Automatic landing control method for shipborne fixed-wing aircraft under strong turbulent flow condition

Abstract

The invention discloses an automatic landing control method for a shipborne fixed-wing aircraft under a strong turbulent flow condition, and belongs to the field of landing of shipborne fixed-wing aircrafts. The problem that high-precision and high-robustness control is difficult to achieve under the strong turbulence condition in an existing scheme is solved. The method comprises the following steps: decoupling landing control of the shipborne fixed-wing aircraft into a longitudinal link and a transverse link, calculating a longitudinal position error through a longitudinal position loop, updating parameters on line based on a neural network, and generating a longitudinal speed deviation; calculating a transverse position error and an error change rate through a transverse position ring, dynamically adjusting parameters, and generating a course angle deviation; an air turbulence compensation module is introduced to compensate the longitudinal speed deviation and generate an expected pitch angle; calculating an expected pitch angle speed operation amount; calculating a yaw angle error to generate an expected roll angle and an expected roll angle speed operation amount; and mapping the expected pitch angular velocity operation quantity and the expected roll angular velocity operation quantity through a distributor, and driving the shipborne fixed-wing aircraft to complete landing. The method is used in marine field.

Description

TECHNICAL FIELD

[0001] The present application belongs to the field of shipborne fixed-wing aircraft landing, and particularly relates to a shipborne fixed-wing aircraft automatic landing control method under strong turbulent flow conditions. BACKGROUND

[0002] The automatic landing task of a shipborne fixed-wing aircraft on a ship deck is a highly challenging part of aviation operations, mainly due to various environmental constraints and system requirements. The ship deck space is extremely limited, and as a moving and unstable platform, its dynamic displacement and swing significantly increase the landing difficulty; at the same time, the sea environment has complex and variable wind field interference, such as strong turbulent flow, turbulent flow and sea state changes, which lead to unstable airflow, posing a severe test to the aircraft attitude and trajectory control. In addition, the shipborne landing process has extremely strict precision requirements, and the aircraft needs to land accurately in a small area to avoid collision or deviation risks. In the traditional manual control mode, the pilot needs to adjust the aircraft thrust and rudder angle in real time under high-risk conditions, which not only puts high requirements on the pilot's skill level and psychological endurance, but also may lead to frequent operation errors in adverse sea conditions or low visibility conditions (such as fog or night), significantly increasing the accident risk. Therefore, developing a high-reliability automatic landing system has become a key technical requirement to improve safety and stability, especially in strong turbulent flow environments, the system must balance robustness and adaptability to offset the nonlinear effects of ship movement and airflow.

[0003] For the design of shipborne fixed-wing aircraft automatic landing algorithms, existing technologies have proposed various solutions, but these solutions have obvious defects in adaptability under complex environments. For example, in "Unmanned Aerial Vehicle Automatic Carrier Landing Net Recovery Technology Research" by Liu Qiang, a longitudinal attitude control and power compensation system and a lateral coordination control law are designed to try to solve the low dynamic pressure carrier landing problem. However, this solution can only guide the unmanned aerial vehicle to directly impact the recovery net at a high speed, causing the aircraft to bear a large impact load and causing significant damage to the aircraft structure, which cannot meet the requirements of smoothness and equipment protection for shipborne landing. For example, in "Research on Autonomous Landing Control Technology of Wheeled Unmanned Aerial Vehicle" by Ji Lili, a height controller outputs a pitch command to the pitch ring, and a track control outputs a roll command to the roll ring. Although this control algorithm simplifies the operation process, it lacks adaptability and cannot adjust control parameters and strategies in time according to the real-time distance between the unmanned aerial vehicle and the landing point or external environmental disturbances (such as dynamic wind field or ship swing), resulting in insufficient robustness and limited control accuracy of the system under strong turbulent flow conditions, making it difficult to respond to sudden disturbances. The core problem of these existing technologies is the fixation and staticization of control parameters, which cannot dynamically respond to disturbance changes, thus easily causing trajectory deviation or response lag in actual combat.

[0004] To sum up, the automatic landing technology of shipborne fixed-wing aircraft still faces significant gaps: the existing scheme is difficult to achieve high precision and high robustness control under strong air disturbance, and its weak adaptive ability limits the system performance in dynamic environment. Traditional manual control relies on highly skilled pilots, and the risk is uncontrollable; while the existing automatic algorithm (such as control based on fixed PID parameters) cannot effectively compensate for the time-varying disturbance of ship movement and air flow, resulting in landing deviation accumulation, structural damage or insufficient stability. Therefore, the present application aims to provide an automatic landing control method for shipborne fixed-wing aircraft under strong air disturbance to solve the problem that the existing scheme is difficult to achieve high precision and high robustness control under strong air disturbance.

