Variable-pitch multi-rotor unmanned aerial vehicle and flight attitude control method thereof

Abstract

The invention belongs to the technical field of unmanned aerial vehicles and flight control thereof, and discloses a variable-pitch multi-rotor unmanned aerial vehicle and a flight attitude control method thereof. Wherein the variable-pitch multi-rotor unmanned aerial vehicle is provided with an active-disturbance-rejection attitude control system, the active-disturbance-rejection attitude control system is provided with a rolling control loop, a pitching control loop and a yawing control loop, and each control loop comprises an extended state observer, a differential tracker and a nonlinear state error feedback. And when the control parameters are optimized, the control parameters are optimized by using an improved cumberfish optimization algorithm, and a dynamic updating strategy of the improved cumberfish optimization algorithm adopts a cumberfish position updating rule based on a variable spiral sine-cosine strategy. According to the technical scheme, the technical problem that control parameters of the unmanned aerial vehicle are difficult to adjust due to factors such as model uncertainty, external interference and nonlinear coupling characteristics caused by variable-pitch rotors in a complex dynamic environment is solved, and the flight performance of the unmanned aerial vehicle can be improved.

Description

Technical Field

[0001] The present invention belongs to the technical field of unmanned aerial vehicles and their flight control, and particularly relates to a variable-pitch multi-rotor unmanned aerial vehicle and a flight attitude control method thereof. Background Art

[0002] Unmanned aerial vehicles have great tactical and strategic value. They can take off and land from short takeoff and landing platforms such as islands, medium and large ships, and offshore oil fields, and perform various tasks such as rescue, patrol, and emergency transportation. In order to further expand the application value of unmanned aerial vehicles and improve their anti-strong wind interference ability so that they can complete maritime flight tasks in harsh environments, it is an important direction in the current research on unmanned aerial vehicle control.

[0003] The variable-pitch multi-rotor unmanned aerial vehicle directly adjusts the thrust by changing the pitch angle through a variable-pitch system, and its response speed is superior to the traditional scheme that relies on motor speed regulation. It is superior to the traditional fixed-pitch design in terms of energy efficiency, adaptability, stability, and economy, and is especially suitable for application scenarios with multiple tasks, long endurance, and complex environments. Currently, the PID (Proportional-Integral-Derivative Controller) controller is the most widely used controller in traditional flight control systems. However, when applied to variable-pitch multi-rotor unmanned aerial vehicles, it faces core bottlenecks such as nonlinearity, coupling, delay, and energy efficiency optimization, and it is unable to accurately compensate for nonlinear effects, making it difficult to obtain ideal flight quality.

[0004] In view of the above situation, the industry has proposed various unmanned aerial vehicle control methods, such as nonlinear control techniques, including sliding mode control, feedback linearization, robust control, etc. However, these control methods are difficult to achieve excellent results in actual flight. Further, the industry has found that the variable-pitch multi-rotor unmanned aerial vehicle system has strong nonlinearity and time-variation. The active disturbance rejection controller does not require an accurate controlled object model. It can design an observer by using an extended state observer (ESO) to estimate the system state and total disturbance in real time. Its core idea is to regard the internal uncertainty and external disturbance of the system as the total disturbance, and dynamically estimate and compensate through the extended state. However, the active disturbance rejection controller has many parameters. When optimizing the parameters under different environments and different targets, its parameter space has a non-linear and complex multi-peak distribution, and it is easy to fall into a local optimum. Currently, the industry has proposed various optimization methods to find the optimal parameters of the active disturbance rejection controller more quickly, such as genetic algorithms, particle swarm optimization methods, gene optimization methods, etc., but they all have technical defects such as long exploration time, low optimization efficiency, and difficulty in ensuring finding the global optimal solution. Summary of the Invention

[0005] The object of the present invention is to provide a variable-pitch multi-rotor unmanned aerial vehicle and its flight attitude control method to solve one or more of the above-mentioned technical problems. In the technical solution disclosed by the present invention, a new active disturbance rejection attitude control system is provided for the variable-pitch multi-rotor unmanned aerial vehicle, which solves the technical problem of difficult adjustment of control parameters existing in the prior art for unmanned aerial vehicles due to factors such as model uncertainty, external interference, and the non-linear coupling characteristics brought by variable-pitch rotors in a complex dynamic environment, and can more accurately control the flight attitude of the unmanned aerial vehicle, improving the flight performance of the unmanned aerial vehicle.

