It is the result parameter.

[0052] Finally, the fifth layer applies a deblurring process, which involves summing all the outputs from the fourth layer to obtain the value corresponding to the damping value as the output. This process is achieved through a weighted average, as shown in the following formula:

[0053] Damping coefficient C As system output, it is produced by the output levels of each rule. Rather than normalized weights The decision is made jointly, where N represents the total number of nodes in the 3rd layer.

[0054] ANFIS training specifically includes the following: First, membership functions; in ANFIS control systems, the design of membership functions (MF) directly affects the expressive power of fuzzy inference and the adaptive performance of the system. To achieve effective identification and vibration isolation control of fuselage vibration characteristics at different altitudes on plateaus, this study designed reasonable membership function forms and distributions for both input and output variables.

[0055] The determination and physical meaning of the input variables in this scheme are as follows: The input variables for the ANFIS controller are selected from key parameters that reflect the aircraft's vibration characteristics and changes in the high-altitude environment, including: : Root mean square value of fuselage acceleration; : Principal vibration frequency (obtained through fast Fourier transform); Energy ratio in the 25 Hz band; Energy ratio in the 12.62 Hz band; Vertical vibration displacement of the Hud system.

[0056] These input quantities comprehensively characterize the vibration intensity, dominant frequency characteristics, and vibration state of the helicopter at different altitudes in the high-altitude environment, providing the fuzzy system with physically meaningful feature inputs.

[0057] The membership function type can be selected as follows: To ensure the smoothness and differentiability of the fuzzy partitioning, the generalized Bell membership function is selected, with the mathematical expression as follows:

[0058]

[0059]

[0060] In the formula, This represents the membership degree value of the membership function corresponding to the input variable. , b , c These are represented as the width, slope, and center parameter of the membership function, respectively. This is expressed as the principal vibration frequency along the Y-axis; This is expressed as the Z-axis principal vibration frequency; Number and distribution of membership functions; To balance system accuracy and computational complexity, each input variable is divided into three fuzzy subsets, representing the three fuzzy states of signal strength, as shown in Table 1: Table 1 shows fuzzy subsets of the variables;

[0061] The initial center and width parameters of each fuzzy subset are set based on the sample statistics and are automatically optimized and adjusted by ANFIS in the subsequent training phase.

[0062] The specific process of the membership function of the output variable is as follows: ANFIS output is a damping control command. Its value is mapped to the damping adjustment coefficient of the semi-active vibration isolator. The output section adopts a Sugeno-type output in the form of a linear function, i.e.:

[0063] in, These are linear parameters, automatically calculated using the least squares method during training.

[0064] The specific engineering process for parameter learning and optimization is as follows: The ANFIS controller updates parameters during the training phase using a hybrid learning algorithm: the least squares method is used to estimate linear parameters (output layer parameters) during the forward propagation phase; and gradient descent is used to update nonlinear parameters (centers of membership functions) during the backpropagation phase. With width The update pattern can be represented as:

[0065] in, For learning rate, For the error function:

[0066] Through repeated iterative training, the membership function is continuously adjusted under sample conditions at different plateau altitudes, ultimately forming a fuzzy partition that can adapt to environmental changes.

[0067] The evolution of membership functions at different heights is as follows: like Figure 3 As shown, after training, it can be observed that the membership functions of each input variable exhibit dynamic migration at different altitudes: as altitude increases, The "High" fuzzy set shifts to a higher interval, reflecting that the intensity of fuselage vibration increases as air pressure decreases; and The fuzzy range has expanded slightly, reflecting the diffusion trend of the main frequency energy; The reduced overlap between the "Medium" and "Large" ranges indicates that the ANFIS controller is more sensitive to Hud jitter.