[0005] To achieve the above-mentioned purpose, the present application adopts the following technical solutions: An automatic landing control method for shipborne fixed-wing aircraft under strong air disturbance, the method comprising: Decoupling the landing control of the shipborne fixed-wing aircraft into longitudinal link and lateral link, and adopting a hierarchical cascade control structure, wherein the longitudinal link includes a longitudinal position loop, a longitudinal velocity loop and a longitudinal attitude loop, and the lateral link includes a lateral position loop, a lateral attitude loop and a lateral roll angle velocity loop; In the longitudinal link, the longitudinal position error is calculated by the longitudinal position loop, and the proportional, integral and differential parameters are updated online based on the neural network to generate the longitudinal velocity deviation; In the lateral link, the lateral position error and error change rate are calculated by the lateral position loop, and the proportional, integral and differential parameters are dynamically adjusted based on fuzzy mathematics to generate the heading angle deviation; the input of the fuzzy mathematics is the lateral position error and error change rate, and the output is the proportional parameter adjustment amount, the integral parameter adjustment amount and the differential parameter adjustment amount, and the fuzzy rule base is constructed based on the fuzzy subsets of error and error change rate for parameter adjustment; An air disturbance compensation module is introduced to feed forward compensate the longitudinal velocity deviation, specifically including: querying the air disturbance intensity coefficient empirical table based on the position of the aircraft relative to the ship body landing point, calculating the actual air disturbance influence coefficient by interpolation, and superimposing it to the longitudinal velocity deviation to obtain the compensated longitudinal velocity deviation; In the longitudinal link, the compensated longitudinal velocity deviation is input into the longitudinal velocity loop to generate the expected pitch angle, and the expected pitch angle is limited and filtered to obtain the equivalent pitch angle; based on the equivalent pitch angle error, the expected pitch angle velocity operation amount is calculated by the proportional-integral controller; In the lateral link, the yaw angle error is calculated based on the heading angle deviation and the aircraft yaw angle to generate the expected roll angle; based on the expected roll angle error, the expected roll angle velocity operation amount is calculated by the proportional-integral controller; The expected pitch angle velocity operation quantity of the longitudinal link output and the expected roll angle velocity operation quantity of the lateral link output are mapped to the rudder, elevator, aileron and throttle actuators through the control surface distributor, so as to drive the fixed-wing aircraft on the ship to complete the automatic landing.

[0006] Further, a preferred mode is also proposed, the neural network adopts a two-layer hidden layer structure, the first hidden layer contains 5 neurons and uses the ReLU activation function, and the second hidden layer contains 5 neurons and uses the tanh activation function; the loss function is defined as:

[0007] Wherein, for weighing the current error, for weighing the current cumulative error, for weighing the current control amount change rate.

[0008] Further, a preferred mode is also proposed, the longitudinal velocity deviation is:

[0009] Wherein, is the longitudinal velocity deviation, is the expected vertical sinking speed, is the real-time vertical position error, is the proportional coefficient of the controller, is the integral coefficient of the controller, is the differential coefficient of the controller.

[0010] Further, a preferred mode is also proposed, the fuzzy subset of fuzzy mathematics includes negative large, negative medium, negative small, zero, positive small, positive medium and positive large, and the membership function adopts a triangular function; the fuzzy rule base is based on the following principles: when the lateral position error and the error change rate are both positive, the proportional parameter and the differential parameter adjustment amount are increased, and the integral parameter adjustment amount is reduced; when the error and the error change rate are both negative, the mirror adjustment parameter is adjusted.

[0011] Further, a preferred mode is also proposed, the air disturbance intensity coefficient empirical table contains the longitudinal distance of a plurality of sampling points and the corresponding air disturbance intensity coefficient, and the actual air disturbance influence coefficient is calculated by interpolation:

[0012] Wherein, is the position of the aircraft relative to the ship landing point, is the position of the first sampling point relative to the ship landing point, is the position of the i-th sampling point relative to the ship landing point, the position of the i+1th sampling point relative to the ship body landing point, the position of the nth sampling point relative to the ship body landing point, the coefficient corresponding to the 1st sampling point in the air disturbance influence coefficient table, the coefficient corresponding to the i th sampling point in the air disturbance influence coefficient table, the coefficient corresponding to the i+1th sampling point in the air disturbance influence coefficient table, the coefficient corresponding to the nth sampling point in the air disturbance influence coefficient table.

[0013] Further, a preferred mode is also proposed, and the calculation mode of the expected pitch angle is:

[0014] wherein, is the actual longitudinal velocity gain of the longitudinal velocity loop, is the expected longitudinal velocity gain of the longitudinal velocity loop, is the actual longitudinal velocity differential gain of the longitudinal velocity loop, is the actual longitudinal acceleration, is the actual longitudinal velocity, is the expected longitudinal velocity after considering the compensation of air disturbance.

[0015] Further, a preferred mode is also proposed, and the calculation mode of the expected roll angle is:

[0016] wherein, is the yaw angle proportional gain of the lateral inner loop, is the current time, is the yaw angle error of the aircraft at time k, is the yaw angle error of the aircraft at time k-1, is the yaw angle differential gain of the lateral inner loop, is the simulation step length.

[0017] Further, a preferred mode is also proposed, and the integral term of the proportional-integral controller is solved by Euler method:

[0018] wherein, is the integral value of the lateral deviation at the current time, is the integral value of the longitudinal deviation at the last time, is the integral value of the lateral deviation at time 0, is the integral reference gain of the lateral deviation, is the integral parameter adjustment amount of the lateral deviation, is the lateral deviation at the current time.

[0019] Based on the same inventive concept, the application further provides a computer device comprising a memory and a processor, wherein the memory stores a computer program, and when the processor runs the computer program stored in the memory, the processor executes the automatic landing control method of a ship-borne fixed-wing aircraft under strong turbulence conditions according to any one of the above.

[0020] Based on the same inventive concept, the application further provides a computer readable storage medium, wherein the computer readable storage medium stores a computer program, and when the computer program is run by a processor, the computer program executes the steps of the automatic landing control method of a ship-borne fixed-wing aircraft under strong turbulence conditions according to any one of the above.

[0021] Compared with the prior art, the application has the following beneficial effects: The application proposes an automatic landing control method of a ship-borne fixed-wing aircraft under strong turbulence conditions, which overcomes the problem of response lag of a static control strategy, and solves the risk of trajectory deviation caused by parameter solidification in the prior art.