[0006] To achieve the above object, the present invention adopts the following technical solutions: In the first aspect of the present invention, a variable-pitch multi-rotor unmanned aerial vehicle is provided. The variable-pitch multi-rotor unmanned aerial vehicle is provided with a flight control system, and the flight control system is used to realize the flight control of the variable-pitch multi-rotor unmanned aerial vehicle; The flight control system is provided with an active disturbance rejection attitude control system, and the active disturbance rejection attitude control system is used to realize the flight attitude control of the variable-pitch multi-rotor unmanned aerial vehicle in a disturbance environment; wherein, the active disturbance rejection attitude control system is provided with three control loops of roll, pitch, and yaw, and each control loop includes an extended state observer, a differential tracker, and a non-linear state error feedback; Among them, in each control loop, the optimization steps of the control parameters of the extended state observer, the differential tracker, and the non-linear state error feedback include: based on the established objective function, using an improved kingfisher optimization algorithm to optimize the control parameters to obtain the optimization result of the control parameters; wherein, the dynamic update strategy of the improved kingfisher optimization algorithm adopts a kingfisher position update rule based on a variable spiral sine-cosine strategy.

[0007] A further improvement of the technical solution of the present invention lies in that, The variable spiral sine-cosine strategy tends to global search in the early stage of optimization and tends to local search in the later stage of optimization.

[0008] A further improvement of the technical solution of the present invention lies in that, A local escape mechanism is set in the process of local search.

[0009] A further improvement of the technical solution of the present invention lies in that, The expression of the objective function is: ; In the formula, is the total objective function of the control loop; is the root mean square error, which is used to measure the tracking accuracy; is the system settling time, which is used to measure the dynamic response; is the overshoot, which is used to measure the system stability; All are weight coefficients.

[0010] A further improvement of the technical solution of the present invention lies in that The step of using the improved kingfisher optimization algorithm to optimize the control parameters and obtaining the optimization result of the control parameters includes: According to the preset objective function value constraint condition, use the improved kingfisher optimization algorithm to find the minimum value of the objective function. After iterative update until the preset condition is satisfied, obtain the optimization result of the control parameters; wherein, first initialize the kingfisher optimization algorithm population and population parameters, and each individual in the population represents a control parameter to be optimized; then in each iteration, use the kingfisher position update rule based on the variable spiral sine-cosine strategy to obtain a new population, and evaluate the fitness value of the population according to the established kingfisher fitness function; finally, after iterative update until the preset condition is satisfied, use the population corresponding to the fitness value closest to the preset optimal fitness value as the optimization result of the control parameters.

[0011] A further improvement of the technical solution of the present invention lies in that In the kingfisher position update rule based on the variable spiral sine-cosine strategy, the update formula in the exploitation stage is as follows: ; In the formula, is the position of individual in the th generation; is the position of individual in the th generation; is the variable spiral coefficient; are respectively the hybrid adaptive coefficient, exploration factor, and step size coefficient; b is the reference offset; is the reference point or target position of individual in the th generation; is the dynamic scaling factor, used to control the step size of the search; is the random angle; and are both random numbers and both follow the uniform distribution of [0, 1].

[0012] A further improvement of the technical solution of the present invention lies in that The expression of the variable spiral coefficient is: ; In the formula, is the parameter for controlling the spiral; is the parameter representing the period of the spiral; l is the linear attenuation factor;t is the number of iterations; is the maximum number of iterations.

[0013] A further improvement of the technical solution of the present invention lies in that in the early stage of optimization, takes a value of 1, which is used to enable the search path to cover a larger range to enhance the global search ability; in the later stage of optimization, as the number of iterations increases, gradually decreases, which is used to narrow the search range to improve the local optimization accuracy.

[0014] A further improvement of the technical solution of the present invention lies in that in each of the control loops, in the optimization steps of the control parameters of the extended state observer, the differential tracker, and the nonlinear state error feedback, the control parameters to be optimized of the extended state observer include the state error feedback gain, the observer nonlinear strength parameter, and the threshold parameter for switching between the linear and nonlinear regions of the observer; the control parameters to be optimized of the nonlinear state error feedback include the error feedback gain, the nonlinear strength, and the threshold parameter for switching between the linear and nonlinear regions; the control parameters to be optimized of the differential tracker include the filtering factor and the speed factor that determines the speed of the system's tracking signal; In addition, it also includes global parameters to be optimized, and the global parameters include the global control gain compensation coefficient.