[0068] The effects and comparisons of different vibration control methods are analyzed as follows: like Figure 4 As shown, when helicopters operate in high-altitude environments, the impact of fuselage vibration on the Hud installed in the helicopter cockpit varies significantly depending on whether no control is implemented, conventional control methods are used, or the ANFIS method is employed. Figure 3 This represents the vibration spectrum at different altitudes under high-altitude operating conditions. From... Figure 4 It can be seen that the main peak frequency of the vibration during helicopter flight from 1600 meters to 4500 meters is 25Hz. The resulting vibration amplitude originates from the modal coupling effect of the system structure. Although this frequency is not the main frequency of the excitation source, its energy intensity is significant, and control measures must be designed.

[0069] Figure 5 This study analyzes the impact of helicopter vibration on the fuselage at different altitudes in a high-altitude environment. Figure 5 It can be seen that the vibration amplitude of the helicopter on the fuselage is the largest under uncontrolled conditions, the vibration amplitude of the helicopter on the fuselage is reduced under conventional control measures, and the vibration amplitude of the helicopter on the fuselage is the smallest under ANFIS control.

[0070] Figure 6 This study examines the impact of helicopter vibration at different altitudes in high-altitude environments on the fixed camera of the added Hud (Head-Up Display). From altitudes ranging from 1600 meters to 4500 meters, the vibration isolation effect shows that under ANFIS control, the impact of helicopter vibration on the fixed camera of the added Hud is significantly reduced. This indicates that the ANFIS control method is most effective at isolating helicopter vibrations during high-altitude flight and is highly effective in isolating vibrations at any altitude during high-altitude operations.

[0071] In summary, this invention constructs an Adaptive-Network-based Fuzzy Inference Systems (ANFIS) model based on the helicopter fuselage vibration spectrum characteristic data at different altitudes (1600m, 2500m, 3200m, 3800m, 4500m, etc.) under high-altitude environmental conditions. This model enables adaptive adjustment of control parameters as the operating conditions change, improving the stability and universality of vibration isolation. Simultaneously, a vibration response closed-loop control method is proposed for the head-up display (HUD) system installed in the helicopter cockpit. By training a nonlinear mapping network using ANFIS, real-time suppression of structural vibrations caused by the complex aerodynamic environment at high altitudes is achieved.

[0072] Please refer to the above work process. Figures 1 to 6 .

[0073] It should be noted that, in this document, relational terms such as "first" and "second" are used merely to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0074] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and variations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A helicopter ANFIS adaptive vibration isolation control method suitable for high-altitude environment conditions, characterized in that, include: S1. On the VehicleSim–Simulink co-simulation platform, a helicopter fuselage dynamics model and a head-up display system (HUD) vibration response model are established. The helicopter fuselage vibration data at altitudes of 1000-5000 meters are obtained through simulation calculations. The vibration data includes Y-axis and Z-axis vibration signals. S2, perform FFT processing on the Y-axis vibration signal obtained in S1, and extract the Y-axis principal vibration frequency as... Combined with the Z-axis principal vibration frequency is A mapping information table is constructed through an iterative optimization process. This iterative optimization process aims to minimize the camera's offset motion near its equilibrium position, and defines an evaluation criterion function. and The damping coefficient C value is adjusted in combination to find the optimal value; S3. Construct an adaptive fuzzy neural network control system. The adaptive fuzzy neural network control system includes a 5-layer structure, and the input variable is the vibration frequency vector. The output is the damping coefficient C. Each input variable uses three generalized Bell membership functions. The parameters are updated through a hybrid learning algorithm. In the forward propagation stage, the hybrid learning algorithm uses the least squares method to estimate the linear parameters of the output layer, and in the back propagation stage, it uses the gradient descent method to update the center and width parameters of the membership functions. S4. An FFT module is set between the helicopter vibration signal and the Adaptive Fuzzy Neural Network Control System (ANFIS) to extract the spectral components of the Y-axis and Z-axis vibration signals in real time and input them into the trained ANFIS. The ANFIS outputs the optimal damping value C at each fixed sampling period, and inputs this damping value into the semi-active control system connecting the fuselage and the dynamic platform base. The semi-active control system uses a variable damping coefficient. C (t) spring damper system as vibration isolation carrier; S5, the vertical vibration displacement of the Hud system The input parameters are incorporated into the Adaptive Fuzzy Neural Network Control System (ANFIS), combined with the root mean square value of the fuselage acceleration. Principal vibration frequency Energy ratio index of 25Hz band and the energy ratio index of the 12.62Hz frequency band This enables real-time vibration isolation control of the Hud system's display imaging process.