[0022] In the automatic landing control method of a ship-borne fixed-wing aircraft under strong turbulence conditions, an air turbulence feedforward compensation module is further provided, which quantifies the disturbance influence as a calculable superimposed signal: a turbulence intensity coefficient is generated based on an empirical table interpolation, and a compensation amount is generated in combination with a real-time turbulence intensity; the compensation amount is superimposed to a longitudinal speed deviation, so that the control system actively offsets the wake sinking effect, and the response speed is improved by more than 40% compared with a traditional feedback control; a pitch angle output by a longitudinal speed loop is strictly limited and filtered by a second-order transfer function, so as to convert a sharp instruction into a smooth S curve, and avoid aileron saturation; a lateral yaw angle error is differentiated and strengthened and controlled by PI, so as to quickly suppress lateral swinging.

[0023] The shipborne fixed-wing aircraft automatic landing control method under strong disturbance conditions provided by the application has the advantages that under the condition of initial longitudinal deviation of -15 meters, lateral deviation of -17 meters and strong disturbance, the final longitudinal deviation is less than or equal to 0.09 meters, and the lateral deviation is less than or equal to 0.11 meters, and the precision is improved by one order of magnitude compared with the prior art.

[0024] The shipborne fixed-wing aircraft automatic landing control method under strong disturbance conditions provided by the application has the advantages that under the condition of initial longitudinal deviation of -15 meters, lateral deviation of -17 meters and strong disturbance, the final longitudinal deviation is less than or equal to 0.09 meters, and the lateral deviation is less than or equal to 0.11 meters, and the precision is improved by one order of magnitude compared with the prior art.

[0025] The application is applied to the field of oceans. DETAILED DESCRIPTION

[0026] The accompanying drawings, which form a part of the present application, are intended to provide further understanding of the present application, and the illustrative embodiments of the present application and their description serve the purpose of explaining the present application. The accompanying drawings should not be regarded as a limitation of the present application. In the drawings: Figure 1 The flow chart of the shipborne fixed-wing aircraft automatic landing control method under strong disturbance conditions described by the application; Figure 2 The control block diagram of the shipborne fixed-wing aircraft described by the application; Figure 3 The schematic diagram of the deviation membership function described by the application; Figure 4 The schematic diagram of the deviation rate membership function described by the application; Figure 5 The schematic diagram of the proportional coefficient membership function described by the application; Figure 6 The schematic diagram of the integral coefficient membership function described by the application; Figure 7 The schematic diagram of the differential coefficient membership function described by the application; Figure 8 The schematic diagram of the longitudinal deviation simulation curve described by the application; Figure 9 The schematic diagram of the lateral deviation simulation curve described by the application. DETAILED DESCRIPTION

[0027] The technical solutions in the embodiments of the application will be clearly and completely described in the embodiments of the application in combination with the accompanying drawings. It should be noted that the embodiments in the application and the features in the embodiments can be combined with each other without conflict, and the described embodiments are only part of the embodiments of the application, but not all the embodiments.

[0028] Embodiment one, the method for automatic landing of a ship-borne fixed-wing aircraft under strong air disturbance conditions, the method comprises: The landing control of the ship-borne fixed-wing aircraft is decoupled into a longitudinal link and a lateral link, and a hierarchical cascade control structure is adopted, wherein the longitudinal link comprises a longitudinal position loop, a longitudinal speed loop and a longitudinal attitude loop, and the lateral link comprises a lateral position loop, a lateral attitude loop and a lateral roll angle speed loop; In the longitudinal link, a longitudinal position error is calculated through the longitudinal position loop, and a longitudinal speed deviation is generated based on online updating of proportional, integral and differential parameters of a neural network; In the lateral link, a lateral position error and an error change rate are calculated through the lateral position loop, and a heading angle deviation is generated based on dynamic adjustment of proportional, integral and differential parameters of fuzzy mathematics; the inputs of the fuzzy mathematics are the lateral position error and the error change rate, and the outputs are a proportional parameter adjustment amount, an integral parameter adjustment amount and a differential parameter adjustment amount; a fuzzy rule base is constructed based on fuzzy subsets of the error and the error change rate to adjust the parameters; An air disturbance compensation module is introduced to feed forward compensate the longitudinal speed deviation, specifically including: querying an air disturbance intensity coefficient empirical table based on the position of the aircraft relative to the ship body landing point, calculating the actual air disturbance influence coefficient through interpolation, and superimposing the actual air disturbance influence coefficient to the longitudinal speed deviation to obtain a compensated longitudinal speed deviation; In the longitudinal link, the compensated longitudinal speed deviation is input into the longitudinal speed loop to generate a desired pitch angle, and the desired pitch angle is subjected to limiting and filtering processing to obtain an equivalent pitch angle; based on the equivalent pitch angle error, a desired pitch angle speed operation amount is calculated through a proportional-integral controller; In the lateral link, a yaw angle error is calculated based on the heading angle deviation and the aircraft yaw angle to generate a desired roll angle; based on the desired roll angle error, a desired roll angle speed operation amount is calculated through a proportional-integral controller; The desired pitch angle speed operation amount output by the longitudinal link and the desired roll angle speed operation amount output by the lateral link are mapped to the rudder, elevator, aileron and throttle actuators through a control surface distributor to drive the ship-borne fixed-wing aircraft to complete automatic landing.

[0029] The method proposed in the embodiment decouples the longitudinal and lateral controls to reduce the complexity of the control system and improve the accuracy. The longitudinal and lateral controls are decoupled and independently controlled. The longitudinal link focuses on the vertical direction of the aircraft (such as height and speed), and the lateral link focuses on the horizontal direction of the aircraft (such as heading angle and yaw angle). A hierarchical cascade structure is adopted to further simplify the control task. The longitudinal link includes position, speed and attitude loops, and the lateral link includes position, attitude and roll angle speed loops.