[0015] In the second aspect of the present invention, a flight attitude control method for a variable-pitch multi-rotor UAV is provided, including: During the process of realizing the flight control of the variable-pitch multi-rotor UAV through the flight control system, the desired pitch angle, the desired yaw angle, and the desired roll angle are input into the active disturbance rejection attitude control system to obtain the rudder deflection value, so as to realize the flight attitude control of the variable-pitch multi-rotor UAV in a disturbed environment through the active disturbance rejection attitude control system.

[0016] Compared with the prior art, the present invention has the following beneficial effects: The present invention specifically discloses a variable-pitch multi-rotor unmanned aerial vehicle (UAV), which is provided with an active disturbance rejection attitude control system newly designed by the present invention. In the active disturbance rejection attitude control system, an improved pied kingfisher optimization algorithm (MPKO) is used to optimize control parameters, solving the problem in the prior art that it is difficult to adjust control parameters of the UAV due to factors such as model uncertainty, external disturbances, and non-linear coupling characteristics brought by variable-pitch rotors in a complex dynamic environment. It has the advantages of shorter exploration time, higher optimization efficiency, and the ability to ensure finding the global optimal solution, etc., and can improve the flight performance of the variable-pitch multi-rotor UAV. Further specifically, the improved pied kingfisher optimization algorithm adopted by the active disturbance rejection attitude control system newly designed by the present invention establishes a pied kingfisher position update rule based on a variable spiral sine-cosine strategy to balance global and local searches. The global search and fine local development capabilities can well adapt to the tuning and optimization requirements of the active disturbance rejection attitude control system, and can effectively avoid falling into local optima. In addition, the improved pied kingfisher optimization algorithm has fewer core parameters (exemplarily, such as only dive speed, angle adjustment step size, etc.), reducing the difficulty of parameter adjustment, shortening the optimization duration, and ensuring finding the global optimal solution with high efficiency.

[0017] In the flight attitude control method disclosed by the present invention, based on the improved pied kingfisher optimization algorithm, the optimal solution of control parameters can be found more quickly, improving the optimization efficiency. Among them, the improved pied kingfisher optimization algorithm tends to global search in the early stage of optimization, increasing the chance of finding the global optimal solution and improving the reliability of the optimization result. In the later stage of optimization, it tends to local search (the exemplary preferred technical solution can also set a local escape mechanism to be able to more finely adjust parameters), improving the optimization accuracy. Further summarily, the technical solution of the present invention is particularly suitable for attitude control of variable-pitch multi-rotor UAVs in a disturbed environment, can more accurately control the flight attitude of the UAV, and improves the flight stability and safety of the UAV. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] In order to more clearly illustrate the technical solutions in the present invention or the prior art, the following will briefly introduce the drawings required for use in the description of the embodiments or the prior art; obviously, the drawings in the following description are some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0019] Figure 1 It is a schematic block diagram of the basic structure of the active disturbance rejection attitude control system in the embodiment of the present invention; Figure 2 It is a schematic flow chart of the improved pied kingfisher optimization algorithm in the embodiment of the present invention; Figure 3 It is a schematic diagram of the current position update process of the kingfisher after the particle swarm is disturbed in the embodiment of the present invention; Figure 4 It is a schematic diagram of the pitch angle control comparison curve in the embodiment of the present invention; Figure 5 It is a schematic diagram of the roll angle control comparison curve in the embodiment of the present invention; Figure 6 It is a schematic diagram of the yaw angle control comparison curve in the embodiment of the present invention. Detailed implementation manners

[0020] To make the objectives, technical solutions and advantages of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention; obviously, the described embodiments of the technical solutions are part of the embodiments of the present invention, rather than all of the embodiments.

[0021] Based on the technical solutions disclosed in the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention. In addition, the terms "include" and "have" and any variations thereof are intended to cover non-exclusive inclusions. For example, a process, method, system, product or device that includes a series of steps or units does not necessarily have to be limited to those steps or units clearly listed, but may include other steps or units not clearly listed or inherent to these processes, methods, products or devices.