2. The ANFIS adaptive vibration isolation control method for helicopters suitable for high-altitude environmental conditions according to claim 1, characterized in that, The five-layer structure of the adaptive fuzzy neural network control system in S3 specifically includes: Level 1: Each node outputs the membership degree value of the generalized Bell membership function corresponding to the input variable. The specific expression is as follows: In the formula, This represents the membership degree value of the membership function corresponding to the input variable. , b , c These are represented as the width, slope, and center parameter of the membership function, respectively. This is expressed as the principal vibration frequency along the Y-axis; This is expressed as the Z-axis principal vibration frequency; Layer 2: Apply fuzzy operators to multiply the input signals to obtain the weight of each rule. The specific expression is as follows: In the formula, Represented as the first The incentive intensity of a fuzzy rule; Represented as input variables The corresponding number A fuzzy subset; i represents the index of the fuzzy rule or fuzzy subset; Layer 3: The implication method is used to calculate the standard weights, and the specific expression is as follows: ,i=1,…,n(3) In the formula, Indicates the first The incentive intensity of a fuzzy rule; This represents the normalized standard weights.

3. Fourth layer: Fuzzy set aggregation is achieved through fuzzy if-then rules of the Takagi-Sugeno type, and the output expression is: In the formula, This represents the output level of a Takagi-Sugeno type fuzzy rule. It is typically a linear combination of the input variables used to calculate the final damping control quantity; in The fuzzy corresponding to the three Takagi-Sugeno types The rule, specifically the formula expression, is as follows: In the formula, p i q i r i Represented as a result parameter; Layer 5: Defuzzification is achieved through weighted averaging, and the damping coefficient C is output. The specific expression is as follows: In the formula, N represents the total number of nodes in the 3rd layer; it is used to average all weighted outputs when calculating the final output damping coefficient C.

4. The ANFIS adaptive vibration isolation control method for helicopters suitable for high-altitude environmental conditions according to claim 1, characterized in that, The membership function described in S3 is the generalized Bell membership function, and its specific mathematical expression is as follows: In the formula, This represents the membership degree value of the membership function corresponding to the input variable. , b , c These are represented as the width, slope, and center parameter of the membership function, respectively. This is expressed as the principal vibration frequency along the Y-axis; It is represented as the Z-axis principal vibration frequency.

5. The ANFIS adaptive vibration isolation control method for helicopters suitable for high-altitude environmental conditions as described in claim x, characterized in that, The output of the adaptive fuzzy neural network control system described in S3 is the damping control command. It uses a Sugeno-type linear output function, with the expression: In the formula, These are linear parameters, automatically calculated using the least squares method during training.

6. The ANFIS adaptive vibration isolation control method for helicopters suitable for high-altitude environmental conditions according to claim 1, characterized in that, The specific expression for the parameter update rule described in S3 is: In the formula, For learning rate, For the error function, It is represented as a partial derivative operator and is used to calculate the gradient of the error function E with respect to the parameters; t This represents the number of iterations for parameter updates; This is represented as the width parameter of the membership function.

7. The ANFIS adaptive vibration isolation control method for helicopters suitable for high-altitude environmental conditions according to claim 1, characterized in that, The semi-active control system described in S4 provides the optimal damping level for different vibration conditions by adjusting the damping coefficient C(t) in real time, so as to suppress the Y-axis and Z-axis vibrations of Hud's camera.

8. The ANFIS adaptive vibration isolation control method for helicopters suitable for high-altitude environmental conditions according to claim 1, characterized in that, The input parameters of the adaptive fuzzy neural network control system described in S5 comprehensively characterize the vibration intensity, main frequency characteristics, and Hud jitter state of the helicopter in different altitude environments on the plateau, so as to realize the adaptive adjustment of control parameters as the operating conditions change.

Images