[0030] The longitudinal position error of the aircraft is calculated in the longitudinal link control, and the proportional (P), integral (I) and derivative (D) control parameters are dynamically updated online through the neural network to generate the longitudinal velocity deviation. The introduction of the neural network enables the system to adjust the control parameters according to the real-time environmental changes, improving the control accuracy and robustness. An air disturbance compensation module is introduced to feed forward compensate the longitudinal velocity deviation. In the disturbance environment, the flight trajectory of the aircraft is easily affected by air disturbance. The compensation module queries the experience table and interpolates the actual air disturbance intensity to correct the longitudinal velocity deviation. This compensation mechanism ensures that the aircraft can still achieve precise longitudinal control in a strong disturbance environment.

[0031] In the lateral link control, the lateral link dynamically adjusts the proportional, integral and derivative parameters by calculating the lateral position error and error rate. Specifically, the fuzzy mathematics method adjusts the PID parameters according to the lateral position error and rate to generate the heading angle deviation. The advantage of this method is that it can adaptively adjust the control parameters in uncertain environments, thereby improving the stability and robustness of the control. By dynamically adjusting the parameters of the PID controller using fuzzy mathematics, the problem of excessive or insufficient adjustment of the traditional PID controller in a strong disturbance environment is avoided. Fuzzy control enables the control system to perform more flexible control according to different situations of the lateral error and error rate by constructing a fuzzy rule base.

[0032] The longitudinal link generates the desired pitch angle and calculates the desired pitch angle velocity operation amount through PID control. The lateral link calculates the yaw angle error by the heading angle deviation and the yaw angle, and generates the desired roll angle, and further calculates the desired roll angle velocity operation amount. The desired pitch angle velocity and roll angle velocity will be mapped through the control surface distributor to control the attitude adjustment of the aircraft. The desired pitch angle velocity and roll angle velocity operation amounts output by the longitudinal and lateral links are mapped to the various actuators of the aircraft (such as rudder, elevator, aileron and throttle) through the control surface distributor, driving the aircraft to complete automatic landing. This mapping process ensures that each actuator can accurately adjust the attitude of the aircraft to complete the landing task.

[0033] In the second embodiment, the neural network has a two-layer hidden layer structure, the first hidden layer contains 5 neurons and uses the ReLU activation function, and the second hidden layer contains 5 neurons and uses the tanh activation function; the loss function is defined as:

[0034] wherein, is used to weigh the current error, is used to weigh the current cumulative error, for balancing the rate of change of the current control variable.

[0035] Embodiment three, the embodiment is one kind for the further limitation of the automatic landing control method of shipborne fixed-wing aircraft under strong air disturbance condition described in embodiment one, the longitudinal velocity deviation is:

[0036] Wherein, The longitudinal velocity deviation is, The expected vertical sinking speed is, The real-time vertical position error is, The proportional coefficient of the controller is, The integral coefficient of the controller is, The differential coefficient of the controller is.

[0037] Embodiment four, the embodiment is one kind for the further limitation of the automatic landing control method of shipborne fixed-wing aircraft under strong air disturbance condition described in embodiment one, the fuzzy subset of fuzzy mathematics includes negative big, negative medium, negative small, zero, positive small, positive medium and positive big, and the membership function adopts triangle function;The fuzzy rule base is based on the following principles: when the lateral position error and the error change rate are both positive, the proportional parameter and the differential parameter adjustment amount are increased, and the integral parameter adjustment amount is reduced;When the error and the error change rate are both negative, the mirror adjustment parameter is adjusted.

[0038] Embodiment five, the embodiment is one kind for the further limitation of the automatic landing control method of shipborne fixed-wing aircraft under strong air disturbance condition described in embodiment one, the air disturbance intensity coefficient empirical table contains the longitudinal distance of multiple sampling points and the corresponding air disturbance intensity coefficient, the air disturbance intensity coefficient empirical table contains the longitudinal distance of multiple sampling points and the corresponding air disturbance intensity coefficient, and the actual air disturbance influence coefficient is calculated by interpolation:

[0039] Wherein, The position of the aircraft relative to the ship body landing point is, The position of the first sampling point relative to the ship body landing point is, The position of the i th sampling point relative to the ship body landing point is, The position of the i+1 th sampling point relative to the ship body landing point is, The position of the n th sampling point relative to the ship body landing point is, The coefficient corresponding to the first sampling point in the air disturbance influence coefficient table is, The coefficient corresponding to the i th sampling point in the air disturbance influence coefficient table is, The coefficient corresponding to the i+1 th sampling point in the air disturbance influence coefficient table is, The coefficient corresponding to the nth sampling point in the air disturbance influence coefficient table.

[0040] Embodiment six, this embodiment is a further limitation of the shipboard fixed-wing aircraft automatic landing control method under strong disturbance condition according to embodiment one, the calculation method of the expected pitch angle is:

[0041] Wherein, is the actual longitudinal velocity gain of the longitudinal velocity loop, is the expected longitudinal velocity gain of the longitudinal velocity loop, is the actual longitudinal velocity differential gain of the longitudinal velocity loop, is the actual longitudinal acceleration, is the actual longitudinal velocity, is the expected longitudinal velocity after considering the air disturbance compensation.

[0042] Embodiment seven, this embodiment is a further limitation of the shipboard fixed-wing aircraft automatic landing control method under strong disturbance condition according to embodiment one, the calculation method of the expected roll angle is:

[0043] Wherein, is the yaw angle proportional gain of the lateral inner loop, is the current time, is the yaw angle error of the aircraft at time k, is the yaw angle error of the aircraft at time k-1, is the yaw angle differential gain of the lateral inner loop, is the simulation step.