[0022] A variable-pitch multi-rotor unmanned aerial vehicle provided in an embodiment of the present invention is provided with a flight control system, and the flight control system is used to implement flight control of the variable-pitch multi-rotor unmanned aerial vehicle; the flight control system is provided with an active disturbance rejection attitude control system, and the active disturbance rejection attitude control system is used to implement flight attitude control of the variable-pitch multi-rotor unmanned aerial vehicle in a disturbed environment; specifically, the active disturbance rejection attitude control system is provided with three control loops for roll, pitch and yaw, and each control loop includes an extended state observer, a differential tracker and a non-linear state error feedback; further specifically, the active disturbance rejection attitude control system is used to input a desired pitch angle, a desired yaw angle and a desired roll angle, and output a rudder deflection value; Among them, the optimization steps of the control parameters of the extended state observer, the differential tracker, and the nonlinear state error feedback include: based on the established objective function, using the improved kingfisher optimization algorithm to optimize the control parameters to obtain the control parameter optimization results; wherein, the dynamic update strategy of the improved kingfisher optimization algorithm adopts the kingfisher position update rule based on the variable spiral sine-cosine strategy; further explanatory, the variable spiral sine-cosine strategy tends to global search in the early stage of optimization, and tends to local search in the later stage of optimization, with the aim of balancing global search and local search; in a further preferred technical solution, the local search process is also provided with a local escape mechanism.

[0023] In the technical solution provided by the embodiment of the present invention, the differential tracker is used to arrange the transition process, solve the contradiction between the rapid response and overshoot of the system, and extract the tracking signal of the input signal and its accurate differential signal, so as to design a reasonable controller; the extended state observer is used to model the inaccurate parts of the drone and the parts that cannot be modeled, including parameter changes, sensor noise, wind field disturbances, etc., and then use the expanded state variables to estimate the total disturbance, and use the estimated value of the total disturbance to reasonably compensate the control signal. In this way, the nonlinear control system is simplified to a higher-order series integral linear control system; the nonlinear state error feedback uses the tracking signal generated by the differential tracker and the observation signal generated by the extended state observer to calculate the tracking error of the system, and then uses the nonlinear combination method to generate the preliminary control quantity, and finally compensates the expanded state variables observed by the extended state observer, that is, the total disturbance is reduced to obtain the actual control quantity.

[0024] In the technical solution disclosed in the embodiment of the present invention, an improved kingfisher optimization algorithm is introduced to solve the technical problems existing in the parameter optimization process of the active disturbance rejection controller, such as long exploration time, low optimization efficiency, and difficulty in ensuring finding the global optimal solution. Specifically, the active disturbance rejection attitude control system in the technical solution of the present invention includes an extended state observer, a differential tracker, and a nonlinear state error feedback. The control parameters of these components need to be optimized through an optimization algorithm. The technical solution of the present invention adopts a control parameter optimization method based on the improved kingfisher optimization algorithm, which can more efficiently find the optimal solution of the control parameters, thereby improving the optimization efficiency, shortening the exploration time, and increasing the possibility of finding the global optimal solution. Further explanatorily, the core technical means of the technical solution of the embodiment of the present invention is the improved kingfisher optimization algorithm. This algorithm is improved on the basis of the traditional kingfisher optimization algorithm, and a kingfisher position update rule introducing a variable spiral sine-cosine strategy is introduced. This strategy makes the algorithm tend to global search in the early stage of optimization, can explore the parameter space more widely, and increase the chance of finding the global optimal solution; in the later stage of optimization, it tends to local search, can adjust the parameters more finely, and improve the optimization accuracy. In addition, a local escape mechanism is set in the local search process to prevent the algorithm from falling into the local optimal solution. To sum up, the technical solution of the embodiment of the present invention improves the traditional kingfisher optimization algorithm, introduces a kingfisher position update rule with a variable spiral sine-cosine strategy, and through the improved algorithm, realizes the balance between global search and local search, not only ensures that the algorithm can widely explore the parameter space, but also can finely adjust the parameters, improves the optimization efficiency and accuracy, is particularly suitable for the flight attitude control of a variable-pitch multi-rotor UAV in a disturbed environment, can more accurately control the flight attitude of the UAV, and can improve the flight stability and safety of the UAV; in addition, the further preferred technical solution of the embodiment of the present invention sets a local escape mechanism in the local search process, which is a further solution proposed for the problem that the algorithm may fall into the local optimal solution.

[0025] In a preferred embodiment of the present invention, the expression of the set objective function is: ; In the formula, is the total objective function of the control loop; is the root mean square error (used to measure the tracking accuracy); is the system settling time (used to measure the dynamic response); is the overshoot (used to measure the system stability); is the weight coefficient.

[0026] The objective function set in the embodiment of the present invention realizes multi-objective optimization by weighted combination of multiple key indicators, balancing tracking accuracy, dynamic response, and system stability.

[0027] As a preferred technical solution of an embodiment of the present invention, the step of optimizing control parameters by using an improved kingfisher optimization algorithm to obtain the optimization result of control parameters specifically includes: According to the constraint conditions of the preset objective function value, use the improved kingfisher optimization algorithm to find the minimum value of the objective function. After iterative update to meet the preset conditions, obtain the optimization result of the control parameters; wherein, First, initialize the kingfisher optimization algorithm population and population parameters. Each individual in the population represents a control parameter to be optimized; In each iteration, use the kingfisher position update rule based on the variable spiral sine-cosine strategy to obtain a new population, and evaluate the fitness value of the population according to the established kingfisher fitness function; After iterative update to meet the preset conditions, the population corresponding to the fitness value closest to the preset optimal fitness value is used as the optimization result of the control parameters.

[0028] In the embodiment of the present invention, the kingfisher position update rule based on the variable spiral sine-cosine strategy (used to balance global and local searches) includes: To improve the global exploration and fine local development capabilities of the algorithm, the improved kingfisher optimization algorithm introduces a variable spiral sine-cosine strategy; wherein, The mathematical model of the sine-cosine algorithm is as follows: ; In the formula, is the position of individual at the th generation; is the position of individual at the th generation; is the reference point or target position of individual at the th generation; is a dynamic scaling factor used to control the search step size, which usually gradually decreases as the number of iterations increases; is a random angle between [0, 2π], which determines the fluctuation mode of position update; and are both random numbers, both following a uniform distribution of [0, 1], used to introduce randomness and diversity in the update formula.

[0029] By gradually narrowing the search range, the algorithm can explore the previously discovered favorable areas more deeply, thereby improving the search ability for local optimal solutions; the dynamic adjustment of the spiral factor enables the search process to gradually transition from global exploration in the initial stage to local refinement search in the later stage, achieving a balance between global search and local optimization.

[0030] The mathematical model of the spiral coefficient is as follows: ; In the formula, represents the variable spiral coefficient, which is a key parameter used to dynamically adjust the scope of spiral search to improve the effectiveness of the algorithm in different search stages; in the initial stage, takes the value of 1, enabling the search path to cover a larger range, thereby enhancing the global search ability; in the later stage of the search, as the number of iterations increases, will gradually decrease, thus narrowing the search scope to improve the accuracy of local optimization. is a parameter used to control the spiral, taking the value of 1 in the early stage and decreasing as the number of iterations increases in the later stage. is a parameter representing the period of the spiral. Generally, is a parameter that linearly decreases from 1 to -1 according to the number of iterations. l is the linear attenuation factor; t is the number of iterations; is the maximum number of iterations.

[0031] An update formula for the development stage of the variable spiral sine-cosine algorithm improved kingfisher optimization algorithm is established. The powerful development ability of the sine-cosine algorithm is utilized to further strengthen the development ability of the kingfisher optimization algorithm, and the oscillating spiral line of the variable spiral coefficient is used to improve the balance ability of the algorithm in exploration and development; Finally, the update formula for the development stage of the improved kingfisher algorithm is as follows: ; In the formula, is the position of individual at the th generation; is the position of individual at the th generation; represents the variable spiral coefficient; are the hybrid adaptive coefficient, exploration factor, and step size coefficient respectively; b is the reference offset; is the reference point or target position of individual at the th generation; is a dynamic scaling factor used to control the search step size, usually gradually decreasing as the number of iterations increases; is a random angle between [0, 2π], which determines the fluctuation mode of position update; and are random numbers, both following the uniform distribution of [0, 1], used to introduce randomness and diversity into the update formula.

[0032] In the specific exemplary technical solution of the present invention, the active disturbance rejection attitude control system includes three control loops for roll, pitch, and yaw. Each control loop consists of a differential tracker, an extended state observer, and a non-linear state error feedback. Among them, the differential tracker is used to receive the error signal and generate a smooth transition trajectory signal and its first-order differential signal. The extended state observer is used to estimate the internal state and total disturbance of the system in real time according to the output signal of the controlled object and the control input. The non-linear state error feedback is used to combine the output of the differential tracker and the estimated value of the extended state observer module to generate a non-linear compensation control quantity. The improved kingfisher optimization algorithm is used to optimize the parameters of the extended state observer to be , the parameters of the non-linear state error feedback are , the parameters of the differential tracker are and the global control gain compensation coefficient b .