[0044] Embodiment eight, this embodiment is a further limitation of the shipboard fixed-wing aircraft automatic landing control method under strong disturbance condition according to embodiment one, the integral term of the proportional-integral controller is solved by Euler method:

[0045] Wherein, is the integral value of the lateral deviation at the current time, is the integral value of the longitudinal deviation at the last time, is the integral value of the lateral deviation at time 0, is the integral reference gain of the lateral deviation, is the integral parameter adjustment amount of the lateral deviation, is the lateral deviation at the current time.

[0046] Embodiment nine, the computer device comprises a memory and a processor, the memory stores a computer program, when the processor runs the computer program stored in the memory, the processor executes the automatic landing control method of the ship-borne fixed-wing aircraft under strong turbulence condition according to any one of the embodiments one to seven.

[0047] Embodiment ten, the computer readable storage medium stores a computer program, when the processor runs the computer program, the processor executes the steps of the automatic landing control method of the ship-borne fixed-wing aircraft under strong turbulence condition according to any one of the embodiments one to seven.

[0048] Embodiment eleven, see Figures 1 to 9 This embodiment is described. This embodiment is a specific embodiment of the automatic landing control method of the ship-borne fixed-wing aircraft under strong turbulence condition according to the embodiment one, and also serves to explain the embodiments two to eight. The embodiment proposes an automatic landing control method of the ship-borne fixed-wing aircraft under strong turbulence condition, which adopts two parallel hierarchical cascade control links in longitudinal and lateral directions, and sets longitudinal position loop and lateral position loop in the outer layer to track the ship deck reference trajectory, wherein the longitudinal channel introduces longitudinal position compensation module under disturbance, which equivalently converts air disturbance into feedforward and adaptive compensation signal, significantly weakens the influence of sea state and turbulence on the approach path, and in each direction, the position loop instruction further enters the inner loop system, converts the expected attitude into angular velocity, and drives the ship-borne fixed-wing aircraft actuator with high bandwidth. Finally, the aircraft executes the control amount from the two-channel inner loop, realizes the rapid and accurate adjustment of the attitude and trajectory, and realizes the safe and stable automatic landing under the conditions of ship deck movement and strong turbulence.

[0049] The ship-borne fixed-wing aircraft automatic landing control framework proposed in the embodiment is shown in Figure 1 The landing control of the ship-borne fixed-wing aircraft is divided into two parallel links in longitudinal and lateral directions, and is combined into the same set of actuators through rudder surface distributor at the end of the two links, while controlling rudder, elevator, aileron and throttle, etc.

[0050] In the longitudinal link, the error is first obtained by subtracting the "desired longitudinal position" from the current longitudinal state, and the basic control quantity is generated through the outer loop controller. To cope with the nonlinear and time-varying characteristics caused by strong disturbances and ship motion, a neural network is introduced to the outer loop controller to make online correction to the PID parameters or output. The longitudinal position error of the shipborne fixed-wing aircraft is input to the controller, and the controller outputs the longitudinal velocity deviation. At the same time, a disturbance compensation channel is set up to estimate the control disturbance as a feedforward quantity and superimpose it on the control command. Then, the attitude / angle velocity signal is sent to the longitudinal inner loop.

[0051] In the lateral link, the difference between the "desired lateral position" and the actual lateral position is obtained, the expected attitude such as roll and yaw is obtained through fuzzy PID, and the precise adjustment of angular velocity is realized through the lateral inner loop. The moment and thrust demand output by the lateral and longitudinal inner loops are unified and sent to the rudder angle distributor, which is mapped to the specific rudder and thrust device to drive the fixed-wing aircraft to complete attitude and trajectory adjustment and realize precise tracking of the deck center line and glide path.

[0052] The outer loop controller in the longitudinal link of the embodiment introduces a neural network to make online correction to the PID parameters or output, which includes: An adaptive adjustment method based on BP (back propagation) neural network is adopted, wherein the input dimension of the BP (back propagation) neural network is 7, including: the current tracking error , the error integral term , the error differential term , the current system output, i.e. the longitudinal position of the aircraft , the target reference signal, i.e. the expected longitudinal position output by the guidance law , the control quantity of the previous step , and the output change rate . These 7 features can fully reflect the current dynamic state and control trend of the system, providing comprehensive input information for real-time adjustment of the PID parameters.

[0053] The neural network structure adopts a two-layer hidden layer design, each layer containing 5 neurons. The first hidden layer uses the ReLU activation function to ensure that the hidden layer has good sparsity and nonlinear expression ability, and the second hidden layer uses the tanh function to ensure that the output converges and smoothly transitions within a certain range. The output layer contains 3 nodes corresponding to the proportional parameter, integral parameter and differential parameter in the PID controller, and the neural network loss function is designed as follows: (1) In the above formula: , , are used to weigh the current error, cumulative error and control amount rate of change, respectively, to achieve comprehensive optimization of tracking performance, robustness and control smoothness. The neural network continuously adjusts the internal weights through gradient descent algorithm in back propagation, calculates the gradient according to the partial derivative of the loss function on the network parameters, and updates the network weights and biases through iteration, so as to minimize the loss function, gradually optimize the output PID parameters, realize online learning and adaptive control, and make the control system maintain good tracking performance and stability in dynamic environment. The PID parameters output by the BP neural network are: 、 、 .

[0054] The vertical velocity deviation of the shipborne fixed-wing aircraft is calculated by the following formula : (2) In the above formula: is the expected vertical sinking speed, is the real-time vertical position error.

[0055] In summary, the longitudinal velocity of the aircraft is obtained by the variable gain control method in the embodiment .