[0033] Compared with the traditional parameter tuning method, the technical solution disclosed in the embodiment of the present invention breaks through the local convergence limitation of traditional optimization based on the improved kingfisher optimization algorithm, realizes the global coverage of the multi-dimensional parameter space, significantly accelerates the convergence speed through a parallel heuristic search architecture, reduces the time cost of method optimization, and improves the parameter optimization efficiency.

[0034] Please refer to Figures 1 to 3 , in the specific exemplary technical solution of the present invention, a variable pitch multi-rotor UAV is provided, which adopts an attitude active disturbance rejection control parameter tuning method based on the improved kingfisher optimization algorithm. Among them, the attitude active disturbance rejection control parameter tuning method specifically includes the following steps: Step 1: Build a dynamic model of the variable pitch multi-rotor UAV and an active disturbance rejection (ADRC) attitude control system for the variable pitch multi-rotor UAV; In the specific exemplary technical solution, a variable pitch multi-rotor UAV model is established to replace the actual controlled object. To ensure the balance between model accuracy and operation complexity, a simplified state space model of the roll, pitch, and yaw channels of the variable pitch multi-rotor UAV is established. A three-channel active disturbance rejection controller model is established, and the control system block diagram is as Figure 2 shown, as the flight attitude control system of the variable pitch multi-rotor UAV.

[0035] An extended state observer is established, and the expression is: ; In the formula, are the estimated values of the state variables attitude angle, attitude angular velocity, and total disturbance; is the state error feedback gain, Affects the estimation of attitude angle, Estimation of the total disturbance affecting the system, Affects the response of the system; is the nonlinear strength parameter of the extended state observer, is the threshold parameter for switching between the linear and nonlinear regions of the extended state observer; is the observation error; is the output signal of the system; is a nonlinear function, is the global control gain compensation coefficient, It is the control input item of the system; Specifically and exemplarily, Expressed as The expression is: ; In the formula, is the function variable. For example, is the error signal, the nonlinear factor, and the threshold value.

[0036] Establish a differential tracker, the expression is: ; In the formula, Indicates the attitude angle at the current moment, Indicates the angular velocity of the current attitude. is the time step, Indicates the attitude angle at the next moment; Indicates the angular velocity of the attitude at the next moment, is the filtering factor, It is the speed factor that determines how fast the system tracks the signal; For a given input signal; The function is the fastest comprehensive function; Establish nonlinear feedback, the expression is:

[0037] In the formula, is the error feedback gain, is the attitude angle tracking error, is the attitude angular velocity tracking error, is the nonlinear strength of the nonlinear feedback, It is the threshold parameter for switching between linear and nonlinear regions of nonlinear feedback.

[0038] The final control amount after compensation is expressed as:

[0039] In the formula, is the global control gain compensation coefficient.

[0040] Step 2: Establish the fitness function of the kingfisher, initialize the parameters of the active disturbance rejection attitude control system, create the initial population of the kingfisher optimization algorithm, and initialize the population parameters; In a specific exemplary technical solution, when initializing the parameters of the active disturbance rejection attitude control system, the control parameters to be optimized include: the extended state observer , the differential tracker parameters , the nonlinear feedback controller parameters and the global control gain compensation coefficient , and define the parameter range to keep the optimization parameters within a reasonable range during the optimization process.

[0041] In the embodiment of the present invention, the search process is initialized by randomly generating a set of starting solutions within the search space, and the formula for generating the initial population is as follows.

[0042] ; In the formula, represents the position of the th bird (total number is n ) in the th dimension (total number is m ) in the kingfisher population, represents a random number between 0 and 1; , represent the search upper bound and the search lower bound.

[0043] Once the initial population is generated, the fitness function is used to evaluate the fitness value of each individual, and it is evaluated according to its problem-solving ability. The individual with the best fitness value will be selected, and then a new generation of population will be generated.

[0044] Step 3: Improve the kingfisher dynamic update strategy, establish a kingfisher position update rule that balances global and local searches through a variable spiral sine-cosine strategy, and adapt to the requirements of tuning the parameters of the active disturbance rejection controller; Among them, the formula for establishing the kingfisher position update rule is as follows: ; In the formula, represents the position of the kingfisher in the next iteration, represents the current position of the kingfisher, represents a random number that follows a normal distribution, represents the population size; the value of the parameter is dynamically determined according to the current strategy. According to whether is greater than 0.5, "perch" or "hover" is selected. The calculation is tailored for each strategy to ensure optimal performance in different modes. In the perching strategy, its calculation formula is as follows: ; In the formula, where is the maximum number of iterations, represents a constant with a value of 8, is a random number between 0 and 1.

[0045] In the hovering strategy, its calculation formula is as follows.

[0046] ; In the formula, , represent the fitness values of the , th kingfishers.

[0047] An updated formula for the development stage of the kingfisher optimized by the variable spiral sine-cosine algorithm is established. The powerful development ability of the sine-cosine algorithm is utilized to further enhance the development ability of the kingfisher optimization algorithm, and the balance ability of the algorithm in exploration and development is improved by using a spiral line with variable spiral coefficient oscillation. The updated formula for the development stage of the improved kingfisher is as follows: .

[0048] Step 4: Fitness evaluation. Compare the updated fitness values, retain the optimal position and fitness, establish a local escape mechanism, and compare the updated positions; Establish a mechanism for the development stage, and its calculation formula is as follows.

[0049] ; In the formula, is the population average position, is a random number uniformly distributed between 0 and 1; is the preset optimal fitness value; Establish a mechanism for the local escape stage.

[0050] ; In the formula, are the maximum and minimum exploration probabilities, are the positions of two randomly selected individuals at time t.

[0051] Randomly select two individuals from the population, and their positions are represented by and respectively. The predation efficiency of the kingfisher is represented by where and The constant values are set to 0.5 and 0 respectively.

[0052] Step 5: Use the kingfisher optimization algorithm to tune the parameters of the three attitude control loops of the UAV respectively, and compare the verified parameters with the original parameters for verification.

[0053] In the technical solution provided by the embodiment of the present invention, the weight coefficients of the objective function are used to balance the tracking accuracy, dynamic response and stability requirements, ensuring that the optimization objectives are comprehensive. Combining random perturbation and dynamic step size adjustment to avoid falling into local optima and improve the global search ability. According to the ADRC theory and practical experience, reasonably set the parameter search range to reduce ineffective search.

[0054] In the specific exemplary technical solution of the present invention, an active disturbance rejection control system simulation platform is built in MATLAB, which includes an active disturbance rejection controller and a three-loop attitude model; the active disturbance rejection controller dynamically compensates for disturbances based on an extended state observer. Take the active disturbance rejection control parameters as optimization variables, initialize the population of the kingfisher optimization algorithm, each "kingfisher" individual represents a set of active disturbance rejection parameter combinations, and set the search range and the number of iterations. Simulate the diving and predation behavior of the kingfisher, update the active disturbance rejection parameter combinations through random perturbation, and expand the parameter search range. According to the position of the current optimal individual, adjust the parameter step size to finely search for potential optimal solutions in the neighborhood. Substitute each set of parameters into the simulation model, calculate the fitness value, and screen out the active disturbance rejection parameter combinations with better performance. Through the "predation - adjustment" mechanism of the kingfisher, continuously update the active disturbance rejection parameters until the convergence condition (such as the fitness threshold or the maximum number of iterations) is met, and output the optimal active disturbance rejection parameter combination.

[0055] Please refer to Figures 4 to 6 , Figures 4 to 6 For the actual pitch, roll and yaw angle changes of the UAV with an expected attitude step of 0.1 rad when running the simulation according to the optimal parameters after optimization, it can be seen that the active disturbance rejection controller shows faster response speed and higher tracking accuracy in the three attitude channels of pitch, roll and yaw, significantly improving the dynamic performance and control accuracy of the system. This comparison verifies the effectiveness and superiority of the parameter tuning method based on the kingfisher optimization algorithm in the optimization of the attitude controller.

[0056] Finally, it should be noted that: the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit them. Although the present invention has been described in detail with reference to the above embodiments, those of ordinary skill in the art should understand that: modifications or equivalent replacements can still be made to the specific embodiments of the present invention, and any modifications or equivalent replacements that do not depart from the spirit and scope of the present invention should be covered within the protection scope of the claims of the present invention.