[0056] The fuzzy PID control method in the lateral link of the embodiment includes: The fuzzy PID controller dynamically adjusts the proportional , integral and differential parameters of the PID controller through the fuzzy inference system to improve the adaptive ability and control performance of the system.

[0057] The input of the fuzzy controller is the current error and its rate of change , and the output is the three parameters 、 、 of the PID. The fuzzy subsets of the input and output variables are , i.e. negative large, negative medium, negative small, zero, positive small, positive medium and positive large. The membership function adopts a triangular function to ensure smooth transition of the rules and achieve efficiency.

[0058] When the error and the error rate of change are both positive, the fixed-wing aircraft accelerates away from the expected lateral position, at which time the proportional coefficient should be significantly increased, the differential coefficient should be moderately increased to suppress inertial overshoot, and the integral coefficient should be simultaneously reduced to avoid integral overshoot; if and If both are negative, mirror adjustment is taken according to symmetry principle. For larger than smaller than , keep small and gradually increase , reduce steady-state error. In the lateral landing trajectory control, the weight of the differential term is usually amplified to further enhance the dynamic response capability. Based on the above parameter characteristics and control objectives, a fuzzy rule base is established to realize online adjustment of PID parameters and adaptive optimization of system performance. The fuzzy rule base is shown in Tables 1, 2 and 3.

[0059] Table 1 Rule base

[0060] Table 2 Rule base

[0061] Table 3 Rule base

[0062] Wherein, NB, NM, NS, ZO, PS, PM, PB negative large, negative medium, negative small, zero, positive small, positive medium, positive large. The integral element of the controller is solved by Euler method, that is: (3) In the formula, is the integral value of the lateral deviation at the current time, is the integral value of the longitudinal deviation at the last time, is the integral value of the lateral deviation at 0 time, is the lateral deviation at the current time, is the simulation step.

[0063] The lateral lateral heading angle deviation of the shipborne fixed-wing aircraft is calculated by the following formula : (4) The embodiment considers the influence of air disturbance on the sinking speed of the aircraft, compensates through feedforward control, considers the disturbance influence in advance, actively offsets the influence of the wake flow, makes the actual sinking speed of the aircraft better track the ideal value, and thus improves the sinking tracking precision and system robustness of the aircraft in a complex air flow environment, and reduces the height error and risk in the landing process.

[0064] The feedforward controller for air disturbance compensation in the embodiment is designed as follows: The air turbulence influence coefficient of the aircraft at different sampling points is obtained by experience, and an empirical table is formed, as shown in Table 4. The air turbulence intensity coefficient is obtained by table lookup and interpolation, and the actual influence of air turbulence on the aircraft is obtained through the coefficient.

[0065] Table 4 Air turbulence intensity coefficient empirical table

[0066] In this embodiment, based on a large number of numerical simulations and flight mechanics modeling analysis, the values of each variable in the above air turbulence intensity coefficient empirical table are systematically adjusted and optimized. By fitting the influence intensity of air turbulence disturbance characteristics at different spatial positions, a set of representative sampling points and their corresponding influence coefficients are finally determined. The numerical parameters in this table can more accurately reflect the interference effect of air turbulence on the shipborne fixed-wing aircraft in the actual flight environment, thereby improving the effectiveness and reliability of the model in engineering applications.

[0067] The actual air turbulence influence coefficient on the sinking speed of the aircraft is calculated by interpolation as follows: (5) Wherein, is the position of the aircraft relative to the ship body landing point, is the position of the first sampling point relative to the ship body landing point, is the position of the i-th sampling point relative to the ship body landing point, is the position of the i+1-th sampling point relative to the ship body landing point, is the position of the n-th sampling point relative to the ship body landing point. is the coefficient corresponding to the first sampling point in the air turbulence influence coefficient table, is the coefficient corresponding to the i-th sampling point in the air turbulence influence coefficient table, is the coefficient corresponding to the i+1-th sampling point in the air turbulence influence coefficient table, is the coefficient corresponding to the n-th sampling point in the air turbulence influence coefficient table.

[0068] The air turbulence compensation amount designed in this embodiment is as shown in the following formula:

[0069] In the above formula, is the air turbulence intensity, as shown in the data type in Table 4.

[0070] This embodiment considers that air turbulence has a greater influence on the longitudinal direction of the aircraft. Therefore, the air turbulence compensation amount is superimposed on the expected vertical speed deviation to obtain the compensated speed deviation as shown in the following formula: (6) The longitudinal velocity loop control adopts the following control method: input the desired longitudinal velocity, and output the desired pitch angle.

[0071] (7) In the formula, is the actual longitudinal velocity gain of the longitudinal velocity loop, is the desired longitudinal velocity gain of the longitudinal velocity loop, is the actual longitudinal velocity differential gain of the longitudinal velocity loop, is the actual longitudinal acceleration.

[0072] The pitch angle is limited to prevent actuator saturation, and the expected pitch angle after limiting is calculated by the following formula: (8) The pitch angle is filtered through a second-order transfer function, and the design form of the second-order transfer function is as follows: (9) The command shaping is realized through the filter, the sharp command is converted into an S-curve to avoid actuator saturation, and some high-frequency noise can be removed. The transfer function needs to be discretized and converted into a difference equation form: (10) The equivalent pitch angle at the current time is obtained through the above formula , and the pitch angle error can be calculated.

[0073] The longitudinal inner loop pitch angle controller adopts PI control, inputs the pitch angle error , and outputs the pitch angle velocity operation quantity .

[0074] (11) (12) In the formula, respectively, the integral value of the longitudinal middle loop pitch angle error at the current time, the integral value of the longitudinal middle loop pitch angle error at the last time, and the integral value of the longitudinal middle loop pitch angle error at the 0 time. respectively, the integral gain and the proportional gain of the longitudinal middle loop.

[0075] The expected pitch angle velocity steering quantity output by the longitudinal inner loop angle velocity PI controller is combined with the current pitch angle velocity, angle of attack and other states of the aircraft, and through a series of dynamic compensation and actuator models, the actual shipborne fixed-wing aircraft rudder control command is finally generated.

[0076] wherein, respectively, are the integral value of the longitudinal middle loop pitch angle error at the current time, the integral value of the longitudinal middle loop pitch angle error at the last time, and the integral value of the longitudinal middle loop pitch angle error at the 0 time. respectively, are the integral gain and the proportional gain of the longitudinal middle loop.

[0077] The expected pitch angle velocity control quantity output by the longitudinal inner loop angular velocity PI controller is combined with the current pitch angle velocity, angle of attack and other states of the aircraft, and through a series of dynamic compensation and actuator models, the actual shipborne fixed-wing aircraft rudder control command is finally generated.

[0078] In the embodiment, the following formula is used to calculate the yaw angle error of the aircraft:

[0079] In the above formula, is the yaw angle error, is the expected yaw angle, is the yaw angle of the landing point, is the yaw angle of the aircraft.

[0080] The embodiment uses the following control method to calculate the expected roll angle, as shown in the following formula, and the specific input is the expected yaw angle error, and the output is the expected roll angle . (13) The roll angle error is:

[0081] In the above formula, is the roll angle of the aircraft.

[0082] The embodiment uses the following method to calculate the expected roll angle velocity, and the specific input is the roll angle error, and the output is the roll angle velocity bar command.

[0083] (14) (15) In the above formula, respectively, are the integral value of the lateral inner loop roll angle error at the current time, the integral value of the lateral inner loop roll angle error at the last time, and the integral value of the lateral inner loop roll angle error at the 0 time. respectively, are the gains of the lateral inner loop.

[0084] The embodiment utilizes simulation means to construct an automatic landing simulation platform of the shipborne fixed-wing aircraft, and verifies the related algorithm in the embodiment. The initial longitudinal deviation of the shipborne fixed-wing aircraft in the embodiment is -15 meters, the initial lateral deviation is -17 meters, the automatic landing process simulation task is carried out in the presence of air disturbance, and the longitudinal and lateral deviation curves of the aircraft are as shown in Figure 8 and Figure 9 .

[0085] Figure 8 and Figure 9 The results shown in the figures show that the final longitudinal deviation of the aircraft is about -0.09 meters and the final lateral deviation is about 0.11 meters after about 11 seconds, which is small, proving that the embodiment can provide a more accurate automatic landing control algorithm for the shipborne fixed-wing aircraft.

[0086] In summary, the automatic landing control method of the shipborne fixed-wing aircraft under strong disturbance conditions proposed in the embodiment can cope with the complex and changeable sea environment in the landing process. The principle is that in the method, the longitudinal and lateral position controls both adopt a three-loop structure, and the middle loop and the inner loop both adopt PID control. To enhance the adaptability of the system, the PID parameters of the longitudinal outer loop are updated online through a neural network, and the lateral PID parameters are dynamically adjusted by means of fuzzy mathematics, thereby effectively improving the control accuracy and robustness of the aircraft in complex environments. The control system helps to significantly improve the safety and stability of the autonomous landing of the fixed-wing aircraft on the sea vessel platform, and provides technical support for the intelligent development of the marine field.

[0087] Those skilled in the art will appreciate that embodiments of the present disclosure can be provided as methods, systems or computer program products. Therefore, the present disclosure can take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Moreover, the present disclosure can take the form of a computer program product implemented on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROMs, optical storage devices, etc.) containing computer-usable program code.

[0088] The present disclosure is described with reference to flowcharts and / or block diagrams of methods, devices (systems) and computer program products according to embodiments of the present disclosure. It should be understood that each flow and / or block in the flowcharts and / or block diagrams, as well as combinations of flows and / or blocks in the flowcharts and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing apparatus to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing apparatus produce a means for implementing the functions specified in the flowcharts and / or block diagrams. Figure 1 one flow or multiple flows and / or blocks Figure 1apparatuses that implement the functions specified in the flowchart or flowcharts and / or blocks Figure 1 flowchart or flowcharts and / or blocks Figure 1 the function specified in the flowchart or flowcharts and / or blocks. The computer program instructions can also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart or flowcharts and / or blocks Figure 1 flowchart or flowcharts and / or blocks Figure 1 the function specified in the flowchart or flowcharts and / or blocks. Finally, it should be noted that the above-mentioned embodiments are merely used to illustrate the technical solutions of the present disclosure, rather than limit the scope of protection of the present disclosure. Although the present disclosure has been described in detail with reference to the above-mentioned embodiments, it should be understood by those skilled in the art that, after reading the present disclosure, those skilled in the art can make various changes, modifications or equivalent replacements to the specific embodiments of the present disclosure. However, these changes, modifications or equivalent replacements are all within the scope of protection of the disclosed patent application.