Claims

1. A variable pitch multi-rotor UAV, characterized in that the variable pitch multi-rotor UAV is provided with a flight control system for realizing the flight control of the variable pitch multi-rotor UAV; the flight control system is provided with an active disturbance rejection attitude control system for realizing the flight attitude control of the variable pitch multi-rotor UAV in a disturbed environment; wherein, the active disturbance rejection attitude control system is provided with three control loops of roll, pitch and yaw, and each control loop includes an extended state observer, a differential tracker and a non-linear state error feedback; wherein, in each control loop, the optimization steps of the control parameters of the extended state observer, the differential tracker and the non-linear state error feedback include: based on the established objective function, using an improved kingfisher optimization algorithm to optimize the control parameters to obtain the optimization result of the control parameters; wherein, the dynamic update strategy of the improved kingfisher optimization algorithm adopts a kingfisher position update rule based on a variable spiral sine-cosine strategy.

2. The variable pitch multi-rotor UAV according to claim 1, characterized in that the variable spiral sine-cosine strategy tends to global search in the early stage of optimization and tends to local search in the later stage of optimization.

3. The variable pitch multi-rotor UAV according to claim 2, characterized in that a local escape mechanism is set in the process of local search.

4. The variable pitch multi-rotor UAV according to claim 1, characterized in that the expression of the objective function is: ； In the formula, is the total objective function of the control loop; is the root mean square error, which is used to measure the tracking accuracy; is the system settling time, which is used to measure the dynamic response; is the overshoot, which is used to measure the system stability; are all weight coefficients.

5. The variable pitch multi-rotor UAV according to claim 1, characterized in that the steps of using the improved kingfisher optimization algorithm to optimize the control parameters to obtain the optimization result of the control parameters include: according to the preset objective function value constraint condition, using the improved kingfisher optimization algorithm to find the minimum value of the objective function, and after iterative update to meet the preset condition, obtaining the optimization result of the control parameters; wherein, first initialize the kingfisher optimization algorithm population and population parameters, and each individual in the population represents a control parameter to be optimized; then in each iteration, adopt the kingfisher position update rule based on the variable spiral sine-cosine strategy to obtain a new population, and evaluate the fitness value of the population according to the established kingfisher fitness function; finally, after iterative update to meet the preset condition, take the population corresponding to the fitness value closest to the preset optimal fitness value as the optimization result of the control parameters.

6. The variable pitch multi-rotor UAV according to claim 1, characterized in that in the kingfisher position update rule based on the variable spiral sine-cosine strategy, the update formula in the exploitation stage is as follows: ； Wherein, is the position of the individual in the generation; is the position of the individual in the generation; is the variable helix coefficient; are the hybrid adaptive coefficient, exploration factor, and step size coefficient respectively; b is the reference offset; is the reference point or target position of the individual in the generation; is the dynamic scaling factor, which is used to control the step size of the search; is the random angle; and are both random numbers and both follow the uniform distribution on [0, 1].

7. The variable pitch multi-rotor UAV according to claim 6, characterized in that Variable helix coefficient The expression is as follows: ； In the formula, is a parameter for controlling the helix; is a parameter representing the period of the helix; l is a linear attenuation factor; t is the number of iterations; is the maximum number of iterations.

8. The variable pitch multi-rotor UAV according to claim 7, characterized in that In the early stage of optimization, takes the value of 1, which is used to enable the search path to cover a larger range to enhance the global search ability; in the later stage of optimization, as the number of iterations increases, gradually decreases, which is used to narrow the search range to improve the local optimization accuracy.

9. The variable pitch multi-rotor UAV according to claim 1, characterized in that in the optimization steps of the control parameters of the extended state observer, the differential tracker and the non-linear state error feedback in each control loop The control parameters to be optimized for the expansion state observer include the state error feedback gain, the observer nonlinearity intensity parameter, and the threshold parameter for switching between the linear and nonlinear regions of the observer; The control parameters to be optimized for the nonlinear state error feedback include the error feedback gain, the nonlinearity intensity, and the threshold parameter for switching between the linear and nonlinear regions; The control parameters to be optimized for the differential tracker include the filtering factor and the speed factor that determines the speed of the system's tracking signal; In addition, it also includes global parameters to be optimized, and the global parameters include the global control gain compensation coefficient.

10. A flight attitude control method for a variable-pitch multi-rotor UAV according to claim 1, characterized in that, It includes: During the process of implementing the flight control of the variable pitch multi-rotor UAV through the flight control system, the desired pitch angle, the desired yaw angle, and the desired roll angle are input into the active disturbance rejection attitude control system to obtain the rudder deflection value, so as to realize the flight attitude control of the variable pitch multi-rotor UAV in a disturbed environment through the active disturbance rejection attitude control system.

Images