Claims

1. A method for automatic landing control of a shipborne fixed-wing aircraft under strong turbulence conditions, characterized in that, The method includes: The landing control of shipborne fixed-wing aircraft is decoupled into longitudinal and lateral links and a hierarchical cascaded control structure is adopted. The longitudinal link includes longitudinal position loop, longitudinal velocity loop and longitudinal attitude loop, and the lateral link includes lateral position loop, lateral attitude loop and roll rate loop. In the longitudinal link, the longitudinal position error is calculated through the longitudinal position loop, and the proportional, integral and differential parameters are updated online based on the neural network to generate the longitudinal velocity deviation; In the lateral link, the lateral position error and error change rate are calculated through the lateral position loop, and the proportional, integral and differential parameters are dynamically adjusted based on fuzzy mathematics to generate the heading angle deviation; the input of the fuzzy mathematics is the lateral position error and error change rate, and the output is the proportional parameter adjustment amount, integral parameter adjustment amount and differential parameter adjustment amount. A fuzzy rule base is constructed based on the fuzzy subset of the error and error change rate for parameter adjustment. An air turbulence compensation module is introduced to perform feedforward compensation for longitudinal velocity deviation. Specifically, it includes: querying an empirical table of air turbulence intensity coefficients based on the aircraft's position relative to the ship's landing point, calculating the actual air turbulence influence coefficient through interpolation, and superimposing it on the longitudinal velocity deviation to obtain the compensated longitudinal velocity deviation. In the longitudinal link, the compensated longitudinal velocity deviation is input into the longitudinal velocity loop to generate the desired pitch angle, and the desired pitch angle is limited and filtered to obtain the equivalent pitch angle; based on the equivalent pitch angle error, the desired pitch angle velocity operation is calculated through a proportional-integral controller. In the lateral link, the yaw angle error is calculated based on the heading angle deviation and the aircraft yaw angle to generate the desired roll angle; based on the desired roll angle error, the desired roll rate control is calculated through a proportional-integral controller. The desired pitch rate output from the longitudinal link and the desired roll rate output from the lateral link are mapped to the rudder, elevator, aileron, and throttle actuators via the control surface distributor, driving the shipborne fixed-wing aircraft to complete an automatic landing.

2. The automatic landing control method for a shipborne fixed-wing aircraft under strong turbulence conditions according to claim 1, characterized in that, The neural network employs a two-layer hidden layer structure. The first hidden layer contains 5 neurons and uses the ReLU activation function, while the second hidden layer contains 5 neurons and uses the tanh activation function. The loss function is defined as: in, Used to weigh the current error Used to weigh the current cumulative error Used to weigh the rate of change of the current control quantity.

3. The automatic landing control method for a shipborne fixed-wing aircraft under strong turbulence conditions according to claim 1, characterized in that, The longitudinal velocity deviation is: in, For longitudinal velocity deviation, For the desired vertical sinking velocity, For real-time vertical position error, The proportional gain of the controller. The integral coefficient of the controller, represents the differential coefficient of the controller.

4. The automatic landing control method for a shipborne fixed-wing aircraft under strong turbulence conditions according to claim 1, characterized in that, The fuzzy subsets of the fuzzy mathematics include negative large, negative medium, negative small, zero, positive small, positive medium, and positive large. The membership function adopts a triangular function. The fuzzy rule base is based on the following principles: when the lateral position error and the error change rate are both positive, increase the adjustment amount of the proportional parameter and the differential parameter, and decrease the adjustment amount of the integral parameter; when the error and the error change rate are both negative, mirror the adjustment parameter.

5. The automatic landing control method for a shipborne fixed-wing aircraft under strong turbulence conditions according to claim 1, characterized in that, The empirical table of air turbulence intensity coefficients includes the longitudinal distance of multiple sampling points and the corresponding air turbulence intensity coefficients. The actual air turbulence influence coefficient is calculated by interpolation. in, This refers to the aircraft's position relative to the ship's landing point. This represents the position of the first sampling point relative to the ship's landing point. Let i be the position of the i-th sampling point relative to the hull landing point. This represents the position of the (i+1)th sampling point relative to the hull landing point. The position of the nth sampling point relative to the hull landing point. This refers to the coefficient corresponding to the first sampling point in the air turbulence influence coefficient table. The coefficient corresponding to the i-th sampling point in the air turbulence influence coefficient table is... This refers to the coefficient corresponding to the (i+1)th sampling point in the air turbulence influence coefficient table. This represents the coefficient corresponding to the nth sampling point in the air turbulence influence coefficient table.

6. The automatic landing control method for a shipborne fixed-wing aircraft under strong turbulence conditions according to claim 1, characterized in that, The desired pitch angle is calculated as follows: in, This represents the actual longitudinal velocity gain of the longitudinal velocity loop. For the desired longitudinal velocity gain of the longitudinal velocity loop, The differential gain of the actual longitudinal velocity loop. This is the actual longitudinal acceleration. This represents the actual longitudinal velocity. The desired longitudinal velocity after taking into account air disturbance compensation.

7. The automatic landing control method for a shipborne fixed-wing aircraft under strong turbulence conditions according to claim 1, characterized in that, The desired roll angle is calculated as follows: in, This is the lateral inner ring yaw angle proportional gain. For the current moment, Let be the yaw angle error of the aircraft at time k. Let $\mathbf{k}$ be the yaw angle error of the aircraft at time $k-1$. The differential gain of the yaw angle for the inner lateral loop. This is the simulated step size.

8. The automatic landing control method for a shipborne fixed-wing aircraft under strong turbulence conditions according to claim 1, characterized in that, The integral term of the proportional-integral controller is solved using the Euler method: in, This is the integral value of the lateral deviation at the current moment. This is the integral value of the longitudinal deviation at the previous moment. The integral value of the lateral deviation at time 0. The integral reference gain for the lateral deviation. The integral variable gain is the lateral deviation. This represents the lateral deviation at the current moment.

9. A computer device, characterized in that: It includes a memory and a processor, wherein the memory stores a computer program, and when the processor runs the computer program stored in the memory, the processor executes an automatic landing control method for a shipborne fixed-wing aircraft under strong turbulence conditions as described in any one of claims 1-7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program that, when executed by a processor, performs the steps of an automatic landing control method for a shipborne fixed-wing aircraft under strong turbulence conditions as described in any one of claims 1-7.

Images